tag:blogger.com,1999:blog-72920731492425065932024-03-12T21:50:03.899-07:00Dr. Alan Reifman's SEM CourseA Resource Page for Graduate Students at Texas Tech University Taking Human Development and Family Studies 6364 (Quantitative Methods III)alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comBlogger49125tag:blogger.com,1999:blog-7292073149242506593.post-54376338931775737452050-01-01T18:19:00.000-08:002018-11-24T11:39:02.357-08:00Welcome!Here are links to our lecture notes on the different course topics...<br />
<br />
Intro: <a href="http://reifman-sem.blogspot.com/2007/01/welcome-to-quantitative-methods-iv.html">The Pyramid of Success</a><br />
<br />
<a href="http://www.webpages.ttu.edu/areifman/hdfs3390relval.html#correlation">Correlation</a>, <a href="http://reifman-sem.blogspot.com/2007/01/heres-photo-from-previous-class-session.html">Least-Squares Principle</a>, and <a href="http://reifman-sem.blogspot.com/2007/01/today-well-go-over-left-side-of-sem.html">Multiple Regression</a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2007/01/below-are-two-photos-from-recent.html">Path Analysis</a><br />
<br />
Exploratory Factor Analysis (<a href="http://reifman-sem.blogspot.com/2007/01/in-wednesdays-class-we-will-work-our.html">here</a>, <a href="http://reifman-sem.blogspot.com/2007/02/as-we-saw-today-promax-oblique-factor.html">here</a>, and <a href="http://reifman-sem.blogspot.com/2010/02/today-well-start-covering-exploratory.html">here</a>)<br />
<br />
Confirmatory Factor Analysis (<a href="http://reifman-sem.blogspot.com/2007/02/im-anticipating-that-well-use-roughly.html">here</a>, <a href="http://reifman-sem.blogspot.com/2007/02/when-learning-sem-important-distinction.html">here</a>, and <a href="http://reifman-sem.blogspot.com/2008/01/during-class-earlier-today-i-encouraged.html">here</a>)...<br />
<br />
...and Associated Basic Concepts (<a href="http://reifman-sem.blogspot.com/2008/02/below-is-photograph-kristina-took-of.html">free/fixed parameters and model identification</a>; <a href="http://reifman-sem.blogspot.com/2007/04/some-of-students-wanted-review-of.html">degrees of freedom</a>; <a href="http://reifman-sem.blogspot.com/2007/02/at-wednesdays-class-well-make-sure.html">model fit</a>; <a href="http://reifman-sem.blogspot.com/2009/03/one-of-our-students-this-semester-susan.html">reporting fit</a>)<br />
<br />
<a href="http://reifman-sem.blogspot.com/2017/02/introduction-to-onyx.html">ONYX Program</a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2013/01/the-semnet-e-mail-discussion-listserv.html">Writing Up SEM/CFA Results</a><br />
<br />
Full Structural Models (<a href="http://reifman-sem.blogspot.com/2007/02/now-that-were-beginning-to-learn-how-to.html">here</a>, <a href="http://reifman-sem.blogspot.com/2008/03/today-id-like-to-cover-interpretive.html">here</a>, and <a href="http://reifman-sem.blogspot.com/2007/04/some-of-students-wanted-review-of.html">here</a>); also see the following article for discussion of what a "model" represents:<br />
<br />
<span style="color: #cc0000;">Rodgers, J. L. (2010). The epistemology of mathematical and statistical modelling. A quiet revolution. <i>American Psychologist, 65</i>, 1-12. </span><br />
<br />
<span style="color: #cc0000;"><span style="color: black;"><a href="http://www.youtube.com/watch?v=EYPapE-3FRw%20%3Chttp://www.youtube.com/watch?v=EYPapE-3FRw&feature=related%3E%20&feature=related">Video clip</a> of legendary physicist Richard Feynman discussing conclusions one can draw from tests of theoretical models.</span></span><br />
<br />
<span style="color: #cc0000;"><span style="color: black;"><a href="http://reifman-sem.blogspot.com/2007/03/as-weve-discussed-part-of-latest.html">Comparative Model Testing and Nestedness</a></span></span><br />
<br />
<a href="http://reifman-sem.blogspot.com/2010/03/now-that-weve-learned-basics-of-full.html"><b>Beyond the Basics of SEM</b></a> (contains all our topics for roughly the second half of the course)<br />
<br />
<a href="http://reifman-sem.blogspot.com/2010/03/following-is-model-for-new-assignment.html">Diagram for Assignment 2</a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2017/04/reminder-of-terminology-in-sem.html">Refresher Diagram</a> on SEM Terminology<br />
<br />
SEM The Musical: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a>, <a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html">5</a>, <a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html">6</a>, <a href="http://reifman-sem.blogspot.com/2013/04/sem-musical-7-april-30.html">7</a>, <a href="http://reifman-sem.blogspot.com/2014/04/sem-musical-8-april-29-2014.html">8</a>, <a href="http://reifman-sem.blogspot.com/2015/04/sem-musical-9.html">9</a>, <a href="http://reifman-sem.blogspot.com/2016/04/sem-musical-10.html">10</a>, <a href="http://reifman-sem.blogspot.com/2017/04/sem-musical-11.html">11</a>, <a href="http://reifman-sem.blogspot.com/2018/04/sem-musical-12.html">12</a>, <a href="http://reifman-sem.blogspot.com/2018/11/sem-musical-125.html">12.5</a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2010/04/on-semnet-discussion-listserv-around.html">Graphic arts programs</a>alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-14697779987242201802018-11-24T11:36:00.001-08:002018-12-19T22:41:29.808-08:00SEM The Musical 12.5<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhOYOhDr1O9UJyURdEJVVz_lnDTFQDMe36J8l4KAPR19IWuNYYVrDEYTjGJ4ddHfuXT0gtmILKJQFQXY39G6ZASIl9FkvGklHL-Sj4aIXuGxCBYOGLzKYF8QE2mkNjc66Ozr27LQkkvink/s1600/sem+musical+12.5.gif" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="501" data-original-width="807" height="247" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhOYOhDr1O9UJyURdEJVVz_lnDTFQDMe36J8l4KAPR19IWuNYYVrDEYTjGJ4ddHfuXT0gtmILKJQFQXY39G6ZASIl9FkvGklHL-Sj4aIXuGxCBYOGLzKYF8QE2mkNjc66Ozr27LQkkvink/s400/sem+musical+12.5.gif" width="400" /></a></div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<br />
We will be performing SEM The Musical <i>12.5</i> this upcoming Thursday, November 29. Why the designation "12.5"? For roughly the last dozen years, SEM had been the fourth course in our HDFS graduate statistics sequence (after Intro, ANOVA/Regression, and Multivariate) and always taught in the spring. However, we revamped the statistics sequence, knocking out the Multivariate course, moving SEM to third in the order, and adding Longitudinal in the fourth position. Starting with the current semester, SEM is now a fall course. Because only six months (rather than 12) have elapsed since the last SEM Musical, we are therefore referring to the upcoming one as 12.5.<br />
<br />
As always, we'll sing some new songs (shown below) and some classics of the previous 12 years. Just click on any of the following numbers to access a prior musical: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a>, <a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html">5</a>, <a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html">6</a>, <a href="http://reifman-sem.blogspot.com/2013/04/sem-musical-7-april-30.html">7</a>, <a href="http://reifman-sem.blogspot.com/2014/04/sem-musical-8-april-29-2014.html">8</a>, <a href="http://reifman-sem.blogspot.com/2015/04/sem-musical-9.html">9</a>, <a href="http://reifman-sem.blogspot.com/2016/04/sem-musical-10.html">10</a>, <a href="http://reifman-sem.blogspot.com/2017/04/sem-musical-11.html">11</a>, <a href="http://reifman-sem.blogspot.com/2018/04/sem-musical-12.html">12</a>.
<br />
<br />
<b>Todd Little Parcels Indicators</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=gr2i0Ld7378">[My Baby Does the] Hanky Panky</a>,” Greenwich/Barry, popularized by Tommy James & the Shondells)<br />
<i>[<a href="https://www.youtube.com/watch?v=eOqKfvvjFck" target="_blank">Video of performance</a>; added 12/4/2018]</i><br />
<br />
Todd Little parcels indicators,<br />
Todd Little parcels indicators,<br />
He’s one of modeling’s innovators,<br />
He checks residuals’ “correlators*,”<br />
Todd Little parcels indicators...<br />
<br />
Todd Little parcels indicators,<br />
Todd Little parcels indicators,<br />
He’s one of modeling’s innovators,<br />
He checks residuals’ “correlators,”<br />
Todd Little parcels indicators...<br />
<br />
You’ve got a bunch of indicators, you know,<br />
You have to decide, how they’re gonna go,<br />
Should you combine them, into smaller sets?<br />
Do so at random or with other intent?<br />
They’re still debating, yeah they’re still debating...<br />
<br />
Todd Little parcels indicators,<br />
Todd Little parcels indicators,<br />
He’s one of modeling’s innovators,<br />
He checks residuals’ “correlators,”<br />
Todd Little parcels indicators...<br />
<br />
(Guitar solo)<br />
<br />
You’ve got a bunch of indicators, you know,<br />
You have to decide, how they’re gonna go,<br />
Should you combine them, into smaller sets?<br />
Do so at random or with other intent?<br />
They’re still debating, yeah they’re still debating...<br />
<br />
Todd Little parcels indicators,<br />
Todd Little parcels indicators,<br />
He’s one of modeling’s innovators,<br />
He checks residuals’ “correlators,”<br />
Todd Little parcels indicators...<br />
<br />
He’s one of modeling’s innovators,<br />
He checks residuals’ “correlators,”<br />
Todd Little parcels indicators ,<br />
Todd Little parcels indicators... (fade out)<br />
<br />
---<br />
*There is, of course, no such term as “correlator.” I made it up to maintain the rhyme. What I’m referring to is how one may choose to combine into a parcel indicators that, while initially separate, show a residual correlation. Little et al. (2013, “Why the items versus parcels controversy needn’t be one,” <i>Psychological Methods</i>) note that: “...when a correlated residual is evident in an item-level solution, the most advantageous parcel solution may be one that aggregates those correlated items together” (p. 290).<br />
<br />
<br />
<b>Oh Mplus!</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=-fufU6fyZAo">Holy War</a>,” Lukather/Vanston/Williams for Toto)<br />
<i>[<a href="https://www.youtube.com/watch?v=Y-IUOmAtLfs" target="_blank">Video of performance</a>; added 12/18/2018]</i><br />
<br />
(Guitar riff four times)<br />
<br />
So, you need, some new, S-E-M software,<br />
From lots of options, you can choose,<br />
I’d say use AMOS, for the basics,<br />
But Mplus, when things, are abstruse,<br />
<br />
Ready, to start,<br />
Your data, must,<br />
Be in plain-text,<br />
With no labels on top,<br />
<br />
Run it, run it,<br />
Then check, warnings,<br />
So you can make sure,<br />
That your run, didn’t stop,<br />
<br />
For more, advanced stuff,<br />
Check out all, the working papers,<br />
Things should, be clear, eventually,<br />
<br />
Oh Mplus!<br />
Yes, you’re such a quirky program,<br />
Adding covs, for which no one asked,<br />
You’ve got quite, a learning curve,<br />
To master, all the details,<br />
It is a, substantial task!<br />
<br />
(Guitar riffs twice)<br />
<br />
Now, if you want latent classes,<br />
Or, multilevel modeling,<br />
Mplus keeps updating, its routines,<br />
So it has got, the things you need,<br />
<br />
You can get help,<br />
In figuring, out, the details,<br />
A book by Geiser’s, crystal clear,<br />
<br />
There also is,<br />
A website, to help you,<br />
Where they do, Q & A,<br />
<br />
Give it a try,<br />
But don’t lose, your patience,<br />
It takes some time, to find your way,<br />
<br />
Oh Mplus!<br />
Yes, you’re such a quirky program,<br />
Adding covs, for which no one asked,<br />
You’ve got quite, a learning curve,<br />
To master, all the details,<br />
It is a, substantial task!<br />
<br />
(Guitar solos)<br />
<br />
Oh Mplus!<br />
Yes, you’re such a quirky program,<br />
Adding covs, for which no one asked,<br />
You’ve got quite, a learning curve,<br />
To master, all the details,<br />
It is a substantial task!<br />
<br />
Oh Mplus!<br />
Yes, you’re such a quirky program,<br />
Adding covs, for which no one asked,<br />
You’ve got quite, a learning curve,<br />
To master, all the details,<br />
It is a substantial task!<br />
<br />alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-10331125446769987142018-04-19T09:28:00.001-07:002018-05-03T09:40:09.920-07:00SEM The Musical 12<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhFpIBSZmFRHfJO0wN3wjB3EO2UCXE3clcPSp3AyT2_gWR1FRY_1THyRA7ctKqMnB_FOxNnJr6OPaAyKgzg4IIGs4biKJj-ed6TvIcF7gOwcud7kM4F1LPnJUQKttbUFDpHFzdOCSVsR90/s1600/SEM+the+Musical+12+logo.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="544" data-original-width="960" height="226" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhFpIBSZmFRHfJO0wN3wjB3EO2UCXE3clcPSp3AyT2_gWR1FRY_1THyRA7ctKqMnB_FOxNnJr6OPaAyKgzg4IIGs4biKJj-ed6TvIcF7gOwcud7kM4F1LPnJUQKttbUFDpHFzdOCSVsR90/s400/SEM+the+Musical+12+logo.jpg" width="400" /></a></div>
<br />
The twelfth annual <i>SEM The Musical</i> will be held on Thursday, May 3, during our class. The above logo was contributed by one of our students, Casey A. Smith. We'll sing new songs (to be listed below as they're written) and some favorites from the previous 11 musicals (links: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a>, <a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html">5</a>, <a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html">6</a>, <a href="http://reifman-sem.blogspot.com/2013/04/sem-musical-7-april-30.html">7</a>, <a href="http://reifman-sem.blogspot.com/2014/04/sem-musical-8-april-29-2014.html">8</a>, <a href="http://reifman-sem.blogspot.com/2015/04/sem-musical-9.html">9</a>, <a href="http://reifman-sem.blogspot.com/2016/04/sem-musical-10.html">10</a>, <a href="http://reifman-sem.blogspot.com/2017/04/sem-musical-11.html">11</a>).<br />
<br />
<b>Welcome to the SEM Parade</b><br />
Lyrics by Jonathan Villarreal<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=pInrJ72eeUU&feature=youtu.be">Welcome to the Black Parade</a>”,Bryar, Iero, Way, Way, & Toro, for My Chemical Romance)
<br />
<br />
When we were, a new class,<br />
Our professor, took us through the basics,<br />
To see a working model,<br />
<br />
He said, class, when you’re finished,<br />
Would you make, a structural equation model,<br />
To find RMSEA,<br />
<br />
He said, will you, connect them,<br />
The pathways, and get degrees of freedom,<br />
And compare your delta chi-squares?<br />
<br />
Because one day, you’ll leave here,<br />
As scholars, to do your own research,<br />
And join the SEM Parade,<br />
<br />
When we were, a new class,<br />
Our professor, took us through the basics,<br />
To see a working model,<br />
<br />
He said, class, when you’re finished,<br />
Would you make, a structural equation model,<br />
To find RMSEA,<br />
<br />
(Drum-led speed-up)<br />
<br />
Sometimes we get the feeling,<br />
These are good factor loadings,<br />
And other times, it feels like it’s all wrong,<br />
<br />
When through it all,<br />
The models we draw,<br />
In Onyx and AMOS,<br />
And when it’s done, we want you all to know,<br />
<br />
We’ll run Mplus, we’ll run Mplus,<br />
And though this class is done, believe us,<br />
We’ll continue to run Mplus,<br />
<br />
We’ll run Mplus,<br />
And three people cannot run it,<br />
Remote desktop will not allow it,<br />
<br />
An error that sends you reeling,<br />
Iterations exceeded,<br />
This model will not run at all,<br />
<br />
So do the math,<br />
And change a path,<br />
Let’s make our syntax clear,<br />
<br />
Triumphant in the end,<br />
We heed the call,<br />
<br />
To run Mplus,<br />
We’ll run Mplus,<br />
And though this class is done, believe us,<br />
We’ll continue to run Mplus,<br />
<br />
We’ll run Mplus,<br />
And through maximum likelihood estimation,<br />
We’ll accept the adjusted,<br />
<br />
Model, and write up our results (oh, oh, oh)<br />
Make sure it’s in STDYX (oh, oh, oh)<br />
<br />
Take a look at analyses,<br />
Cause it does not fit at all,<br />
<br />
The CFI, Is below point 90,<br />
Didn’t calculate, Degrees of freedom,<br />
It’s too low, The Tucker-Lewis,<br />
<br />
We compared it all,<br />
We want to cite our source,<br />
For best fit,<br />
David Kenny,<br />
Don’t forget,<br />
To discuss correlations,<br />
“Causal” paths,<br />
For all our factors,<br />
<br />
List them all,<br />
Or at least if significant,<br />
We’re just a class,<br />
We’re not statisticians,<br />
Just a class, who had to run these tests,<br />
We’re just a class,<br />
We’re not Todd Little,<br />
WE – DID – IT,<br />
<br />
We’ll run Mplus,<br />
We’ll run Mplus,<br />
And though this class is done, believe us,<br />
<br />
We’ll continue to run Mplus,<br />
We’ll run Mplus,<br />
And through maximum likelihood estimation,<br />
We’ll accept the adjusted model,<br />
<br />
The CFI,<br />
Is below point 90,<br />
Didn’t calculate,<br />
Degrees of freedom,<br />
It’s too low,<br />
The Tucker-Lewis,<br />
We compared it all,<br />
We want to cite our source,<br />
<br />
The CFI (we’ll run Mplus),<br />
Is below point 90 (we’ll run Mplus),<br />
Didn’t calculate (we’ll run Mplus),<br />
Degrees of freedom,<br />
It’s too low,<br />
The Tucker-Lewis,<br />
We compared it all,<br />
We want to cite our source (we’ll run Mplus)<br />
<br />
<b>The CFI</b><br />
Lyrics by Alan Reifman<br />
May be sung to the tune of “<a href="https://www.youtube.com/watch?v=r8IahJPQL40">English Eyes</a>” (Kimball/Paich/J. Porcaro /S. Porcaro for Toto)<br />
<br />
What you’ve run, you want to see how well it fits,<br />
Do the known, and the implied r’s, match bit-by-bit?<br />
The NFI, is one way, but it rises just by adding paths,<br />
Can parsimony, be embedded, right there in the <a href="http://www.slate.com/blogs/lexicon_valley/2014/12/08/math_versus_maths_how_americans_and_brits_deploy_the_collective_noun.html">maths</a>?<br />
<br />
It takes account, com-plex-i-ty, CFI,<br />
You want to get, values above point-9-5,<br />
<br />
(Instrumentals)<br />
<br />
It compares, your model to the null version, which has no links,<br />
To ensure, your model fits better, than one you know that stinks,<br />
In the <a href="https://pdfs.semanticscholar.org/f1f2/d2485e7312bd2ea45b7675fd07f811138577.pdf">formula</a>, the df track, how many paths you use,<br />
In this way, the more you saturate, the more you lose,<br />
<br />
It takes account, com-plex-i-ty, CFI,<br />
You want to get, values above point-9-5,<br />
CFI!<br />
CFI!<br />
<br />
(Keyboard/guitar back-and-forth)<br />
<br />
[<i>Slow and quiet:</i><br />
How's your fit?<br />
What indices should you be using now?]<br />
<br />
Of sample-size bias, the CFI is relatively free,<br />
As a fit index, it enjoys great popularity,<br />
It’s in programs, such as AMOS, Onyx, and Mplus,<br />
So you can find it, without going through, any fuss,<br />
<br />
It takes account, com-plex-i-ty, CFI,<br />
You want to get, values above point-9-5,<br />
It takes account, com-plex-i-ty, CFI,<br />
You want to get, values above point-9-5,<br />
CFI!<br />
CFI!<br />
<br />
(More instrumentals)<br />
<br />
CFI!<br />
<br />
(Guitar solo)<br />
<br />
CFI!<br />
<br />
CFI!<br />
<br />
<b>Dr. Cong </b>(pronounced “Tsong” like tsunami)<br />
Lyrics by Alan Reifman<br />
May be sung to the tune of “<a href="https://www.youtube.com/watch?v=6u9vPkVr7CU">Miss Sun</a>” (David Paich, popularized by Boz Scaggs)<br />
<br />
Been teaching stats, a long time,<br />
Since you, came from U-S-C,<br />
You’ve taught, lots of students,<br />
In QM 1, and 2, and 3,<br />
<br />
Dr. Cong, what can we say?<br />
We wish you, all the best, out at U-T-A,<br />
We hope it isn’t long,<br />
Before our paths, will cross again, in some way,<br />
<br />
You’ve served, on our committees,<br />
Methods quals, won’t be the same,<br />
Who’s going to, ask the students,<br />
With a sample, what’s your aim?<br />
<br />
Dr. Cong, what can we say?<br />
We wish you, all the best, out at U-T-A,<br />
We hope it isn’t long,<br />
Before our paths, will cross again, in some way (Cross again in some way)<br />
<br />
(Guitar solo)<br />
<br />
Dr. Cong, what can we say?<br />
You’ve been a, friend of ours, for 10 years, every day,<br />
We hope it isn’t long,<br />
Before our paths, will cross again, in some way,<br />
...In some way...<br />
<br />
(Brief interlude)<br />
<br />
Dr. Cong, what can we say?<br />
We wish you, all the best, out at U-T-A,<br />
We hope it isn’t long,<br />
Before our paths, will cross again, in some way...<br />
<br />
(Instrumentals)<br />
<br />alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-67267299744067076902017-04-30T20:34:00.004-07:002017-05-05T15:49:18.376-07:00SEM The Musical 11<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgAcfl-zOfuJXC7sIGocv6jKuXuE4lqa92_L__YeuSd19UeaQCW9dtxmau0hnZ5zvjOSW5B_X_hrZ81q1sNN-4aVapjKjawTaeuAX9jBO_m3Y3ov8R1Zc96qXGk0HrSyFGfNbNtO_kJUmE/s1600/sem_musical_11_logo.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgAcfl-zOfuJXC7sIGocv6jKuXuE4lqa92_L__YeuSd19UeaQCW9dtxmau0hnZ5zvjOSW5B_X_hrZ81q1sNN-4aVapjKjawTaeuAX9jBO_m3Y3ov8R1Zc96qXGk0HrSyFGfNbNtO_kJUmE/s320/sem_musical_11_logo.jpg" width="137" /></a></div>
<div style="text-align: left;">
<br />
<i>(Updated May 5, 2017)</i><br />
<br />
The eleventh annual <i>SEM The Musical</i> was held Thursday, May 4, during our class. We had two new songs this year, one by Dr. Reifman and one by student Derrick Holland. Derrick's song keeps our streak alive of having at least one student-written song every year. We also, of course, sang a bunch of favorites from the previous ten musicals (links: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a>, <a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html">5</a>, <a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html">6</a>, <a href="http://reifman-sem.blogspot.com/2013/04/sem-musical-7-april-30.html">7</a>, <a href="http://reifman-sem.blogspot.com/2014/04/sem-musical-8-april-29-2014.html">8</a>, <a href="http://reifman-sem.blogspot.com/2015/04/sem-musical-9.html">9</a>, <a href="http://reifman-sem.blogspot.com/2016/04/sem-musical-10.html">10</a>). See below for this year's new songs...<br />
<br />
<b>Why Won’t It Run? </b><br />
Lyrics by Alan Reifman<br />
May be sung to the tune of “<a href="https://www.youtube.com/watch?v=XkvqYxdSd20">On the Run</a>” (Lukather/Paich/Waybill for Toto)<br />
<br />
SEM involves, some complex math,<br />
Before you go, you need to check, you’ve added each path,<br />
Lots of little details, for you to keep in sight,<br />
Formatting the data, and making sure, your syntax is right,
<br />
<br />
Have you verified, what you’ll fix to one?<br />
Otherwise, your troubles, have just begun,
<br />
<br />
Why won’t it run, why all these error signs?<br />
Why won’t it run? Only hints, of what could’ve, gone wrong,<br />
Why won’t it run? Check your steps, line-by-line,<br />
You can put, all your angst, into song!
<br />
<br />
(Instrumental)
<br />
<br />
You never know, what a new model, can bring,<br />
Punctuation, constraints, it could be anything,<br />
Maybe what you have, is a Heywood Case?<br />
You’ll need a sharp eye, to keep things in place,
<br />
<br />
Maximum likelihood, seeks a minimum to achieve,<br />
Gonna take a miracle, for you to receive,
<br />
<br />
Oh, oh, oh, why won’t it run, why won’t the steps converge?<br />
Why won’t it run, are the magnitude scales far apart?<br />
Why won’t it run, when will I be, on the verge?<br />
Doing this, can tax your heart!
<br />
<br />
(Instrumental and Guitar Solo)
<br />
<br />
<b>Hungry Like a Low Chi-Square</b><br />
Lyrics by Derrick Holland<br />
May be sung to the tune of "<a href="https://www.youtube.com/watch?v=IxW0n5Fe5CY">Hungry Like a Wolf</a>" (Duran Duran)<br />
<br />
Open the connection, get ready to run,<br />
Make sure variables, are in a dot-dat file,<br />
Do do do do do do do dodo dododo dodo,<br />
<br />
List, all your variables, that you will use,<br />
Make sure all missing variables, are -99*,<br />
Do do do do do do do dodo dododo dodo,<br />
<br />
All pathways are free,<br />
Unless you fix a path to 1,<br />
The very end goal,<br />
It’s important you know,<br />
And I'm hungry, like a low chi-square,<br />
<br />
Straddle the line,<br />
With comparative models,<br />
I'm on the hunt, for a good CFI,<br />
Check your TLI, and RMSEA,<br />
And I'm hungry, like a low chi-square, <br />
<br />
You get an error, so you start to freak out,<br />
Mplus tells you, that variables are not defined,<br />
Do do do do do do do dodo dododo dodo,<br />
<br />
You get a low CFI, important paths are behind,<br />
You search in theory, for paths that are not benign, <br />
Do do do do do do do dodo dododo dodo, <br />
<br />
All pathways are free,<br />
Unless you fix one path to 1,<br />
The very end goal,<br />
It’s important you know,<br />
And I'm hungry, like a low chi-square,<br />
<br />
Straddle the line,<br />
With comparative models,<br />
I'm on the hunt, for a good CFI,<br />
Check your TLI, and RMSEA,<br />
And I'm hungry, like a low chi-square, <br />
<br />
Searching for paths,<br />
I break from theory,<br />
I'm on the hunt,<br />
But I won’t get pubbed,<br />
<br />
Latent constructs, made up of manifest<br />
And I'm hungry like low chi-squares,<br />
<br />
Draw many lines,<br />
If you use ONYX,<br />
I'm on the hunt, for a good CFI,<br />
Check your TLI, and RMSEA,<br />
And I'm hungry, like a low chi-square,<br />
<br />
---<br />
<br />
*This is a specification in the Mplus program</div>
alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-60679777941291738902017-04-29T21:24:00.001-07:002017-04-29T21:24:34.800-07:00Partial Least Squares (Small-Sample Alternative to Conventional SEM)Partial Least Squares (PLS) is a variation on Structural Equation Modeling (SEM). Riou, Guyon, and Falissard (2016) state that, relative to conventional SEM, PLS “is more suitable to … work with smaller sample sizes.” PLS is recommended for exploratory purposes, and is often used with single-indicator constructs. The technique seems to be used predominantly within the field of Management Information Systems (MIS).<br />
<br />
Significance testing is done through bootstrapping, with 100 random variations of the original data set being generated and the model rerun in each random data set. An actual path coefficient from one’s model can then be evaluated for extremity, relative to the distribution of the same coefficient estimated 100 times from the bootstrap.<br />
<br />
Though PLS may have reputation for making it easier to obtain significant results, this view appears overstated; a simulation study found that “for N = 40, PLS had 3% and 1% higher power than regression for strong and medium effect sizes [and…] the same power as regression at weak effect size” (Goodhue, Lewis, & Thompson, 2006).<br />
<br />
Fit indices, such as NFI, CFI, RMSEA, are not available.<br />
<br />
<a href="http://warppls.com/">WarpPLS</a> (Kock, 2015) is a program I've found useful and that has a three-month free trial version. Note that the probabilities given in WarpPLS output are one-tailed, so that if you want to report two-tailed <i>p</i>-values, you must double the printed value (e.g., <i>p</i> = .02 one-tailed represents <i>p</i> = .04 two-tailed).<br />
<br />
Discussion of the pros and cons of PLS, and of the circumstances for which it may -- or may not -- be appropriate, is available in Goodhue, Thompson, and Lewis (2013); Marcoulides, Chin, and Saunders (2009); McIntosh, Edwards, and Antonakis (2014); and other sources. See also this <a href="https://groups.google.com/forum/#!msg/pls-sem/VbezAqfqoCI/9W3rdfUbXucJ">discussion piece</a> by Kock.<br />
<br />
<b>References</b><br />
<br />
Goodhue, D., Lewis, W., & Thompson, R. 2006. “PLS, small sample size and statistical power in MIS research,” in Proceedings of the 39th Hawaii International Conference on System Sciences, R. Sprague Jr. (ed.), Los Alamitos, CA: IEEE Computer Society Press. (<a href="https://www.computer.org/csdl/proceedings/hicss/2006/2507/08/250780202b.pdf">link</a>)<br />
<br />
Goodhue D. L., Thompson R. L., & Lewis W. (2013). Why you shouldn’t use PLS: Four reasons to be uneasy about using PLS in analyzing path models. In 46th Hawaii International Conference on System Sciences (pp. 4739–4748). Wailea, HI: HICSS.
<br />
<br />
Kock, N. (2015). <i>WarpPLS 5.0 User Manual</i>. Laredo, TX: ScriptWarp Systems. (<a href="http://warppls.com/">link</a>) <br />
<br />
Marcoulides, G. A., Chin, W. W., & Saunders, C. (2009). A critical look at partial least
squares modeling. MIS Quarterly, 33(1), 171-175.
(<a href="https://www.researchgate.net/publication/253551778_A_Critical_Look_at_Partial_Least_Squares_Modeling">link</a>)<br />
<br />
McIntosh, C. N., Edwards, J. R., & Antonakis, J. (2014). Reflections on partial least squares path modeling. <i>Organizational Research Methods, 17</i>, 210-251. (<a href="http://journals.sagepub.com/doi/abs/10.1177/1094428114529165">abstract</a>)<br />
<br />
Riou, J., Guyon, H., & Falissard, B. (2016). An introduction to the partial least squares approach to structural equation modelling: A method for exploratory psychiatric research. <i>International Journal of Methods in Psychiatric Research, 25</i>,<i> </i>220-231. Published online first at doi: 10.1002/mpr.1497.alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-10512940011321582442017-04-04T07:34:00.000-07:002017-04-04T07:34:25.692-07:00Reminder of Terminology in an SEMTo the beginning SEM practitioner, terms such as "parameter," "factor loading," and "directional path" may be confusing. Here's a drawing on the whiteboard (with some touch-ups in PowerPoint) to help clarify proper usage. Thanks to the students who photographed the board!<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj8o1WvbWWwa2Sz17QCbEtS0-WF7p27CdmSOCYAaWnAKxBNRTP9A3FpjHvrOd9fxUS20A3gsab7x-zg6fNtYRfuSTSFVraeYZU5FLF_uiJmRNRx0dBnkBIJrwD6cODPnV4-l7aE-_ZhVWo/s1600/SEM+terminology+review.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="305" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj8o1WvbWWwa2Sz17QCbEtS0-WF7p27CdmSOCYAaWnAKxBNRTP9A3FpjHvrOd9fxUS20A3gsab7x-zg6fNtYRfuSTSFVraeYZU5FLF_uiJmRNRx0dBnkBIJrwD6cODPnV4-l7aE-_ZhVWo/s400/SEM+terminology+review.jpg" width="400" /></a></div>
<br />alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-19940524408820868292017-02-18T20:24:00.001-08:002018-09-26T22:07:31.210-07:00Introduction to ONYXONYX is a <a href="http://onyx.brandmaier.de/">free SEM package</a> developed in Germany. We will use it for Assignment 1, a CFA on the Hendrick and Hendrick love styles. ONYX is a graphic-arts-based program (like the commercial product AMOS), so your first experience designing a structural equation model will involve what I hope is an intuitive approach of drawing a picture (before we switch to the more technical, but more broadly applicable, Mplus for later assignments). Here are some tips I have come up with for using ONYX, given its differences from other SEM programs:<br />
<br />
1. Everything is done through right-clicking to bring up menus.<br />
<br />
2. You can use an SPSS data file or a plain-text (tab-delimited) ".dat" file saved from an SPSS data file. The ONYX user manual lists available options for designating missing data. Once you've drawn your model, you can use "Load Data" to connect to the .dat file, yielding what's called a "Data Panel."<br />
<br />
3. Use the "Create Variable" option to generate either latent or observed variables.<br />
<br />
4. You should name latent variables (in ALL CAPITALS) via the right-clicking. However, but you’ll have to drag in the measured variables from the "Data Panel" to the variables' respective boxes in the model. By hovering over the measured-variable boxes, you can verify that the data have been linked. <br />
<br />
5. By right-clicking on top of a variable, you can use the "Add Path" tool (the default is to draw unidirectional "causal" paths, whereas holding down the Shift key while using "Add Path" yields dual-headed correlational arrows).<br />
<br />
6. All unstandardized factor loadings start out fixed at 1; you should free all of them (i.e., letting them take on freely estimated values). To identify the model (i.e., make sure you're not estimating more quantities than you have information for), construct variances should be fixed to 1.*<br />
<br />
7. The default settings yield an unstandardized solution, whereas usually we're interested in a standardized one. You can obtain a standardized solution by right-clicking on each indicator’s box and selecting “z-score Transform.”<br />
<br />
8. Covariances (correlations between factors) are also fixed and should be freed. <br />
<br />
9. Unlike other programs, which have you submit a "job" or a "run," ONYX is constantly running in the background and responds to changes you make in model specifications.
Right-clicking and selecting “Show Estimate Summary” will show current results.<br />
<br />
10. The following article provides a concise summary of ONYX, including the claim that its method of seeking a solution is superior to that of other programs (see Figure 6 and the text beginning on the prior page at "Multiple Optima").<br />
<br />
<span style="color: blue;">von Oertzen, T., Brandmaier, A. M., & Tsang, S. (2015). Structural equation modeling with Ωnyx, <i>Structural Equation Modeling, 22</i>,
148-161. </span><br />
<span style="color: blue;"><br /></span>
---<br />
*Fixing (or constraining) variables and (under)identification are discussed <a href="http://reifman-sem.blogspot.com/2008/02/below-is-photograph-kristina-took-of.html">here</a>.alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-12089614707969687252016-05-25T20:43:00.000-07:002016-05-25T16:56:50.735-07:00SEM The Musical 10<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhBz5f_k-8vkJf43rxmAJc3NlJfBvrJvZ9XGhe_VGTeV6pWiEfLKc_My4pm5KL04UBvBVHtVbktRPjy8HIy4dIGjRtF7L5ZIXw-Xby7QJaCDQSXAfQLqdf1vgjqMrpC18tRfmZyl6MuYbg/s1600/SEM+Musical+X.JPG" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="200" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhBz5f_k-8vkJf43rxmAJc3NlJfBvrJvZ9XGhe_VGTeV6pWiEfLKc_My4pm5KL04UBvBVHtVbktRPjy8HIy4dIGjRtF7L5ZIXw-Xby7QJaCDQSXAfQLqdf1vgjqMrpC18tRfmZyl6MuYbg/s200/SEM+Musical+X.JPG" width="157" /></a></div>
<br />
<div class="separator" style="clear: both; text-align: center;">
</div>
The tenth annual SEM The Musical was held on Thursday, May 5. We performed a few new songs this year, as shown below. We also performed songs from previous SEM Musicals. Three older songs we performed this year are available on YouTube (thanks to SH for filming). These songs are "<a href="https://www.youtube.com/watch?v=4xQyS8U1UWc">Common Model Mistakes</a>" (originally from SEM The Musical 9), "<a href="https://www.youtube.com/watch?v=2ZaPJcxoxN0">Saturated Your Model</a>" and "<a href="https://www.youtube.com/watch?v=lAtRdJkBy9E">If You Wanna Join My Construct (You've Gotta Load with My Friends)</a>," the latter two from SEM The Musical 8. To see the lyrics from these (and other) older songs, just click on the year number of the musical: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a>, <a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html">5</a>, <a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html">6</a>, <a href="http://reifman-sem.blogspot.com/2013/04/sem-musical-7-april-30.html">7</a>, <a href="http://reifman-sem.blogspot.com/2014/04/sem-musical-8-april-29-2014.html">8</a>, <a href="http://reifman-sem.blogspot.com/2015/04/sem-musical-9.html">9</a>.<br />
<br />
<br />
<b>SEM Musical <span style="color: red;">TEN</span><span style="color: #333333;">!</span></b><br />
Lyrics by Alan Reifman (retread from previous years)<br />
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=IKqV7DB8Iwg">Let’s Get it Started</a>,” Will Adams et al. for the Black Eyed Peas)<br />
<br />
(Softly) The models keep runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and...<br />
<br />
We’re back again, to have some fun,<br />
We’re gonna bust some rhyme, have a good time,<br />
We’re gonna sing some songs, about SEM technique,<br />
Access your inner geek, let your voices speak,<br />
<br />
SEM is different, your measurement model’s explicit,<br />
The whole model, gets tested for fit,<br />
Is it identified? We know how hard you’ve tried,<br />
Knowns and unknowns, side by side,<br />
It takes you on a ride, finally you’re satisfied,<br />
Your output’s now just fine, you’ve arrived, you can take pride…<br />
<br />
NFI, TLI, CFI,<br />
Calculate estimates, let it run, have some fun, yeah…<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
Yeah...<br />
<br />
Build your constructs, get this straight,<br />
Make sure the indicators, correlate,<br />
Draw your pathways, residuals too,<br />
Don’t leave out, the fixed 1 value,<br />
<br />
Take your time, think it through,<br />
Don’t worry if you’re new, we’ll walk with you,<br />
Step by step, right up the pyramid,<br />
For SEM, we’re really groovin,’<br />
Hope you get an acceptable solution,<br />
Submit your model and get it movin,’<br />
<br />
NFI, TLI, CFI,<br />
Calculate estimates, let it run, have some fun, yeah…<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">TEN</span>!), SEM Musical (HERE!),<br />
Yeah...<br />
<br />
<br />
<b>The Part That’s Error-Free </b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=60eqU0V0jTI">Biggest Part of Me</a>,” David Pack for Ambrosia)<br />
<br />
Boxes, they hold the manifestations,<br />
Bubbles, are error locations,<br />
Constructs, house the shared variation,<br />
They're the part, that’s error-free,<br />
<br />
Loadings, show measures, are correlated,<br />
That makes, indicators validated,<br />
Errors, in the bubbles, they are gated,<br />
So constructs, are error-free,<br />
<br />
Well...<br />
You remove error,<br />
And the paths, become more true*,<br />
This is such, a key thing,<br />
Latent constructs, do for you,<br />
<br />
So draw it now,<br />
Tell measurement error, to shoo.<br />
You can estimate, the paths,<br />
Without error, troubling you,<br />
<br />
Sometimes, you have just, total-scale measures,<br />
Of those, certain constructs, that you treasure,<br />
Alpha, gives a way to block displeasure,<br />
Controls, unreliability,<br />
<br />
Parcels, a technique, that can’t be plainer,<br />
Items, placed into, random containers,<br />
These sets, can then serve, as indicators,<br />
Constructs now, are error-free,<br />
<br />
Well-l-l-l-l...<br />
You remove error,<br />
And the paths, become more true,<br />
This is such, a key thing,<br />
Latent constructs, do for you,<br />
<br />
So draw it now,<br />
Tell the measurement error, to shoo,<br />
You can estimate, the paths,<br />
Without error, troubling you,<br />
<br />
(Instrumentals)<br />
<br />
It’s an SEM hallmark,<br />
Going back to CFA,<br />
It’s a major advantage, of using LV’s,<br />
<br />
Not all techniques, give you this,<br />
Measurement error, doesn’t go away,<br />
So use latent constructs, to be error-free,<br />
Be error-free,<br />
Be error-free...<br />
<br />
<span style="color: red;">*Stephenson, M. T., & Holbert, R. L. (2003). A Monte Carlo simulation of observable versus latent variable structural equation modeling techniques. <i>Communication Research, 30</i>, 332-354.</span><br />
<span style="color: red;"><br /></span>
<span style="color: red;">See also previous lecture modules <a href="http://reifman-sem.blogspot.com/2007/02/im-anticipating-that-well-use-roughly.html">here</a> and <a href="http://reifman-sem.blogspot.com/2012/04/we-recently-discussed-how-to-handle.html">here</a>.</span><br />
<br />
<br />
<b>Those Kinds of Paths (Are Autoregressive)</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=SH8Ik9xh5JM">Because the Night</a>,” Springsteen/Smith)<br />
<br />
Panel models, longitudinally,<br />
Follow the same people, over time,<br />
Each major construct, we include repeatedly,<br />
It gets us the time-ordering, of causality,<br />
<br />
So, come on now, no hand-calculated math,<br />
In cross-lagged models, we run paths,<br />
From Construct A at one time, to B at the next,<br />
We also have paths, from the same construct,<br />
Time 1 to Time 2, and Time 2 to Time 3,<br />
<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
<br />
Autoregressive paths, play a crucial role,<br />
They control for earlier levels, of a later DV,<br />
So when a cross-lagged path, is significant,<br />
It shows association, beyond stability,<br />
<br />
So, come on now, no hand-calculated math,<br />
In cross-lagged models, we run paths,<br />
From Construct A at one time, to B at the next,<br />
We also have paths, from the same construct,<br />
Time 1 to Time 2, and Time 2 to Time 3,<br />
<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
<br />
(Guitar solo)<br />
<br />
These kinds, of paths,<br />
Predict, to later versions, of themselves,<br />
Without them, analyses would lack rigor,<br />
So include them...<br />
<br />
Time 1 to Time 2, Time 2 to Time 3,<br />
Time 1 to Time 2, Time 2 to Time 3,<br />
Time 1 to Time 2, Time 2 to Time 3,<br />
<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability,<br />
Those kinds of paths, are auto-regressive,<br />
Those kinds of paths, test stability...<br />
<br />
<br />
<b>Constructs (Don’t be Afraid of Changing!)</b><br />
Lyrics by Diane Wittie<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=H_CwT7p8-e8">Landslide</a>,” Stevie Nicks)<br />
<br />
Gathered the data, and they abound,<br />
I cleaned them up, then I went to town,<br />
And I saw some variables, that looked interesting,<br />
And now my sleep, would be sound,<br />
<br />
Oh, yes I can begin, naming latent constructs,<br />
But will those, constructs make any sense?<br />
Will they adequately represent,<br />
What I envision?<br />
Can I implement, my central concepts?<br />
<br />
Uh-hum, I do think so,<br />
<br />
Well, don’t be, afraid of changing,<br />
’Cause your constructs, need to make sense,<br />
Think through, your decisions,<br />
You may need, revisions,<br />
Don't do anything, you'll rue,<br />
<br />
(Brief guitar)<br />
<br />
So, don’t be, afraid of changing,<br />
’Cause your constructs, need to make sense,<br />
Think through, your decisions,<br />
You may need, revisions,<br />
Don't do anything, you'll rue,<br />
To your theory, be true,<br />
<br />
So, analyze your data, see what you've found,<br />
Your model, may earn great renown!<br />
If you see factor loadings, at plus or minus .4,<br />
Well maybe, high points you will score,<br />
If you see structural paths, that are significant,<br />
Yes, high points, you will score!alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-2136936975501174052016-04-07T10:31:00.000-07:002020-02-25T13:21:09.305-08:00Mediational Models<i>(<span style="color: red;">Updated February 25, 2020</span>)</i><br />
<br />
Many SEM-based studies examine <i>mediation</i> between variables. To mediate is to go in the middle, like a negotiation mediator comes between the labor union and management.<br />
<br />
In statistical analysis, we often start out with a relationship between two variables. Using an example from one of my grad-school mentors, Patricia Gurin, cigarette smoking and lung cancer are positively associated.<br />
<br />
<div style="text-align: center;">
<b>Cigarette Smoking ==> Lung Cancer</b></div>
<br />
Why does this relationship exist? A more fine-grained understanding would be that smoking leads to lung tissue damage, and tissue damage leads to cancer. Tissue damage would thus be considered the mediator or mechanism.<br />
<br />
<div style="text-align: center;">
<b>Cigarette Smoking ==> <span style="color: red;">Tissue Damage</span> ==> Lung Cancer</b></div>
<br />
Reuben Baron and David Kenny published an <a href="http://psycnet.apa.org/psycinfo/1987-13085-001">article</a> in 1986 on mediation that has been cited nearly 80,000 times (as of November 2018)! Kenny summarizes the process in a nutshell <a href="http://davidakenny.net/cm/mediate.htm">here</a>. In the following figure, I apply Baron and Kenny's "old school" method to Gurin's example. Note that one would<span style="color: red;"><b> run the model twice</b></span>.<br />
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiTmki1KvUFU279GVmL3nXFxhyphenhyphenq_Q5YHu0kUwHVVav_IrvFbgv46bN4_z6qw1R5YIGyC7VMXilZamiYMoR4fUZ7v4amhA-ntSzZrzzgLiiO1KG9eRmQX0KOPxUp2zRyqWZmkr8mZb6anhc/s1600/mediation+example.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="225" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiTmki1KvUFU279GVmL3nXFxhyphenhyphenq_Q5YHu0kUwHVVav_IrvFbgv46bN4_z6qw1R5YIGyC7VMXilZamiYMoR4fUZ7v4amhA-ntSzZrzzgLiiO1KG9eRmQX0KOPxUp2zRyqWZmkr8mZb6anhc/s400/mediation+example.jpg" width="400" /></a></div>
<br />
<i>(Illustration of Baron and Kenny's, 1986, logic. Example from Patricia Gurin, University of Michigan, circa 2002-2003, <a href="http://diversity.umich.edu/admissions/research/gurin.html">link</a>)</i><br />
<br />
The above diagram presents the scenario of <i>full</i> mediation (i.e., the initially significant direct path from antecedent to outcome becomes nonsignificant). One can then say that the mediator accounts fully for the antecedent-outcome relationship. If the initial direct path from antecedent to outcome remains significant after addition of the two mediational paths, but the initial direct path is <i>reduced in magnitude</i>, one can claim <i>partial</i> mediation (see Huselid and Cooper, 1994, "<a href="http://www.ncbi.nlm.nih.gov/pubmed/7822560">Gender roles as mediators of sex differences in expressions of pathology</a>").<br />
<br />
<i>(As noted above, my perspective is that the antecedent and outcome should first be shown to relate significantly, before one pursues the further steps to test mediation. For an opposing view, that a significant antecedent-outcome path should not be a "gatekeeper" for mediation analyses, see <a href="https://stats.stackexchange.com/questions/185626/what-if-path-c-isnt-significant-but-paths-a-and-b-are-indirect-effect-in-medi">here</a>.)</i><br />
<br />
As Kenny writes on his website, "More contemporary analyses focus on the indirect effect." The leading names associated with contemporary mediational analysis are <a href="http://afhayes.com/index.html">Andrew Hayes</a> and <a href="http://quantpsy.org/pubs.htm">Kristopher Preacher</a>, who indeed emphasize indirect effects. The indirect effect can be calculated by multiplying the standardized paths from antecedent to mediator, and from mediator to outcome (think back to our unit on <a href="http://reifman-sem.blogspot.com/2007/01/below-are-two-photos-from-recent.html">path-analysis tracing rules</a>).<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgia-Qi3CP1TL1TMn4dE0l9qfEZBWjDfqgu0mfCLE-qIN1r1t1u4gSXqDkcP5KvepTq4I-AUjiWQTob3fNNJtxIxdYGB78dOCys-iA-S7XWTWnrkzoXLMFvWFTGhQEgeBJ0MGT9kA90E9A/s1600/mediation+example+2.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="82" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgia-Qi3CP1TL1TMn4dE0l9qfEZBWjDfqgu0mfCLE-qIN1r1t1u4gSXqDkcP5KvepTq4I-AUjiWQTob3fNNJtxIxdYGB78dOCys-iA-S7XWTWnrkzoXLMFvWFTGhQEgeBJ0MGT9kA90E9A/s400/mediation+example+2.jpg" width="400" /></a></div>
<br />
The indirect effect is .15 in the above example. If each of the two segments of the indirect effect (A to M, and M to O) is each statisically significant (i.e., different from zero), we would be confident that the indirect effect also is significant. As Hayes (2009, "<a href="http://www.tandfonline.com/doi/full/10.1080/03637750903310360">Beyond Baron and Kenny: Statistical mediation analysis in the new millennium</a>") notes, however, "it is possible for an indirect effect to be detectably different from zero even though one of its constituent paths is not." What is called for is a significance test of the indirect effect of .15 (or whatever value one has).<br />
<br />
The problem is that there is no existing theoretical distribution such as the <i>z</i>, <i>t</i>, <i>F</i>, or chi-square distribution to judge the statistical significance of an indirect effect (i.e., whether or not one's obtained indirect effect falls in the upper or lower 2.5% of the distribution for a two-tailed <i>p</i> < .05 significance level). Therefore, researchers use a "synthetic" statistical distribution for testing the significance of indirect effects, known as a "bootstrap" distribution. Kenny discusses this on his website and it is also illustrated in slide 6 of this <a href="http://slideplayer.com/slide/8340508/">slideshow</a>.<br />
<br />
<div style="color: #cc0000;">
<span style="color: black;">An additional source for studying mediation in SEM is:</span></div>
<div style="color: #cc0000;">
<br /></div>
<div>
<div style="color: #cc0000;">
Li, S. (2011). Testing mediation using multiple regression and structural modeling analyses in secondary data. <i>Evaluation Review, 35</i>, 240-268.</div>
<div style="color: #cc0000;">
<br /></div>
One sometimes posits multiple mediators in a model. For an illustration of what are known as <i>parallel </i>mediation (i.e., the mediators have no causal relations amongst themselves) and <i>serial</i> mediation (i.e., one mediator causes another mediator and so on, in a sequence), see:<br />
<br />
<span style="color: #cc0000;">Guthrie Yarwood, M. F. (2013).The relationship between love styles and digital dating outcomes: A multiple mediation test of the Perceived Importance of Dating Features Scale. Dissertation, Texas Tech University (</span><a href="https://ttu-ir.tdl.org/handle/2346/50725">LINK</a><span style="color: #cc0000;">).</span> See especially pp. 56-62.<br />
<br />
I recently discovered a <a href="http://web.pdx.edu/~newsomj/semclass/ho_mediation.pdf" target="_blank">worksheet</a> from Jason Newsom that contains a nice four-step chart for implementing Baron and Kenny's framework for mediation.</div>
alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-65060596030223748422015-04-16T06:43:00.001-07:002016-05-25T17:13:39.256-07:00SEM The Musical 9<div class="separator" style="clear: both; text-align: center;">
</div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg1qVOV7jyFaFzgEQforOa6JUXscaVbH018UDgTk_ji2irvV_6HAZnLxotpCN2ZRguKSnlR5xzxKtm2KWg46ae46xY69iB2KOUp_eubrdtMueDlPbvrGgjvxztgT9pN-JPYF-Lf0zWBGOc/s1600/SEM+musical+9.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="400" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg1qVOV7jyFaFzgEQforOa6JUXscaVbH018UDgTk_ji2irvV_6HAZnLxotpCN2ZRguKSnlR5xzxKtm2KWg46ae46xY69iB2KOUp_eubrdtMueDlPbvrGgjvxztgT9pN-JPYF-Lf0zWBGOc/s1600/SEM+musical+9.jpg" width="256" /></a></div>
<br />
<br />
Our ninth annual SEM The Musical was held on April 30, 2015. We performed some <span style="background-color: white; font-family: "georgia" , serif; font-size: 13.3333330154419px;">new songs this year, as shown below. We also performed songs from previous SEM Musicals (links: </span><a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.3333330154419px; text-decoration: none;">1</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.3333330154419px;">, </span><a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.3333330154419px; text-decoration: none;">2</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.3333330154419px;">, </span><a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.3333330154419px; text-decoration: none;">3</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.3333330154419px;">, </span><a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.3333330154419px; text-decoration: none;">4</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.3333330154419px;">, </span><a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.3333330154419px; text-decoration: none;">5</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.3333330154419px;">, </span><a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.3333330154419px; text-decoration: none;">6</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.3333330154419px;">, </span><a href="http://reifman-sem.blogspot.com/2013/04/sem-musical-7-april-30.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.3333330154419px; text-decoration: none;">7</a>, <a href="http://reifman-sem.blogspot.com/2014/04/sem-musical-8-april-29-2014.html">8</a>).<br />
<br />
<b style="background-color: white; color: #333333; font-family: Georgia, serif; font-size: 13px;">SEM Musical </b><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span><span style="color: #333333;">!</span></b><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Lyrics by Alan Reifman (retread from previous years)</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">(May be sung to the tune of “</span><a href="http://www.youtube.com/watch?v=IKqV7DB8Iwg" style="background-color: white; color: #00cc66; font-family: Georgia, serif; font-size: 13px; text-decoration: none;">Let’s Get it Started</a><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">,” Will Adams et al. for the Black Eyed Peas)</span><br />
<br style="background-color: white; color: #333333; font-family: Georgia, serif; font-size: 13px;" />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">(Softly) The models keep runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and...</span><br />
<br style="background-color: white; color: #333333; font-family: Georgia, serif; font-size: 13px;" />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">We’re back again, to have some fun, </span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">We’re gonna bust some rhyme, have a good time,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">We’re gonna sing some songs, about SEM technique, </span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Access your inner geek, let your voices speak,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM is different, your measurement model’s explicit, </span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">The whole model, gets tested for fit, </span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Is it identified? We know how hard you’ve tried,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Knowns and unknowns, side by side, </span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">It takes you on a ride, finally you’re satisfied, </span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Your output’s now just fine, you’ve arrived, you can take pride…</span><br />
<br style="background-color: white; color: #333333; font-family: Georgia, serif; font-size: 13px;" />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">NFI, TLI, CFI, </span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Calculate estimates, let it run, have some fun, yeah…</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Yeah,</span><br />
<br style="background-color: white; color: #333333; font-family: Georgia, serif; font-size: 13px;" />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Build your constructs, get this straight,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Make sure the indicators, correlate,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Draw your pathways, residuals too,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Don’t leave out, the fixed 1 value,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Take your time, think it through,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Don’t worry if you’re new, we’ll walk with you,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Step by step, right up the pyramid,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">For SEM, we’re really groovin,’</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Hope you get an acceptable solution,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Submit your model and get it movin,’</span><br />
<br style="background-color: white; color: #333333; font-family: Georgia, serif; font-size: 13px;" />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">NFI, TLI, CFI,</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Calculate estimates, let it run, have some fun, yeah…</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">SEM Musical (</span><b style="background-color: white; font-family: Georgia, serif; font-size: 13px;"><span style="color: red;">N</span><span style="color: red;">INE</span></b><span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">!), SEM Musical (HERE!),</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;">Yeah…</span><br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;"><br /></span>
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;"><br /></span>
<b>Let's Run the S-E-M</b><br />
Lyrics by Tobi Ruwase<br />
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=Dg2HKMFsers">Let's Call the Whole Thing Off</a>," George & Ira Gershwin)<br />
<br />
We have finally, come to the end,<br />
Of Dr. Alan Reifman’s class,<br />
QM 1 to 4 have taken two years,<br />
From correlation to regression,<br />
<br />
Goodness knows, what the end will be,<br />
As we prep, for our exams,<br />
It’s time for us, to go down memory lane... (slight pause)<br />
Some things, that we’ve learnt:<br />
<br />
We need constructs and we need items,<br />
We need items, for each of our constructs,<br />
Constructs and items, items and constructs,<br />
Let’s run the SEM,<br />
<br />
Open the data, run your bivariates,<br />
Check for loadings, higher than the cut-off,<br />
Correlations! Loadings! Inform your decisions,<br />
Let’s run the SEM,<br />
<br />
But oh! If we run the SEM (slow),<br />
There may be a glitch,<br />
And oh! If we get a glitch,<br />
Then AMOS would not run,<br />
<br />
So, we’ve got correlations, we proceed to AMOS,<br />
Select the data file, from SPSS,<br />
Click on OK, now we’ve got our data,<br />
Now we run SEM,<br />
Oh! Let’s run the SEM,<br />
<br />
We call it, the BAM TOOL!!!<br />
In the AMOS toolbar,<br />
It draws your constructs, and then your items,<br />
Constructs and items, items and constructs,<br />
Let’s run the SEM,<br />
<br />
Using your cursor, for two types of arrows,<br />
Uni-directed or two-headed arrows,<br />
Construct to items, Oh, structural paths,<br />
Let’s run the SEM,<br />
<br />
But oh! If we run the SEM,<br />
There may be a glitch,<br />
And oh! If we get a glitch,<br />
Then AMOS would not run,<br />
<br />
So label your constructs, don’t forget items,<br />
Time to run the AMOS, don’t forget properties,<br />
Means and intercepts, for the missing data,<br />
Now we run SEM,<br />
Oh! Lets’ run the SEM,<br />
<br />
We need construct and we need items,<br />
We need items, for each of our constructs,<br />
Constructs and items, items and constructs,<br />
Let’s run the SEM,<br />
<br />
Open the data, run your bivariates,<br />
Check for loadings, higher than the cut-off,<br />
Correlations! Loadings! Inform your decisions,<br />
Let’s run the SEM,<br />
<br />
But oh! If we run the SEM,<br />
There may be a glitch,<br />
And oh! If we get a glitch,<br />
Then AMOS would not run,<br />
<br />
So, we’ve got correlations, we proceed to AMOS,<br />
Click on OK, now we’ve got our data,<br />
Now we run SEM,<br />
Oh! Let’s run the SEM,<br />
Let’s run the SEMMMMMMMMMMMM........<br />
<span style="background-color: white; color: #333333; font-family: "georgia" , serif; font-size: 13px;"><br /></span>
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgSAXYQr9LJe8x8kkxcYKoEQxOq9MJGfGdvmKpFeZIMMA4lCtxAXHJNeBNsEDessPILGo32gpvXHGCk6ZdOKdgszsLNF_g4-Abi1jBCUXb3jM75jSOYUVPJGvy3jZwPXlnXoP-YcAsQJ5U/s1600/click+hold+and+drag.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="337" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgSAXYQr9LJe8x8kkxcYKoEQxOq9MJGfGdvmKpFeZIMMA4lCtxAXHJNeBNsEDessPILGo32gpvXHGCk6ZdOKdgszsLNF_g4-Abi1jBCUXb3jM75jSOYUVPJGvy3jZwPXlnXoP-YcAsQJ5U/s1600/click+hold+and+drag.jpg" width="400" /></a></div>
<br />
<br />
The AMOS Structural Equation Modeling program has a lot of graphical features, which the beginning SEM student must adjust to. Let's do an <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">earlier song</a> ("Once You Work in AMOS") on the topic before our new one.<br />
<b><br /></b>
<b>Click, Hold, and Drag</b><br />
Lyrics by Alan Reifman, inspired by Tobi Ruwase<br />
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=AryHJdadezM&list=RDAryHJdadezM">Jump, Jive, and Wail</a>," Louis Prima, popularized in recent decades by the Brian Setzer Orchestra)<br />
<br />
<i><a href="https://www.youtube.com/watch?v=vXiAvFtnfoM">Video of this song being performed.</a></i><br />
<br />
Tobi, Tobi, drew a big model, on her pad,<br />
Tobi, Tobi, drew a big model, on her pad,<br />
When you learn AMOS,<br />
You gotta make the paths, zig and zag,<br />
<br />
Oh, you gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag away...<br />
<br />
(Saxophone solo)<br />
<br />
All these shapes, with a label, she's got to tag,<br />
All these shapes, with a label, she's got to tag,<br />
With the variable names,<br />
From SPSS, in the bag,<br />
<br />
Oh, you gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag away...<br />
<br />
(Guitar solo)<br />
<br />
A model is a model, and an AMOS error, is a nag,<br />
A model is a model, and an AMOS error, is a nag,<br />
You gotta draw things right,<br />
So other statisticians, will not rag,<br />
<br />
Let's make sure, her drawing work, doesn't lag,<br />
Let's make sure, her drawing work, doesn't lag,<br />
So that she can get,<br />
Her model to run, without a snag,<br />
<br />
Oh, you gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag away...<br />
<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag away...<br />
<br />
Oh, you gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag,<br />
You gotta click and hold, then you drag away...<br />
<br />
You gotta click and hold...<br />
You gotta click and hold...<br />
You gotta click and hold...<br />
<br />
(Guitar flourish)<br />
<br />
<b><br /></b>
<b>Common Model Mistakes</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=AmIlUKo4dQc">My Favorite Mistake</a>," Crow/Trott)<br />
<br />
<i>Performance videos of this song from SEM The Musical <a href="https://www.youtube.com/watch?v=AB01Q7Y0g6U">9</a> and <a href="https://www.youtube.com/watch?v=4xQyS8U1UWc">10</a>.</i><br />
<br />
Omitting, residual bubbles,<br />
Will surely, get you in trouble,<br />
When AMOS gives your model, a run,<br />
<br />
Deleting a, fixed-one loading,<br />
Should trigger a sense, of foreboding,<br />
The error messages, are no fun,<br />
<br />
You should know, as you go,<br />
When you're, just beginning,<br />
These are some of, the more subtle errors,<br />
<br />
You should know, as you go,<br />
These are, common mistakes,<br />
<br />
(Instrumental)<br />
<br />
Misnaming, your indicators,<br />
Will bring an, emotional nadir,<br />
You'll have to find out, just where you failed,<br />
<br />
Grouping scales, with low correlation,<br />
Your constructs, will bring frustration,<br />
Check Pearson <i>r</i>'s, and then you'll sail,<br />
<br />
You should know, as you go,<br />
When you're, just beginning,<br />
These are some of, the more subtle errors,<br />
<br />
You should know, as you go,<br />
These are common mistakes,<br />
These are common mistakes,<br />
<br />
(Bridge)<br />
<br />
Well, SEM is, quite technical,<br />
Little things, will send a ripple,<br />
If you get, an error message,<br />
Look at the, above suggestions,<br />
They should help you, find your way,<br />
<br />
(Instrumental)<br />
<br />
Keep in mind, you will find,<br />
These aren't, the only ones,<br />
That you'll encounter,<br />
Other things, can go wrong,<br />
<br />
Keep your concentration, high,<br />
These mistakes, can make you cry,<br />
<br />
These are common mistakes,<br />
These are common mistakes,<br />
These are common mistakes...<br />
<br />
<b>Fit It</b><br />
Lyrics by Brandon Logan<br />
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=QDWlNv5XFFM">Whip It</a>," G. Casale/M. Mothersbaugh for Devo)<br />
<br />
<i><a href="https://www.youtube.com/watch?v=7fwjpRhAZ7E">Performance video.</a></i><br />
<br />
Check that fit,<br />
Really question it,<br />
Pick out a stat,<br />
Take a look at that,<br />
<br />
When a matrix, comes along,<br />
You must fit it,<br />
To prove that, the model’s strong,<br />
You must fit it,<br />
When something’s going wrong,<br />
You must fit it,<br />
<br />
Now fit it,<br />
N-F-I,<br />
Get it high,<br />
Com-par...<br />
...i-tive or,<br />
Absolute,<br />
Try to increase it,<br />
The C-F-I,<br />
Go fit it,<br />
Fit it good,<br />
<br />
Minimum is not achieved,<br />
You won’t fit it,<br />
Constraints to be released,<br />
So you can fit it,<br />
This must be policed,<br />
For you can fit it,<br />
<br />
I say fit it,<br />
Fit it good,<br />
I say fit it,<br />
Fit it good,<br />
<br />
(Interlude)<br />
<br />
Check that fit,<br />
Really question it,<br />
Pick out a stat,<br />
Take a look at that,<br />
<br />
When a matrix comes along,<br />
You must fit it,<br />
To prove that, the model’s strong,<br />
You must fit it,<br />
When something’s going wrong,<br />
You must fit it,<br />
<br />
Now fit it,<br />
N-F-I,<br />
Get it high,<br />
Com-par...<br />
...i-tive or,<br />
Absolute,<br />
Try to increase it,<br />
The C-F-I...<br />
<br />
Now fit it,<br />
N-F-I,<br />
Get it high,<br />
Com-par...<br />
...i-tive or,<br />
Absolute,<br />
Try to increase it,<br />
The C-F-I<br />
Go fit it,<br />
Oh, fit it good!<br />
<br />
<br />
<b>We'll now perform some "classics" and, finally, our traditional closing number: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">Parsi-Mony</a></b><br />
<br />
<br />alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-55409385373601810872014-04-09T18:18:00.000-07:002018-05-02T20:29:33.565-07:00SEM The Musical 8<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiC0MocENo2LwpmEVu5JkkmBL_CADV6C82wu1Se2k7y9Qp5AiRyr8YcEBeHJkoPMgg4V4K2No3WkYJFprnKMgTPudpUnczoBTnKiaSAGZGfocpHkdCR-nSdEFOgcjxpRgyJoduP_texEok/s1600/SEM+Musical+8.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em; text-align: center;"><img border="0" height="280" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiC0MocENo2LwpmEVu5JkkmBL_CADV6C82wu1Se2k7y9Qp5AiRyr8YcEBeHJkoPMgg4V4K2No3WkYJFprnKMgTPudpUnczoBTnKiaSAGZGfocpHkdCR-nSdEFOgcjxpRgyJoduP_texEok/s1600/SEM+Musical+8.jpg" width="320" /></a></div>
<br />
<b><span style="color: red;">UPDATE:</span></b> Our eighth annual SEM The Musical was held on April 29, 2014. We had three <span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;">new songs this year, which are shown below. We also performed some songs</span><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;"> from previous SEM Musicals (links: </span><a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.333333015441895px; text-decoration: none;">1</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;">, </span><a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.333333015441895px; text-decoration: none;">2</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;">, </span><a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.333333015441895px; text-decoration: none;">3</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;">, </span><a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.333333015441895px; text-decoration: none;">4</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;">, </span><a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.333333015441895px; text-decoration: none;">5</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;">, </span><a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.333333015441895px; text-decoration: none;">6</a><span style="background-color: white; font-family: "georgia" , serif; font-size: 13.333333015441895px;">, </span><a href="http://reifman-sem.blogspot.com/2013/04/sem-musical-7-april-30.html" style="background-color: white; font-family: Georgia, serif; font-size: 13.333333015441895px; text-decoration: none;">7</a>).<br />
<br />
<i>Ivette Noriega, sporting her homemade <a href="http://www.thedailybeast.com/articles/2013/05/21/daft-punk-goes-back-to-the-future-with-random-access-memories.html">Daft Punk helmet</a>, and Dr. Reifman perform "Saturated Your Model." You may click on the photo to enlarge it. </i><br />
<div class="separator" style="clear: both; text-align: center;">
<br /></div>
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhggdhULOse7PSi97cM7W6Ikaji0h6Wc_oZ9_aCFG7V0ip-VPVJcQNBbDah90KawhwMSFKAyi6goKjjhcxOEAkbgkTjU57rNNt5TvO16tfJcl1hBmooF4D4ryqJ5BsH1eM0XrP46eC4g5A/s1600/ivette.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhggdhULOse7PSi97cM7W6Ikaji0h6Wc_oZ9_aCFG7V0ip-VPVJcQNBbDah90KawhwMSFKAyi6goKjjhcxOEAkbgkTjU57rNNt5TvO16tfJcl1hBmooF4D4ryqJ5BsH1eM0XrP46eC4g5A/s1600/ivette.jpg" width="176" /></a></div>
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhO2xbHe7kWSPDKrnrBg7o0nUsCbtAl0QdwPJwZbrLXwKn668nq6hxPckUz9fPWF4mmkr_2iNbGumnAzpVIPUZubDEhyphenhyphen2mW9gPEQ_mKBUjXIqNCwr7Y0T3N4lIeT9f0IpSE7iKLrhoUkeo/s1600/daft+punk+--+saturated+your+model.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="225" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhO2xbHe7kWSPDKrnrBg7o0nUsCbtAl0QdwPJwZbrLXwKn668nq6hxPckUz9fPWF4mmkr_2iNbGumnAzpVIPUZubDEhyphenhyphen2mW9gPEQ_mKBUjXIqNCwr7Y0T3N4lIeT9f0IpSE7iKLrhoUkeo/s1600/daft+punk+--+saturated+your+model.jpg" width="400" /></a></div>
<b><br /></b>
<b>Saturated Your Model </b>(<a href="https://www.tandfonline.com/doi/full/10.3402/ejpt.v2i0.5652">example</a>)<br />
Lyrics by Ivette Noriega and Alan Reifman<br />
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=5NV6Rdv1a3I">Get Lucky</a>,” Bangalter/de Homem-Christo/Williams/Rodgers) <br />
<br />
<i>Performance videos of this song from SEM The Musical <a href="https://www.youtube.com/watch?v=tiO9WQbTpnk">9</a> and <a href="https://www.youtube.com/watch?v=2ZaPJcxoxN0">10</a>.</i><br />
<br />
In the world, of SEM graphs,<br />
All the paths, have beginnings,<br />
It keeps, statisticians spinning (uh-huh),<br />
AMOS will be helping,<br />
<br />
(Look)<br />
<br />
You've, gone too far,<br />
You’ve linked all, paths there are,<br />
None you’ve left out,<br />
<a href="http://theartofmodelling.wordpress.com/2012/01/12/deriving-models-that-are-simple-but-not-too-simple/">Einstein’s quote</a>, did you flout?<br />
<br />
Fit indices are at 1,<br />
Degrees of freedom are none,<br />
You’ve got to know, what you’ve done,<br />
You’ve saturated your model!<br />
<br />
Fit indices are at 1,<br />
Degrees of freedom are none,<br />
You’ve got to know, what you’ve done,<br />
You’ve saturated your model!<br />
<br />
You’ve saturated your model,<br />
You’ve saturated your model,<br />
You’ve saturated your model,<br />
You’ve saturated your model,<br />
<br />
(Instrumental)<br />
<br />
S-E-M, has no limits,<br />
Your theory, is depicted,<br />
What is it, you’re testing?<br />
Parsimony says, leave out paths (uh-huh),<br />
<br />
You've, gone too far,<br />
You’ve linked all, paths there are,<br />
None you’ve left out,<br />
Einstein’s quote, did you flout?<br />
<br />
Fit indices are at 1,<br />
Degrees of freedom are none,<br />
You’ve got to know, what you’ve done,<br />
You’ve saturated your model!<br />
<br />
Fit indices are at 1,<br />
Degrees of freedom are none,<br />
You’ve got to know what you’ve done,<br />
You’ve saturated your model!<br />
<br />
You’ve saturated your model,<br />
You’ve saturated your model,<br />
You’ve saturated your model,<br />
You’ve saturated your model,<br />
<br />
<span style="color: red;">(Voice-synthesizer in background, shown in red)</span><br />
<span style="color: red;"><br /></span>
<span style="color: red;">Fit indices are at 1, </span><br />
<span style="color: red;">Fit indices are at 1, </span><br />
<span style="color: red;">Fit indices are at 1, </span><br />
<span style="color: red;">Fit indices are at 1 </span><br />
<span style="color: red;"><br /></span>
<span style="color: red;">Fit indices are at 1 (all of them), </span><br />
<span style="color: red;">Fit indices are at 1 (it's hard to interpret), </span><br />
<span style="color: red;">Fit indices are at 1, </span><br />
<span style="color: red;">Fit indices are at 1,</span><br />
<span style="color: red;"><br /></span>
<span style="color: red;">You’ve saturated your model! </span><br />
<span style="color: red;">You’ve saturated your model! </span><br />
<span style="color: red;">You’ve saturated your model! </span><br />
<span style="color: red;">You’ve saturated your model!</span><br />
<span style="color: red;"><br /></span>
<span style="color: red;">You’ve saturated your model! </span><br />
<span style="color: red;">You’ve saturated your model! </span><br />
<span style="color: red;">You’ve saturated your model! </span><br />
<span style="color: red;">You’ve saturated your model!</span><br />
<br />
You've, gone too far,<br />
You’ve linked all, paths there are,<br />
None you’ve left out,<br />
Einstein’s quote, did you flout?<br />
<br />
Fit indices are at 1,<br />
Degrees of freedom are none,<br />
You’ve got to know, what you’ve done,<br />
You’ve saturated your model!<br />
<br />
Fit indices are at 1,<br />
Degrees of freedom are none,<br />
You’ve got to know, what you’ve done,<br />
You’ve saturated your model!<br />
<br />
You’ve saturated your model!<br />
You’ve saturated your model!<br />
You’ve saturated your model!<br />
You’ve saturated your model!<br />
<br />
You’ve saturated your model!<br />
You’ve saturated your model!<br />
You’ve saturated your model!<br />
You’ve saturated your model...<br />
<br />
<div style="text-align: center;">
***</div>
<br />
<i>The "Spice Guys," Nicholas (Hairy Spice) and Alan (Veggie Spice), perform "If You Wanna Join My Construct."</i><br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjXWmrkoDDvRKvXBOXGYisIH1_UQ6hEra4P0chWvPwmC16OqQ0OFibMMPNc2yGVUuK8ZT7_S4_ZC9d_vYPUlcwOGGg6083IDK5jU2FFXUznUkg6V4mPS5NIUmMGIn7XMDR5rbYhIcmIo5M/s1600/spice+guys+--+if+you+want+to+join.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="204" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjXWmrkoDDvRKvXBOXGYisIH1_UQ6hEra4P0chWvPwmC16OqQ0OFibMMPNc2yGVUuK8ZT7_S4_ZC9d_vYPUlcwOGGg6083IDK5jU2FFXUznUkg6V4mPS5NIUmMGIn7XMDR5rbYhIcmIo5M/s1600/spice+guys+--+if+you+want+to+join.jpg" width="320" /></a></div>
<br />
<br />
<b>If You Wanna Join My Construct (You've Got to Load with My Friends) </b><br />
Lyrics by Nicholas Johnston and Alan Reifman<br />
May be sung to the tune of “<a href="https://www.youtube.com/watch?v=gJLIiF15wjQ">Wannabe</a>” (Spice Girls/Rowe/Stannard)<br />
<br />
<i>Performance video of this song from SEM The Musical <a href="https://www.youtube.com/watch?v=536BCVqFLR8">9</a> and <a href="https://www.youtube.com/watch?v=lAtRdJkBy9E">10</a>.</i><br />
<br />
Yo, I’ll tell you what to draw, what you really need to draw,<br />
So, tell me what to draw, what I really need to draw,<br />
I’ll tell you what to draw, what you really need to draw,<br />
So, tell me what to draw, what I really need to draw,<br />
I need a circle, need a box, I need a circle, need a box...,<br />
Really, really, really, really, really, need a box,
<br />
<br />
If you like my construct, then load significantly,<br />
If you wanna join me, minimize residuality,<br />
Now don't go wasting, iterations,<br />
Get your<i> r</i>'s together, we could load just fine,<br />
<br />
I’ll tell you what to draw, what you really need to draw,<br />
So, tell me what to draw, what I really need to draw,<br />
I need a circle, need a box, I need a circle, need a box...,<br />
Really, really, really, really, really, need a box,<br />
<br />
If you want to join my construct, you gotta load with my friends,<br />
Sharing variation, on that constructs depend,<br />
If you want to join my construct, you have got to show,<br />
High <i>r</i>’s with the other, manifests, you know,<br />
<br />
What do you think about that, now that you know the deal?<br />
Say you fit my construct, is your manifest for real?<br />
Got a small residual, I'll give you a try, <br />
If the construct won't account for your variance, then I'll say goodbye,<br />
<br />
Yo, I’ll tell you what to draw, what you really need to draw,<br />
So, tell me what to draw, what I really need to draw,<br />
I need a circle, need a box, I need a circle, need a box...,<br />
Really, really, really, really, really, need a box,<br />
<br />
If you want to join my construct, you gotta load with my friends,<br />
Sharing variation, on that constructs depend,<br />
If you want to join my construct, you have got to show,<br />
High <i>r</i>’s with the other, manifests, you know,<br />
<br />
So here's a story, from <i>r</i> to <i>p</i>,<br />
You wanna get with me, you gotta load significantly,<br />
We got CFA tests in place, and coefficients to taste,<br />
You then see, on your screen, which V loads, on the C,<br />
All your V's, you can see, reflect variance, manifestly,<br />
And if you please, you'll see...<br />
<br />
Get your constructs drawn, and run your model now,<br />
Get your constructs drawn, and run your model now,<br />
<br />
If you want to join my construct, you gotta load with my friends,<br />
Sharing variation, on that constructs depend,<br />
If you want to join my construct, you have got to show,<br />
High <i>r</i>’s with the other, manifests, you know,<br />
<br />
If you want to join my construct...<br />
You gotta, you gotta, you gotta,
you gotta, you gotta, load, load, load, load....<br />
<br />
Get your constructs drawn and run your model now,<br />
Get your constructs drawn and run your model now (uh, uh, uh, uh...).<br />
Get your constructs drawn and run your model now,<br />
Get your constructs drawn zigazig-ah,<br />
<br />
If you want to join my construct...<br />
<br />
<br />
<b>Non-Exchangeable </b><br />
Lyrics by Alan Reifman<br />
May be sung to the tune of “<a href="https://www.youtube.com/watch?v=Aebjhd7srqk&spfreload=10">Unforgettable</a>” (Irving Gordon; popularized by Nat King Cole)
<br />
<br />
Non-exchangeable,<br />
Some dyads’ fate,<br />
They’re arrange-able,<br />
By role, or trait,<br />
<br />
Such as hetero, spouses or steadies,<br />
Teacher-student pairs, boss and employees,<br />
(Slow) These are studied,<br />
From a distinguishable, view,<br />
<br />
But, exchangeable,<br />
Are some, you see,<br />
It’s not absolute,<br />
Who’s A, and B,<br />
<br />
Friends, or twins, or old college roommates,<br />
Same-sex spouses, or pairs who go on dates,<br />
More complex stats,<br />
Will be needed, for you...<br />
<br />
[Interlude -- Instrumental and vocal improvisation]<br />
<br />
Non-exchangeable,<br />
Some dyads’ fate,<br />
They’re arrange-able,<br />
By role, or trait,<br />
<br />
Other pairs are, interchangeable,<br />
Their data are, re-arrangeable,<br />
So their, APIM models,<br />
Are harder, to do...
<br />
<br />
<i><span style="color: red;">Thanks to Satabdi and Rebecca for the photos!</span></i>alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-91260051197706362992013-04-06T12:34:00.000-07:002018-11-24T10:15:10.226-08:00SEM The Musical 7<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiBV-Mo91czOxpnZTbACV-_AnE8LhDLHe4Ypp6gNU5tXvQ9_3s4xNEqzm0vkytJYmquXxNBVJ9r0ZMeKfcEqetG61neH7z6tTrnHbufAm9cGSvhug6_c2p4fMXNki3QD7PuYzBTddo2__o/s1600/sem+musical+7.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="400" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiBV-Mo91czOxpnZTbACV-_AnE8LhDLHe4Ypp6gNU5tXvQ9_3s4xNEqzm0vkytJYmquXxNBVJ9r0ZMeKfcEqetG61neH7z6tTrnHbufAm9cGSvhug6_c2p4fMXNki3QD7PuYzBTddo2__o/s400/sem+musical+7.jpg" width="245" /></a></div>
<br />
SEM The Musical 7 is now complete. New songs for this year are listed below, along with photos from some of the performances. Thanks to Hannah Korkow, Andrea Parker, Nancy Trevino Schafer, and Paulina Velez for the pictures. Links to the songs from our previous musicals are as follows: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a>, <a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html">5</a>, <a href="http://reifman-sem.blogspot.com/2012/04/new-songs-to-come-for-sem-musical-6.html">6</a>.<br />
<br />
<b>The Road to S-E-M</b><br />
Lyrics by Alan Reifman<br />
May be sung to the tune of “<a href="http://www.youtube.com/watch?v=RWgxgfMWm6A%20Factor%20analysis,%20and%20how%20you%20correlate,">Shambala</a>” (Daniel Moore, popularized by Three Dog Night)<br />
<br />
Factor analysis, and how to correlate,<br />
On the road to S-E-M,<br />
Need to know regression, and draw paths so straight,<br />
On the road to S-E-M,<br />
<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
Run your S-E-M,<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
Run your S-E-M,<br />
<br />
You’ll draw little boxes, and these larger hoops,<br />
On the road to S-E-M,<br />
You’ll run panel models, and multiple groups,<br />
On the road to S-E-M,<br />
<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
Run your S-E-M,<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
Run your S-E-M,<br />
<br />
What... is your NFI,<br />
Once you’ve run, your S-E-M?<br />
What... is your TLI,<br />
Once you’ve run, your S-E-M?<br />
<br />
(Brief guitar solo)<br />
<br />
The measurement model, that’s a CFA,<br />
On the road to S-E-M,<br />
There’s the structural part, paths that flow one way,<br />
On the road to S-E-M,<br />
<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
Run your S-E-M,<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
Run your S-E-M,<br />
<br />
What... is, Hoelter’s CN,<br />
Once you’ve run, your S-E-M?<br />
What... is your CFI,<br />
Once you’ve run, your S-E-M?<br />
<br />
And, chi-square to df,<br />
Once you’ve run, your S-E-M?<br />
And, RM-SEA,<br />
Once you’ve run, your S-E-M?<br />
<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
Run your S-E-M,<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
On the road to S-E-M,<br />
<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
On... the... road...<br />
Ooh-ooh-hoo, ooh-ooh-ooh, yeah...<br />
On the road to S-E-M...<br />
<br />
(More guitar)<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj2eO0yqD2MGnmAL1Nw-Ng1h23ZLIpHzTjEOrnTNdv8lo-rK1s0qyYXec_iS-P0Q_xvfHVaeILgxUHATDlirBlFUaT0c6lIZvSlptH47ne5E-kVfsGYb0OPoMd7XQzrjoWfyJWkuq_oWa0/s1600/air+guitar.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="400" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEj2eO0yqD2MGnmAL1Nw-Ng1h23ZLIpHzTjEOrnTNdv8lo-rK1s0qyYXec_iS-P0Q_xvfHVaeILgxUHATDlirBlFUaT0c6lIZvSlptH47ne5E-kVfsGYb0OPoMd7XQzrjoWfyJWkuq_oWa0/s400/air+guitar.jpg" width="361" /></a></div>
<br />
<div style="text-align: center;">
<i><span style="color: #38761d;"><b>His right hand a blur, Dr. Reifman shows off some of his air-guitar technique.</b></span></i></div>
<br />
<b>---</b><br />
<b><br /></b>
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgW_LU2bzmFtszHkD8Iw2fQG5zhUWiB1ZT8NJTZqmSiAJXvNHQl0meFsy9HPxwNklyJ956CMsdRP14k-hPOe0-aCogrxLDPBohZ8llomzL38CclpkA9xVZ7Pp6DRH8JrTsD2kBt6rJfBxw/s1600/SEM+nestedness.jpg" imageanchor="1" style="clear: left; cssfloat: left; float: left; margin-bottom: 1em; margin-right: 1em;"><img border="0" height="141px" i8="true" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgW_LU2bzmFtszHkD8Iw2fQG5zhUWiB1ZT8NJTZqmSiAJXvNHQl0meFsy9HPxwNklyJ956CMsdRP14k-hPOe0-aCogrxLDPBohZ8llomzL38CclpkA9xVZ7Pp6DRH8JrTsD2kBt6rJfBxw/s200/SEM+nestedness.jpg" width="200px" /></a></div>
<br />
<b><br /></b>
<b><br /></b>
<b><br /></b>
<b><br /></b>
<b><br /></b><br />
<b><br /></b>
<b><br /></b>
<b><br /></b>
<b><br /></b>
<b><br /></b><b>Why Don’t You Run a Set of Nested Models?</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of [I Don’t Know Why You Don’t Take Me] <a href="https://www.youtube.com/watch?v=WjsEkZf9b7E">Downtown</a>; Laird/McAnally/Hemby; popularized by Lady Antebellum)<br />
<br />
Well, I was trying to get a model going,<br />
Staring at AMOS, bright on my screen,<br />
I tried to think, but not really knowing,<br />
Had a few pathways,<br />
But really now, what does it mean?<br />
<br />
Knew a scale, which could be a mediator,<br />
Could draw paths, right through X-Y-Z,<br />
Inside my mind, I became a debater,<br />
Should I model cause-to-effect, directly?<br />
But then... a low voice, said to me:<br />
<br />
“Why don’t you run, a set of nested models?<br />
Why don’t you run, a test, to check and see,<br />
If the new paths, lower the chi-square?<br />
I mean, significant-ly?”<br />
<br />
“It is a very simple, calculation,<br />
Subtract the smaller, from the larger chi-square,<br />
Then you compare results, to a table,<br />
And that’s how you, test nested models, if you dare,<br />
Oh-oh-oh, if you dare...”<br />
<br />
I want to keep my, number of paths low,<br />
To follow notions, of parsimony,<br />
Einstein said to keep, scientific theories,<br />
As simple as they, can possibly be,<br />
But no simpler!<br />
<br />
And here... comes that voice...<br />
<br />
“Why don’t you run, a set of nested models?<br />
Why don’t you run, a test, to check and see,<br />
If the new paths, lower the chi-square?<br />
I mean, significant-ly?”<br />
<br />
“It is a very simple, calculation,<br />
Subtract the smaller, from the larger chi-square,<br />
Then you compare results, to a table,<br />
And that’s how you, test nested models, if you dare...”<br />
<br />
(Guitar solos)<br />
<br />
“Why don’t you run, a set of nested models?<br />
Why don’t you run, a test, to check and see,<br />
If the new paths, lower the chi-square?<br />
I mean, significant-ly?”<br />
<br />
“It is a very simple, calculation,<br />
Subtract the smaller, from the larger chi-square,<br />
Then you compare results, to a table,<br />
And that’s how you, test nested models, if you dare,<br />
Oh-oh-oh, if you dare...”<br />
<br />
“Yeah, why don’t you run a set of nested models?”<br />
<br />
OK, I’ll run a set of nested models...<br />
(Music fades out)<br />
<br />
Now I get it...<br />
<br />
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgWvqMMXpd8rxTjReayb0Yeu4M8qjThbtp0fu_4eIm2Bd5xeV37z5V4fiK3vXqvslqY-gpz8zW__2Yu3az0a42XVq1A0JcQorovTOKEtx55oaBPEHq5jTyomuz8lzlG4_flXDveiKwMxIA/s1600/violeta+solo.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="245" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgWvqMMXpd8rxTjReayb0Yeu4M8qjThbtp0fu_4eIm2Bd5xeV37z5V4fiK3vXqvslqY-gpz8zW__2Yu3az0a42XVq1A0JcQorovTOKEtx55oaBPEHq5jTyomuz8lzlG4_flXDveiKwMxIA/s400/violeta+solo.jpg" width="400" /></a></div>
<br />
<div style="text-align: center;">
<b> <i><span style="color: #38761d;">Violeta Kadieva, the first student ever to write two songs for a single musical (see next two songs).</span></i></b></div>
<br />
<br />
<b>Trouble When I Ran You</b><br />
Lyrics by Violeta Kadieva<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=bEh2iECS0To">I Knew You Were Trouble</a>,” Swift/Martin/Shellback)<br />
<br />
I discovered how, to draw AMOS models,<br />
And I was having fun, getting it to run,<br />
It dawned on me, it dawned on me, it dawned on me…. (ee-ee-ee-ee-ee)<br />
<br />
I had to run a, new measurement model,<br />
I used AMOS, to fit hypothesized paths,<br />
And draw the diagram, draw the diagram, draw the diagrammmm (ee-ee-ee-ee-ee)<br />
<br />
And I raaannnn the modeelll, with all the vaaaariables,<br />
And I also included, the factor indicators there,<br />
<br />
And I knew you were trouble when I ran you,<br />
So shame on me that,<br />
I did not draw, another alternative model,<br />
To check if it, could be a better fit,<br />
<br />
And I knew you were trouble when I ran you,<br />
So shame on me that,<br />
I did not draw another, alternative model,<br />
Noooow I am so wondering about it, <br />
<br />
Runs? No! Trouble, trouble, trouble,<br />
Runs? No! Trouble, trouble, trouble,<br />
<br />
So what can I do? Let’s run another one,<br />
Explore another one, by adding other paths,<br />
And check if it's a better one, it's a better one, and improves the model (ee-ee-ee-ee-ee)<br />
<br />
I checked the chi squares, with and without new paths,<br />
So, the model changed degrees, the delta of the change is, <br />
We'll just have to see, we'll just have to see, we'll just have to see (ee-ee-ee-ee-ee),<br />
<br />
The chi squaaaare table, showwwwed signiiiificant improvement,<br />
And I realized having more paths, improves the model signiiificantly,<br />
<br />
I knew you were trouble when I ran you,<br />
So shame on me that,<br />
I did not draw another alternative model,<br />
To check if it could be a better fit,<br />
<br />
And I knew you were trouble when I ran you,<br />
So shame on me that,<br />
I did not draw any equivalent models,<br />
Now I am so wondering about it,<br />
<br />
Runs? No! Trouble, trouble, trouble,<br />
Runs? No! Trouble, trouble, trouble,<br />
<br />
So when the model, is improved significantly,<br />
By costing us only, a few degrees of freedom,<br />
And lowering the chi square significantly,<br />
We have to accept the alternative model as a better model to use,<br />
Yes exactlyyy…<br />
<br />
(Lengthy sound-effects riff)<br />
<br />
I knew you were trouble when I ran you,<br />
So shame on me that,<br />
I did not draw another alternative model,<br />
To check if it could be a better fit,<br />
<br />
I knew you were trouble when I ran you,<br />
So shame on me that,<br />
I did not draw any equivalent models either,<br />
Now I am so wondering about it,<br />
<br />
Runs? No! Trouble, trouble, trouble,<br />
Runs? No! Trouble, trouble, trouble,<br />
<br />
I knew you were trouble when I ran you,<br />
Trouble, trouble, trouble,<br />
<br />
I knew you were trouble when I ran you,<br />
Trouble, trouble, trouble...<br />
<br />
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEip3xW-ec7rFvXRcpRHwRdQi9IGhBmDNvS-QishR3NpdTMTNHSg59sBgqKpiJ-FJ1hjw3Au6ujRSL5_n3e8SDjdJDD0je3z2as4uB2ExPXQhemHa0btmCw9PPBOoPSKYHawPZf0bDdcg5s/s1600/violeta+&+esperanza.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="400" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEip3xW-ec7rFvXRcpRHwRdQi9IGhBmDNvS-QishR3NpdTMTNHSg59sBgqKpiJ-FJ1hjw3Au6ujRSL5_n3e8SDjdJDD0je3z2as4uB2ExPXQhemHa0btmCw9PPBOoPSKYHawPZf0bDdcg5s/s400/violeta+&+esperanza.jpg" width="337" /></a></div>
<br />
<div class="separator" style="clear: both; text-align: center;">
</div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<!--[if gte mso 9]><xml>
<w:LatentStyles DefLockedState="false" LatentStyleCount="156">
</w:LatentStyles>
</xml><![endif]--><!--[if gte mso 10]>
<style>
/* Style Definitions */
table.MsoNormalTable
{mso-style-name:"Table Normal";
mso-tstyle-rowband-size:0;
mso-tstyle-colband-size:0;
mso-style-noshow:yes;
mso-style-parent:"";
mso-padding-alt:0in 5.4pt 0in 5.4pt;
mso-para-margin:0in;
mso-para-margin-bottom:.0001pt;
mso-pagination:widow-orphan;
font-size:10.0pt;
font-family:"Times New Roman";
mso-fareast-font-family:"Times New Roman";
mso-ansi-language:#0400;
mso-fareast-language:#0400;
mso-bidi-language:#0400;}
</style>
<![endif]--><b>A Terrifying Model</b><br />
<div class="MsoNormal">
Lyrics by Violeta Kadieva (performance accompanied by <span class="nDesc">Esperanza Bregendahl, right</span>)</div>
<div class="MsoNormal">
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=tHpXxBDXsJw">Terrified</a>," DioGuardi/Reeves, popularized by Katharine
McPhee) </div>
<br />
You, AMOS stats,<br />
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Are the greatest... find,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
In the world, of software,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
You're <i>the</i> S-E-M package...
</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Did not make it, with my loadings, to the standard, </div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Of the point-4 magnitude...</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
I ran it again, with a new notion,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Each iteration, sends my heart, like a shooting star,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
I'm looking at, the weak connections,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
But I think that, I might not be too far... </div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And I-I-I-I-I... </div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And I-I-I-I-I... I'm terrified…</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
For the first time, and hopefully the last time,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
In S-E-M life (ummm-mmm)
</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
This, could be good,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
I'm ready to update, my diagram,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And nothing's worse<span style="mso-spacerun: yes;">,</span></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<span style="mso-spacerun: yes;">Than not seeking..., a novel path, </span></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And this could be, all that I need,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Maybe if I try... </div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
My revised model, is now in motion,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Each iteration, seems likes it getting near,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
I'm looking at, much stronger loadings,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Looking at the fit, I now cheer!
</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And I-I-I-I-I... ,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And I-I-I-I-I... I'm still terrified,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
For the first time and hopefully the last time</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
In S-E-M… life,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
I only, just looked, at the Root Mean Square*,</div>
And the Root Mean Square's, below oh-5,<br />
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
So don't you doubt, what I've been running,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
To point-9, the fit indices are close,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
As a modeler, I feel alive...</div>
<br />
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
I checked it again, looked at the fit measures,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
Each iteration, is giving a small chi-square, </div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And its ratio, to degrees of freedom's, not even 3,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<br /></div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And I-I-I-I-I... I did it,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
And I am not terrified anymore,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
For the first time and hopefully the last time,</div>
<div class="MsoNormal" style="line-height: normal; margin-bottom: .0001pt; margin-bottom: 0in;">
<span style="font-family: inherit; font-size: small;">In S-E-M… life…life…life,</span></div>
<span style="line-height: 115%;">S-E-M life...</span><br />
<br />
<span style="color: #6aa84f;"><span style="font-family: inherit;"><span style="font-size: 11pt; line-height: 115%;"><i>[*Root Mean Square Error of Approximation, more commonly known as the RMSEA.]</i></span></span></span><br />
<br />
<span style="color: #6aa84f;"><span style="font-family: inherit;"><span style="font-size: 11pt; line-height: 115%;"><i> </i></span></span></span>
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjd5A4x7WKiFCWE4KAkLSy81wSI9E620jSxnCqZVnndI9oWkR7ooVMkRGj0nR6OD0Sk-9Oa8BE96A2v2TNUe9zZz9VU85eqqXajT1Q9tXLLwnl2EoYrRVZcGBzjOnTKnBqvqtDK8ASebe8/s1600/kaitlin+%2526+lisa.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="355" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjd5A4x7WKiFCWE4KAkLSy81wSI9E620jSxnCqZVnndI9oWkR7ooVMkRGj0nR6OD0Sk-9Oa8BE96A2v2TNUe9zZz9VU85eqqXajT1Q9tXLLwnl2EoYrRVZcGBzjOnTKnBqvqtDK8ASebe8/s400/kaitlin+%2526+lisa.jpg" width="400" /></a></div>
<br />
<div style="text-align: center;">
<i><span style="color: #38761d;"><b>The next song, "Nice Nice Beta," was performed by Lisa Merchant (left, with brass knuckles spelling out "AMOS"), Kaitlin Leckie (with the Beta necklace), and I-Shan Yang (not pictured)</b></span></i></div>
<br />
<b>Nice Nice Beta </b><br />
Lyrics by Kaitlin Leckie,<br />
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=rog8ou-ZepE">Ice Ice Baby</a>," Vanilla Ice/Earthquake/M. Smooth; based on earlier song "Under Pressure," Queen [Deacon/May/Mercury/Taylor] and Bowie)<br />
<br />
Yo SEM
let’s run it<br />
<br />
(Hook)
Nice nice Beta<br />
Nice nice Beta<br />
All right stop.<br />
<br />
Correlate and regress them, SEM is back with a brand new edition, <br />
Something grabs a hold of me tightly, could be a residual influence slightly, <br />
Will it be significant? I don't know,<br />
Minimum achieved? Fo sho'!<br />
To the output, I must get a handle,
Maximum Likelihood estimate? Full.<br />
Husband and wife, data distinguished,<br />
Exchangeable? Plan is extinguished,<br />
Who checks Betas? Should be everybody,<br />
Anything less than .20, is a felony,<br />
Means and intercepts, with missing, estimate,<br />
Better hit the minimum, the model won’t wait,<br />
If there was an error, yo I’ll will solve it,<br />
Check out the model, while AMOS resolves it<br />
<br />
(Hook) Nice nice Beta, <br />
Nice nice Beta,<br />
.20 is a nice nice Beta, <br />
.20 is a nice nice Beta,<br />
<br />
Now that the arrows are jumping, the constructs kicked in,<br />
The indicators are pumping, stats to the point,<br />
To the point of no faking, cooking up models, like a pound of bacon,<br />
Burning them if you ain’t thorough and nimble,<br />
I go crazy when I see a Greek symbol,<br />
And a dataset with the groups all stacked,<br />
I’m on a roll and it’s time to see impact, <br />
AMOS version 21-point-0,<br />
With estimates on, so missings won’t blow,<br />
Means on standby, waiting just to say hi,<br />
Did you stop? No I just standardized,<br />
Kept on pursuing the solution, find it relates but no causal attribution,<br />
Effects immense, yo so I determined it’s Actor Partner Interdependence,<br />
Results so hot they’re creatin’ haters, <br />
Research hovers on single-level data,<br />
Jealous ‘cause my model’s lookin’ fine,<br />
TLI with nine-five; CFI with point-nine,<br />
How’s that for goodness of fit?<br />
The haters acting ill, because the model’s <span style="color: #6aa84f;">such a hit*</span>, <br />
<br />
Factor loadings, rang out like a bell,<br />
Check my indicators-all I see is swell,<br />
Following the constructs real fast,<br />
Judged by the indicators-shows that they’ll last,<br />
Factor to factor the model’s packed,<br />
I’m trying to determine if the model lacks,<br />
Effects on the actors and partners you see,<br />
Daring dyadic data analyses,<br />
If there was an error, yo I’ll solve it,<br />
Check out the model, while AMOS resolves it<br />
<br />
Nice Nice Beta, <br />
.20 is a nice nice Beta
(repeat)<br />
<br />
<i><span style="color: #6aa84f;">*[One could substitute "hot sh--," but we're a family-oriented group!]</span></i><br />
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
</div>
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhm0xzBRJzh2rD2xsUzPEGW9C294sqQ5XYf-blvrCLcS4PuxS6E58NNXQLGVlMNkw7lEfzP0iOaOqIgJZVfhWszb2sWZYehdHcYYu5v8ulyl9Z6TLeblyuQqF_oKYsdSzy9l_2194pJHOE/s1600/andrea+anuradha+paulina.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="338" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhm0xzBRJzh2rD2xsUzPEGW9C294sqQ5XYf-blvrCLcS4PuxS6E58NNXQLGVlMNkw7lEfzP0iOaOqIgJZVfhWszb2sWZYehdHcYYu5v8ulyl9Z6TLeblyuQqF_oKYsdSzy9l_2194pJHOE/s400/andrea+anuradha+paulina.jpg" width="400" /></a></div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<div class="separator" style="clear: both; text-align: center;">
</div>
<br />
<b>Fit & Fine </b><br />
Lyrics by Andrea Parker, Anuradha Sastry, and Paulina Velez<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=IsUsVbTj2AY">Suit & Tie</a>,” Timberlake/Mosley/Carter/Harmon/Fauntleroy/Stubbs/Wilson/Still; performed by Justin Timberlake, featuring Jay Z)<br />
<br />
Checking on the CFI, TLI, NFI,<br />
Checking on the CFI, TLI, NFI,<br />
Can I show you a few things?<br />
A few things, a few things, how the model fits,<br />
<br />
Checking on the CFI, TLI, NFI,<br />
Checking on the CFI, TLI, NFI,<br />
Let me show you a few things,
Let me show you a few things,<br />
Are you ready Reifman?<br />
<br />
[Verse 1]<br />
I can't wait, till I get to run my model in AMOS,<br />
Got a large data set, just like a census,<br />
I cleaned up the data and I just have to run them,<br />
Hope it fits fine, cause it's all mine,<br />
Hey baby, I have three latent constructs,<br />
If the loadings are above 0.4,<br />
We might learn something,<br />
Hoping that the minimum is achieved when we run it,<br />
Fits so fine, tonight,<br />
<br />
[Hook]<br />
And as long as I've got my CFI,<br />
I'mma move on to the NFI,<br />
And they are both above 0.9,<br />
This might be a good fit,<br />
TLI is also high,<br />
RMSEA as low as Kenny likes,<br />
Fit is tested in AMOS, tonight,<br />
<br />
Let me show you a few things,<br />
Let me show you a few things,<br />
Show you a few things about fit,<br />
While we’re running a model,<br />
This is a good fit,<br />
Show you a few things about fit,<br />
Hey,<br />
<br />
[Verse 2]<br />
Stop, let me get a good look at Hoelter’s,<br />
Ohhh so neat, now I know why Berndt likes it,<br />
Ohhh chi-square is big and might make it rubbish,<br />
But that's alright, cause the rest are fine,<br />
Ohhh go on and celebrate with a party,<br />
I guess friends are mad, cause they wish they had it,<br />
Uuuu my model, the fittest, yeah you're a classic,<br />
And you're all mine tonight,<br />
<br />
[Hook]<br />
And as long as I've got my CFI,<br />
I'mma move on to the NFI,<br />
And they are both above 0.9,<br />
This might be a good fit,<br />
TLI is also high,<br />
RMSEA as low as Kenny likes,<br />
Fit is tested in AMOS tonight,<br />
Let me show you a few things,<br />
Let me show you a few things,<br />
Show you a few things about fit,<br />
While we’re running a model,<br />
This is a good fit,<br />
Show you a few things about fit...
alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-35436536838964673992012-09-30T18:34:00.000-07:002013-01-16T10:40:07.251-08:00Writing Up SEM FindingsThe SEMNET e-mail discussion listserv had some useful references recently on how to write up SEM findings (thanks to Debbi Bandalos, Mbaye Fall Diallo, Jonathon Little, Amin Mousavi, Christian M. Ringle):<br />
<br />
Bandalos, D.L., & Finney, S.J. (2010). Factor analysis: Exploratory and confirmatory. In G.R. Hancock & R.O. Mueller (Eds.), <i>The reviewer’s guide to quantitative methods in the social sciences</i> (pp. 125-155). New York: Routledge. <br />
<br />
Boomsma, A. (2000). Reporting analyses of covariance structures (Teacher's Corner). <i>Structural Equation Modeling, 7</i>, 461-83.<br />
<br />
Boomsma, A., Hoyle, R. H., & Panter, A. T. (2012). The structural equation modeling research report. In R. H. Hoyle (Ed.), <i>Handbook of structural equation modeling</i> (pp. 341-358). New York: Guilford Press.<br />
<br />
Gefen, D., Rigdon, E. E., & Straub, D.W. (2011). Editor's comment: An update and extension to SEM guidelines for administrative and social science research. <i>MIS Quarterly, 35</i>, iii-xiv.<br />
<br />
Hoyle, R. H., & Panter, A. T. (1995). Writing about structural equation models. In R. H. Hoyle (Ed.), <i>Structural equation modeling: Concepts, issues and applications</i> (pp. 158-176). Thousand Oaks, CA: Sage.<br />
<br />
Jackson, D. L., Gillaspy J. A., & Purc-Stephenson R. (2009). Reporting practices in confirmatory factor analysis: An overview and some recommendations. <i>Psychological Methods, 14</i>, 6–23.<br />
<br />
<div class="body-paragraph">
Mueller,
R. O., &
Hancock,
G. R. (
2010).
Structural equation modeling. In
G. R.Hancock &
R. O.Mueller (
Eds.),
<i>The reviewer's guide to quantitative methods in the social sciences</i> (pp.
371–
383).
New York:
Routledge.
</div>
<br />
Raykov, T., Tomer, A., & Nesselroade, J. R. (1991). Reporting structural equation modeling results in <i>Psychology and Aging</i>: Some proposed guidelines. <i>Psychology and Aging, 6</i>, 499-503. <br />
<br />
Schreiber, J. B., Nora, A., Stage, F. K., Barlow, E. A., & King, J. (2006). Reporting structural equation modeling and confirmatory factor analysis results: A review. <i>Journal of Educational Research, 99</i>, 323-337.alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-45514060705388130132012-04-18T19:43:00.002-07:002015-04-23T19:01:52.984-07:00SEM The Musical 6<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgb-xpO8EWU3bEkFhGb66mdTTXBCutBy1klLDA0XunD6QbGUHqj8ljCNJFWhzHUUOOgboQvKJBa1DXQCbEyqUWTSy5eZO5jt3fUea-n_T-vOiawHHjwIZCKQz-em7O3dDOKcin5GichQtI/s1600/sem+musical+6.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgb-xpO8EWU3bEkFhGb66mdTTXBCutBy1klLDA0XunD6QbGUHqj8ljCNJFWhzHUUOOgboQvKJBa1DXQCbEyqUWTSy5eZO5jt3fUea-n_T-vOiawHHjwIZCKQz-em7O3dDOKcin5GichQtI/s400/sem+musical+6.jpg" height="377" width="400" /></a></div>
<br />
<b style="color: #cc0000;">UPDATE May 10.</b><i> SEM The Musical 6</i> is now on the books. Shown below are the new songs written this year, along with some photos of the performances (thanks to Chris Bedard and Xiaohui Tang for the pictures).<br />
<br />
We had a surprise "visit" via video feed from three friends of the musical who moved last August from Texas Tech to Virginia Tech: Professor Anisa Zvonkovic (below, center), now department chair at VT; Kyung-Hee Lee (left), who finished her Ph.D. at TTU and is now a post-doc at VT; and Andrea "Hermione" Swenson (right), who finished her Master's at TTU and is studying for her Ph.D. at VT. The song they wrote is shown further down the page.<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhpxXyWqVuw9Ojzg0bvLHD2w7C5-lqRt8-_DA4wkZ20xMUoV6MITfR5rxfR6082CuH2gey92hlvWDLq80NQufBZEcD4WhHmM7jhMLJeGC2qNO1F530CuSW8-T3LLAJaObSopV9H1X-doag/s1600/vt3.JPG" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhpxXyWqVuw9Ojzg0bvLHD2w7C5-lqRt8-_DA4wkZ20xMUoV6MITfR5rxfR6082CuH2gey92hlvWDLq80NQufBZEcD4WhHmM7jhMLJeGC2qNO1F530CuSW8-T3LLAJaObSopV9H1X-doag/s400/vt3.JPG" height="298" width="400" /></a></div>
<br />
We also sang several songs from the previous five musicals (links: <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a>, <a href="http://reifman-sem.blogspot.com/2011/04/heres-announcement-for-this-years.html">5</a>). For the first time ever, we ran a poll (see right-hand column) to help determine which songs from previous years to sing. The song "Parsimony," from <i>SEM The Musical 1</i>, is always what we close the show with, so it was not included in the poll. That's enough background. Here are this year's new songs...<br />
<br />
<div class="MsoNormal">
<b style="mso-bidi-font-weight: normal;">Let’s Sing About It</b></div>
<div class="MsoNormal">
Lyrics by Alan Reifman</div>
<div class="MsoNormal">
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=Nd4E9ePUrok">Tell Her About It</a>,” Billy Joel)</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Listen all you, SEM
users,</div>
<div class="MsoNormal">
It’s our yearly, singing day,</div>
<div class="MsoNormal">
To review the things we’ve learned, this semester,</div>
<div class="MsoNormal">
In a most unusual, way,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
From your model fit, </div>
<div class="MsoNormal">
To max-i-mum likelihood,</div>
<div class="MsoNormal">
You’ve learned practices that, </div>
<div class="MsoNormal">
Expert modelers should,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Listen all you, SEM
users,</div>
<div class="MsoNormal">
You’ve really come, a long way,</div>
<div class="MsoNormal">
From regression, correlation, and pathways,</div>
<div class="MsoNormal">
Through factors and CFA,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
From constraining paths,</div>
<div class="MsoNormal">
And checking, delta chi-square,</div>
<div class="MsoNormal">
To run multi-groups and,</div>
<div class="MsoNormal">
Seeing what paths, they share,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Let’s sing about it,</div>
<div class="MsoNormal">
Cover, all the things, we do,</div>
<div class="MsoNormal">
So that, all the concepts,</div>
<div class="MsoNormal">
Now seem old, although they’re new,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Let’s sing about it,</div>
<div class="MsoNormal">
From the blueprint, to the fit,</div>
<div class="MsoNormal">
How to judge, a model,</div>
<div class="MsoNormal">
Knowing how, to interpret,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
(brief interlude)</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Listen all you, SEM
users,</div>
<div class="MsoNormal">
You can run a model, now,</div>
<div class="MsoNormal">
You can draw the boxes, circles, and arrows,</div>
<div class="MsoNormal">
Those are things, that you know how,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
But to understand,</div>
<div class="MsoNormal">
In depth, is what we seek,</div>
<div class="MsoNormal">
To make each, of you,</div>
<div class="MsoNormal">
A real, SEM
geek,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Let’s sing about it,</div>
<div class="MsoNormal">
Cover, all the things, we do,</div>
<div class="MsoNormal">
So that, all the concepts,</div>
<div class="MsoNormal">
Now seem old, although they’re new,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Let’s sing about it,</div>
<div class="MsoNormal">
From the blueprint, to the fit,</div>
<div class="MsoNormal">
How to judge, a model,</div>
<div class="MsoNormal">
Knowing how, to interpret,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
So, now and then, you get an error,</div>
<div class="MsoNormal">
Cause you’ve drawn, something in AMOS, that is wrong,</div>
<div class="MsoNormal">
Make sure to use, the correct tools,</div>
<div class="MsoNormal">
So figuring out the right way, won’t take long,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Listen all you, SEM
users,</div>
<div class="MsoNormal">
You’ve, really, come, a long way,</div>
<div class="MsoNormal">
From regression, correlation, and pathways,</div>
<div class="MsoNormal">
Through factors and CFA,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
From constraining paths,</div>
<div class="MsoNormal">
And checking, delta chi-square,</div>
<div class="MsoNormal">
To run multi-groups and,</div>
<div class="MsoNormal">
Seeing what paths, they share,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Let’s sing about it,</div>
<div class="MsoNormal">
Cover, all the things, we do,</div>
<div class="MsoNormal">
So that, all the concepts,</div>
<div class="MsoNormal">
Now seem old, although they’re new,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
Let’s sing about it,</div>
<div class="MsoNormal">
From the blueprint, to the fit,</div>
<div class="MsoNormal">
How to judge, a model,</div>
<div class="MsoNormal">
Knowing how, to interpret,</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
<span class="lineline-s">Let’s sing about it!</span></div>
<div class="MsoNormal">
<span class="lineline-s">Sing about what, you’ve learned here,</span></div>
<div class="MsoNormal">
<span class="lineline-s">We’ve got to, all sing about it, </span></div>
<div class="MsoNormal">
<span class="lineline-s">How degrees-of-freedom work,</span></div>
<div class="MsoNormal">
<span class="lineline-s">We’ve got to, all sing about it, </span></div>
<div class="MsoNormal">
<span class="lineline-s">Finding out the, errors that lurk, </span></div>
<div class="MsoNormal">
<span class="lineline-s">We’ve got to, all sing about it, </span></div>
<div class="MsoNormal">
<span class="lineline-s">We’ve all got, models to run </span></div>
<div class="MsoNormal">
<span class="lineline-s">We’ve got to, all sing about it </span></div>
<div class="MsoNormal">
<span class="lineline-s">You know, we want to have fun,</span></div>
<div class="MsoNormal">
<span class="lineline-s">We got to, all sing</span><span class="lineline-shover"> about it… </span><br />
<br />
<div class="x_MsoNoSpacing">
<b>You Have Not Shown It’s Causal</b></div>
<div class="x_MsoNoSpacing">
Lyrics (and performance) by Devin DuPree</div>
<div class="x_MsoNoSpacing">
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=1SCOimBo5tg">When You Say Nothing at All</a>,” Overstreet/Schlitz, popularized by Alison Krauss)<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjxQzQ-ZivpgY6aDKKw9qZc-TT32oY9mw_uWIh13iZPMxESyyCqJYX7FCBnhy0dDHyLrSU5V0cZnhZyac6aHNTuSNMRbzSeApZYV6uoAOxGQta-s2OhF_TwCgyfMFgMvmg8X-5X593CgpE/s1600/dd.JPG" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjxQzQ-ZivpgY6aDKKw9qZc-TT32oY9mw_uWIh13iZPMxESyyCqJYX7FCBnhy0dDHyLrSU5V0cZnhZyac6aHNTuSNMRbzSeApZYV6uoAOxGQta-s2OhF_TwCgyfMFgMvmg8X-5X593CgpE/s400/dd.JPG" height="400" width="293" /></a></div>
</div>
<div class="x_MsoNoSpacing">
<br />
Sometimes you may want, to show how, two constructs relate,</div>
<div class="x_MsoNoSpacing">
And you may want to show, that B’s caused by A.</div>
<div class="x_MsoNoSpacing">
Try as you may, it is hard to define, </div>
<div class="x_MsoNoSpacing">
If Y causes X, or if X causes Y.</div>
<div class="x_MsoNoSpacing">
<br /></div>
<div class="x_MsoNoSpacing">
Even with a, significant correlation,</div>
<div class="x_MsoNoSpacing">
There are two other things, you need to, show causation,</div>
<div class="x_MsoNoSpacing">
Time ordering and, ruling out third variables...</div>
<div class="x_MsoNoSpacing">
<br /></div>
<div class="x_MsoNoSpacing">
And without that, </div>
<div class="x_MsoNoSpacing">
You have not shown, it’s causal.</div>
<div class="x_MsoNoSpacing">
<br /></div>
<div class="x_MsoNoSpacing">
There's a link between, <a href="http://www.reuters.com/article/2007/08/08/us-implants-suicide-idUSN0836919020070808?feedType=RSS&rpc=22&sp=true">breast implants and suicide</a>,</div>
<div class="x_MsoNoSpacing">
The risk for women, who do is, three times as high.</div>
<div class="x_MsoNoSpacing">
Is this because, they don’t like, what they’ve done,</div>
<div class="x_MsoNoSpacing">
Or are both, caused by, body dis-affection? </div>
<div class="x_MsoNoSpacing">
<br /></div>
<div class="x_MsoNoSpacing">
Even with a, significant correlation,</div>
<div class="x_MsoNoSpacing">
There are two other things, you need to, show causation,</div>
<div class="x_MsoNoSpacing">
Time ordering and, ruling out third variables...</div>
<div class="x_MsoNoSpacing">
<br /></div>
<div class="x_MsoNoSpacing">
And without that,</div>
<div class="x_MsoNoSpacing">
You have not shown, it’s causal...</div>
<div class="x_MsoNormal">
<br /></div>
<div class="MsoNormal">
<b style="mso-bidi-font-weight: normal;">Theory on Your Screen</b></div>
<div class="MsoNormal">
Lyrics by Alan Reifman</div>
<div class="MsoNormal">
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=q4mVv1wb4A0">Music of the Night</a>,” Lloyd
Webber/Hart, from <i>Phantom of the Opera</i>)</div>
<div class="MsoNormal">
<br /></div>
<div class="MsoNormal">
(SLOWLY and SOFTLY)</div>
<div class="MsoNormal">
Thinking over,<span style="color: red;"> </span>every
implication,<span style="color: red;"><br />
</span>Direct pathways,<span style="color: red;"> </span>maybe mediation,<span style="color: red;"> <br />
</span>Carefully consider, and overcome your jitter,</div>
<div class="MsoNormal">
You have to convey, all that you mean,</div>
<div class="MsoNormal">
For you record, the theory on your screen,</div>
<div class="MsoNormal">
<span style="color: red;"><br />
</span>Slowly, gently, contemplate each linkage, <br />
Parsimony, compels one, to shrinkage,<br />
Draw the paths you say, underlie the works at play,<span style="color: red;"> <br />
</span>Think of every type of route, though serpent-teen,<span style="color: red;"> <br />
</span>And listen to, the theory on your screen,</div>
<div class="MsoNormal">
<span style="color: red;"><br />
</span>Close your eyes and surrender to creative schemes! <br />
Push notions beyond, what you’ve read before! <br />
Close your eyes, let your thinking start to soar! <br />
And draw something, that gets right to the core!<span style="color: red;"> <br />
<br />
</span>Constructs, measures, hypotheses surround you ... <br />
Hear them, feel them, closing in around you ... <br />
Open up your mind, new ideas let it find, <span style="color: red;"><br />
</span>Try to make sense, of what all these things can mean,<span style="color: red;"><br />
</span>The beauty of, the theory on your screen,</div>
<div class="MsoNormal">
<span style="color: red;"><br />
</span>Try and model a journey, through a strange, new world!<span style="color: red;"> <br />
</span>Go beyond what, the world has known before! <br />
Let arrows, fill in what you think you see! <br />
Only then can you let the paths run free!<br />
<span style="color: red;"><br />
</span>Theories, concepts, find your inspiration! <br />
Later, check your, identification! <br />
Let it calculate, the paths that you estimate, </div>
<div class="MsoNormal">
Find the fruits, of the ideas, that you glean, </div>
<div class="MsoNormal">
The power of, the theory on your screen,</div>
<div class="MsoNormal">
<span style="color: red;"> </span></div>
<div class="MsoNormal">
<br />
ORCHESTRA CRESCENDO</div>
<div class="MsoNormal">
<span style="color: red;"><br />
</span>You have to convey, all that you mean,<span style="color: red;"></span></div>
<div class="MsoNormal">
For you record, the theory on your screen...<br />
<br />
<b>S-E-M</b><br />
Lyrics by Anisa Zvonkovic and Kyung-Hee Lee<br />
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=8bL2BCiFkTk">Edelweiss</a>," Rodgers/Hammerstein, from <i>The Sound of Music</i>)<br />
<br />
S-E-M, S-E-M,<br />
Every model, we're building,<br />
<br />
Small but right, fit and tight,<br />
Keeps us happy, analyzing,<br />
<br />
Helping our vitas, to bloom and grow,<br />
Bloom and grow, for tenure,<br />
<br />
Texas Tech, Virginia Tech,<br />
Bless Dr. Reifman, forever </div>
</div>
alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-27476999675048695922012-04-02T16:07:00.004-07:002018-11-23T12:57:00.256-08:00Ways to Treat Single-Indicator Variables<i>(Updated April 17, 2018)</i><br />
<br />
Often, a researcher will have one or more single-indicator variables within his or her model. It could be a demographic variable such as gender or age, or a total scale score for some social/psychological questionnaire (e.g., Rosenberg Self-Esteem Scale).<br />
<br />
With multiple manifest indicators for a latent construct, the construct is <b>automatically rendered "error-free,"</b> with measurement error segregated out into each indicator's residual "tiny bubble." Relations between constructs will be stronger when they are error free. Single-indicator variables, when left to stand alone, usually have measurement error, but are assumed to be perfectly measured.<br />
<br />
Here are five scenarios in which a researcher was interested in studying self-esteem (thanks to CRO for the photograph of the board).<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiE0_NmXQMdT-LdFcAGq2LZ85PbXKbRjSNUeUKq5HWJP2lwQjbMeMF2T2CJW5fUjli-I7vv0Fy0PCzPdcoFQ3g0TFxczIsIyq9xSR_74c4a3nhIMHZ6QGEtQtblJUOGn9OTGKS3KuFJMkM/s1600/single+indicator+constructs.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="300" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiE0_NmXQMdT-LdFcAGq2LZ85PbXKbRjSNUeUKq5HWJP2lwQjbMeMF2T2CJW5fUjli-I7vv0Fy0PCzPdcoFQ3g0TFxczIsIyq9xSR_74c4a3nhIMHZ6QGEtQtblJUOGn9OTGKS3KuFJMkM/s1600/single+indicator+constructs.jpg" width="400" /></a></div>
<br />
As shown in the photo, reliability-corrected single-indicator constructs are a way to account for measurement error in single-indicator variables (lower-right). The following is a quote from Choi et al. (2011): “To account for imperfect reliability of the scale scores, we created latent variables to represent the … constructs with each latent variable being measured by its corresponding scale score and the <span style="color: red;">residual variance of the scale score fixed to </span><b style="color: red;">(1-scale reliability) * scale variance</b><span style="color: red;"> </span>(Hayduk, 1987).” Cronbach's<a href="http://www.webpages.ttu.edu/areifman/hdfs3390relval.html#alpha"> alpha (internal consistency)</a> is often used as the reliability value. A made-up example of this procedure is shown in the photo.<br />
<br />
I previously created the following graphic to illustrate further the difference between keeping single variables as they are and using reliability correction.<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhc4C5vXCcWIWYgBqwdYNDPQkMGNzKy2dyCy4mGpbZI0hJ91U8ww-1ud3OJKOnN5gvrfSH2SSp5fanwhH9U9Js-ft0qLHBBLUwe09Xi6uKS6DHQ02P56SRfoXqlIvzqjlBiwTlpR7PpV68/s1600/reliability+corrected+sem+for+blog.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="272" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhc4C5vXCcWIWYgBqwdYNDPQkMGNzKy2dyCy4mGpbZI0hJ91U8ww-1ud3OJKOnN5gvrfSH2SSp5fanwhH9U9Js-ft0qLHBBLUwe09Xi6uKS6DHQ02P56SRfoXqlIvzqjlBiwTlpR7PpV68/s400/reliability+corrected+sem+for+blog.jpg" width="400" /></a></div>
<br />
<div class="separator" style="clear: both; text-align: center;">
<span style="font-size: 12pt;"></span></div>
<span style="color: red; font-weight: bold;">NEW! </span>Video of Todd Little speaking at Texas Tech about parceling (February 2, 2018). Dr. Little discusses strategic parceling approaches, as opposed to random parceling. Follow this <a href="https://www.depts.ttu.edu/immap/Resources/videos.php">link</a> to the video (limited to TTU); listed under Daniel Bontempo, organizer of IMMAP series.<br />
<b><br /></b>
<b>References and Resources</b><br />
<br />
Choi, K. H., Bowleg, L., & Neilands, T. B. (2011). The effects of sexism, psychological distress, and difficult sexual situations on U.S. women's sexual risk behaviors. <i>AIDS Education and Prevention, 23</i>(5), 397-411. (<a href="http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3198821/%20">LINK</a>)<br />
<br />
Cole, D. A., & Preacher, K. J. (2014). Manifest variable path analysis: Potentially serious and misleading consequences due to uncorrected measurement error. <i>Psychological Methods, 19, </i>300-315. (<a href="http://quantpsy.org/pubs/cole_preacher_2014.pdf">LINK</a>)<br />
<br />
<span class="element-citation" id="__element-citationid1889122">Hayduk L. A. (</span><span class="element-citation" id="__element-citationid1889122">1987</span><span class="element-citation" id="__element-citationid1889122">). <i><span class="ref-journal">Structural equation modeling with LISREL: Essentials and advances.</span></i> Baltimore, MD, USA: Johns Hopkins University Press. </span><br />
<span class="element-citation"><br /></span>alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-23929184468318045232011-04-06T13:26:00.000-07:002015-04-23T19:09:27.287-07:00SEM The Musical 5Here's the announcement for this year's musical!<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhIaGQnXhgNdZTBuH_rwer6WcdA-H22X7fxZmgraB3xoVsF3a_-xoBkxKm31tBSLIYqO303Vx5xiHtGlx7rEupqW1iziz_d2ZrdXoHeGz8Mu_b1bLU2uAXiEBl5QSbzLO-EuzgI1DyVDx0/s1600/SEM+Musical+5.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhIaGQnXhgNdZTBuH_rwer6WcdA-H22X7fxZmgraB3xoVsF3a_-xoBkxKm31tBSLIYqO303Vx5xiHtGlx7rEupqW1iziz_d2ZrdXoHeGz8Mu_b1bLU2uAXiEBl5QSbzLO-EuzgI1DyVDx0/s400/SEM+Musical+5.jpg" height="400px" r6="true" width="292px" /></a></div>
<div class="separator" style="clear: both; text-align: left;">
<br /></div>
<div class="separator" style="clear: both; text-align: left;">
We'll have a new song or two, plus we'll be singing some "oldies" from SEM the Musical <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a>, and <a href="http://reifman-sem.blogspot.com/2010/04/blog-post.html">4</a> (just click directly on the numbers to access previous years' lyrics).</div>
<br />
<br />
<b>SEM Musical <span style="color: red;">FIVE</span>!</b><br />
Lyrics by Alan Reifman (retread from last year)<br />
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=IKqV7DB8Iwg">Let’s Get it Started</a>,” Will Adams et al. for the Black Eyed Peas)<br />
<br />
(Softly) The models keep runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and...<br />
<br />
We’re back again, to have some fun, <br />
We’re gonna bust some rhyme, have a good time,<br />
We’re gonna sing some songs, about SEM technique, <br />
Access your inner geek, let your voices speak,<br />
SEM is different, your measurement model’s explicit, <br />
The whole model, gets tested for fit, <br />
Is it identified? We know how hard you’ve tried,<br />
Knowns and unknowns, side by side, <br />
It takes you on a ride, finally you’re satisfied, <br />
Your output’s now just fine, you’ve arrived, you can take pride…<br />
<br />
NFI, TLI, CFI, <br />
Calculate estimates, let it run, have some fun, yeah…<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
Yeah,<br />
<br />
Build your constructs, get this straight,<br />
Make sure the indicators, correlate,<br />
Draw your pathways, residuals too,<br />
Don’t leave out, the fixed 1 value,<br />
Take your time, think it through,<br />
Don’t worry if you’re new, we’ll walk with you,<br />
Step by step, right up the pyramid,<br />
For SEM, we’re really groovin,’<br />
Hope you get an acceptable solution,<br />
Submit your model and get it movin,’<br />
<br />
NFI, TLI, CFI,<br />
Calculate estimates, let it run, have some fun, yeah…<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
SEM Musical (<span style="color: red;">FIVE</span>!), SEM Musical (HERE!),<br />
Yeah…<br />
<br />
<b>SEM Pyramid of Success</b> (<a href="http://reifman-sem.blogspot.com/2007/01/welcome-to-quantitative-methods-iv.html">explanation</a>)<br />
Lyrics by Andrea Swenson<br />
(May be sung to the tune of "<a href="https://www.youtube.com/watch?v=UvyHuse6buY">Seasons of Love</a>," Jonathan Larson, from the musical "Rent")<br />
<br />
(Long opening on piano, about 40 seconds)<br />
<br />
One hundred, thirty-nine thousand, two hundred seconds,<br />
One hundred, thirty-nine thousand, moments to learn,<br />
One hundred, thirty-nine thousand, two hundred seconds,<br />
That is, how long, we sit in this class,*<br />
<br />
It starts with, correlation,<br />
Regression, and path an-al-y-sis,<br />
E-F-A, builds into,<br />
C-F-A, in time,<br />
<br />
One hundred, thirty-nine thousand, two hundred seconds,<br />
How, do you start? And, where do you go?<br />
<br />
To get to, S.....E.....M.....<br />
To get to, S.....E.....M.....<br />
To get to, S.....E.....M.....<br />
Measure it well....<br />
<br />
Pyramid of.... (slow) success,<br />
Pyramid of.... (slow) success,<br />
<br />
One hundred, thirty-nine thousand, two hundred seconds,<br />
One hundred, thirty-nine thousand, moments to learn,<br />
One hundred, thirty-nine thousand, two hundred seconds,<br />
That is, how long, we sit in this class,<br />
<br />
Starting with, correlation,<br />
Moving up to, regression,<br />
In exploring, factors,<br />
And confirming them,<br />
<br />
It’s time now, to remember,<br />
To bring it all together,<br />
Let's, bring it all together, to do SEM,<br />
<br />
Remember the pyramid (Oh you got to you got to remember the pyramid)<br />
Remember the pyramid (You know that SEM is a starts from the <i>r</i>)<br />
Remember the pyramid (regress, factor, SEM)<br />
Assess the model (Learn, learn SEM)<br />
<br />
Pyramid of success<br />
Pyramid of success (that’s how we learn, learn SEM)<br />
<br />
---<br />
<i>*Number of seconds in the class, based on 29 periods of 80 minutes each.</i><br />
<br />
<br />
<b>Hey, Hey, Heywood Cases</b><br />
Lyrics by Nora "Felix" Phillips<br />
(May be sung to the <a href="https://www.youtube.com/watch?v=ksJ6QP8BYn0">theme</a> from "The Monkees," Boyce/Hart)<br />
<br />
Let it run, the computations go through,<br />
You get an error message, it leaves you feeling blue,<br />
<br />
Hey Hey Heywood Cases!<br />
Bringing, my AMOS, model down,<br />
With your, negative variance,<br />
You know that, isn't allowed,<br />
<br />
Mis-specification,<br />
Of the model, that you've drawn,<br />
Or maybe, your own sample,<br />
Was just, a tad bit, too small?<br />
<br />
Hey Hey Heywood Cases!<br />
I won't let you bring me down,<br />
I can constrain, residuals,<br />
To a, small positive, amount!<br />
<br />
---<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgW_LU2bzmFtszHkD8Iw2fQG5zhUWiB1ZT8NJTZqmSiAJXvNHQl0meFsy9HPxwNklyJ956CMsdRP14k-hPOe0-aCogrxLDPBohZ8llomzL38CclpkA9xVZ7Pp6DRH8JrTsD2kBt6rJfBxw/s1600/SEM+nestedness.jpg" imageanchor="1" style="clear: left; cssfloat: left; float: left; margin-bottom: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgW_LU2bzmFtszHkD8Iw2fQG5zhUWiB1ZT8NJTZqmSiAJXvNHQl0meFsy9HPxwNklyJ956CMsdRP14k-hPOe0-aCogrxLDPBohZ8llomzL38CclpkA9xVZ7Pp6DRH8JrTsD2kBt6rJfBxw/s200/SEM+nestedness.jpg" height="141px" i8="true" width="200px" /></a></div>
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<br />
<b><br /></b>
<b>Nestedness</b><br />
Lyrics by Alan Reifman<br />
May be sung to the tune of “<a href="http://www.youtube.com/watch?v=ObAuJDCHw2I">Yesterday</a>” (Lennon/McCartney)<br />
<br />
Nestedness,<br />
It’s the way, models can be compared,<br />
Should new paths be added in or spared?<br />
The delta-test needs nestedness,<br />
<br />
Can’t you see?<br />
One model might have simplicity,<br />
But more paths increase fidelity,<br />
Which one to choose, the chi-square’s key,<br />
<br />
Inside, the big one, the small one, is self-contained,<br />
One has, extra paths, the other, does not maintain...<br />
<br />
Nestedness,<br />
To the baseline, you can only add,<br />
Or only subtract, paths you once had,<br />
You can’t do both, for nestedness,<br />
<br />
Inside, the big one, the small one, is self-contained,<br />
One has, extra paths, the other, does not maintain...<br />
<br />
Look, shall we?<br />
One model could have, paths “A” and “B,”<br />
They would nest in, model “A/B/C,”<br />
A/B’s contained, in A/B/C…<br />
<br />
<br />
<b>Maximum Likelihood</b><br />
Lyrics by Alan Reifman<br />
May be sung to the tune of “<a href="http://www.youtube.com/watch?v=malnd19Ttyk">Pink Houses</a>” (John Mellencamp)<br />
<br />
The computer, runs your model, looking for a solution,<br />
It seeks to maximize, or maybe minimize,<br />
Some function, seen in, a distribution,<br />
<br />
You have least squares, which tries to put, the best-fit line near the dots,<br />
But ML, seeks equations, so your findings, will come out on top,<br />
<br />
Oh, maximum likelihood, that’s what we use,<br />
Maximum likelihood, it tends to confuse,<br />
Maximum likelihood, underlying values, that make your results, most probable,<br />
And that’s, big news!<br />
<br />
Sir Ronald Fisher, statistician, <br />
Developed the, ML perspective,<br />
It will iterate, till it’s really great,<br />
But it’s so, calculation intensive,<br />
<br />
For a long time, ML sat there, <br />
Its steps were, so hard to reckon,<br />
But computers, came along, and sped things up,<br />
And now ML, runs in mere seconds, <br />
<br />
Oh, maximum likelihood, that’s what we use,<br />
Maximum likelihood, it tends to confuse,<br />
Maximum likelihood, underlying values, that make your results, most probable,<br />
And that’s, big news!<br />
<br />
Instrumental<br />
<br />
Well there are data, and more data,<br />
What do they show?<br />
With its complex math, on a tricky path,<br />
ML tells you, what you, need to know,<br />
<br />
Oh yeah,<br />
<br />
Well some data, might be missing,<br />
But there’s no need, for frustration,<br />
’Cause you can, estimate the means, and intercepts,<br />
To get ML, with full, information,<br />
<br />
Oh, maximum likelihood, that’s what we use,<br />
Maximum likelihood, tends to confuse,<br />
Maximum likelihood, underlying values, that make your results, most probable,<br />
And that’s, big news!alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-9735514653767714902010-04-20T10:17:00.000-07:002013-04-30T08:59:15.147-07:00SEM The Musical 4<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgbq6NzTxlXtgWmWvnXbl9G2XzazMGX_SfiwalNf6JVOEPaNN7IgmWxyBL8jbaPS8uH9v5gct1V2p9qLbUPgztBto2R4m2ovIv24Ayyx8RRj7VK1hz_1YG5PEHsUnmyfeiC2CEvGRjX0fk/s1600/sem+the+musical+4.jpg"><img alt="" border="0" id="BLOGGER_PHOTO_ID_5462270248204504242" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgbq6NzTxlXtgWmWvnXbl9G2XzazMGX_SfiwalNf6JVOEPaNN7IgmWxyBL8jbaPS8uH9v5gct1V2p9qLbUPgztBto2R4m2ovIv24Ayyx8RRj7VK1hz_1YG5PEHsUnmyfeiC2CEvGRjX0fk/s400/sem+the+musical+4.jpg" style="cursor: hand; display: block; height: 362px; margin: 0px auto 10px; text-align: center; width: 400px;" /></a><br />
<br />
Below is a sneak peek at our new songs for this year. We'll also be singing some "oldies" from SEM the Musical <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">1</a>, <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">2</a>, and <a href="http://reifman-sem.blogspot.com/2009/04/here-are-some-newly-written-songs-for.html">3</a> (just click directly on the numbers to access previous years' lyrics).<br />
<br />
<b>SEM Musical FOUR!</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=IKqV7DB8Iwg">Let’s Get it Started</a>,” Will Adams et al. for the Black Eyed Peas)<br />
<br />
(Softly) The models keep runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and runnin-runnin, and...<br />
<br />
We’re back again, to have some fun, <br />
We’re gonna bust some rhyme, have a good time,<br />
We’re gonna sing some songs, about SEM technique, <br />
Access your inner geek, let your voices speak,<br />
SEM is different, your measurement model’s explicit, <br />
The whole model, gets tested for fit, <br />
Is it identified? We know how hard you’ve tried,<br />
Knowns and unknowns, side by side, <br />
It takes you on a ride, finally you’re satisfied, <br />
Your output’s now just fine, you’ve arrived, you can take pride…<br />
<br />
NFI, TLI, CFI, <br />
Calculate estimates, let it run, have some fun, yeah…<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
Yeah,<br />
<br />
Build your constructs, get this straight,<br />
Make sure the indicators, correlate,<br />
Draw your pathways, residuals too,<br />
Don’t leave out, the fixed 1 value,<br />
Take your time, think it through,<br />
Don’t worry if you’re new, we’ll walk with you,<br />
Step by step, right up the pyramid,<br />
For SEM, we’re really groovin,’<br />
Hope you get an acceptable solution,<br />
Submit your model and get it movin,’<br />
<br />
NFI, TLI, CFI,<br />
Calculate estimates, let it run, have some fun, yeah…<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
SEM Musical (FOUR!), SEM Musical (HERE!),<br />
Yeah…<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgL6xRz9ucMxJxiT8cEez7WXZ0-4KaPhgN2G9ZN_ylFec3gkS_SAar9XufKGr3xsJ62AlBQZpvFT2thNLbdJfRSpumQkVDbSvYw-TM_IBC6gshSM77sIB6FTahzCSOIBrYOVgdBW0-8zKc/s1600/amos.jpg" imageanchor="1" style="clear: right; float: right; margin-bottom: 1em; margin-left: 1em;"><img border="0" height="343" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgL6xRz9ucMxJxiT8cEez7WXZ0-4KaPhgN2G9ZN_ylFec3gkS_SAar9XufKGr3xsJ62AlBQZpvFT2thNLbdJfRSpumQkVDbSvYw-TM_IBC6gshSM77sIB6FTahzCSOIBrYOVgdBW0-8zKc/s400/amos.jpg" width="400" /></a></div>
<b>Once You Work in AMOS</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=W8gZ2Ie232w">Once in Love with Amy</a>,” Frank Loesser)<br />
<br />
Once you work, in AMOS,<br />
Find every quirk, in AMOS,<br />
Construct by construct, you can draw your picture,<br />
Using all the gadgets, is fun,<br />
<br />
Learn the rules, in AMOS,<br />
Use all the tools, in AMOS,<br />
Circles and boxes, and you can add arrows,<br />
Soon your model’s, ready to run,<br />
<br />
The moving truck, the sizer, and the bubble,<br />
Your choices, are vast,<br />
And even if, you find yourself in trouble,<br />
You can fix things fast,<br />
<br />
So, once you work, in AMOS,<br />
Find every quirk, in AMOS,<br />
Each time you use it, your skills are expanded,<br />
And you’ll understand, what you see,<br />
Cause, in the end, it’s fixed, or it's free…<br />
<br />
<b>Prof. Reifman</b><br />
Lyrics by Kim Corson and Janis Henderson<br />
(May be sung to the tune of "<a href="http://www.youtube.com/watch?v=dQsjAbZDx-4">Fernando</a>," Ulvaeus, Andersson, & Anderson, for ABBA)<br />
<br />
Can you hear the songs, Prof. Reifman?<br />
We remember long ago, in intro stats you sang like this,<br />
At the front of class, Prof. Reifman,<br />
You were humming to yourself, and softly strumming air guitar,<br />
We could hear the distant drums, <br />
And suddenly, the answers didn't seem so far,<br />
<br />
We’re much closer now, Prof. Reifman,<br />
Every box, every circle, seems to make more sense to us,<br />
We are not afraid, Prof. Reifman,<br />
We sit here so full of life; all of us are prepared to try,<br />
And we're not ashamed to say,<br />
The songs of SEM the Musical 4 helped us get by,<br />
<br />
There was something in the air that day,<br />
The fog went away, Prof. Reifman,<br />
He was talking about SEM,<br />
And our heads didn't swim, Prof. Reifman,<br />
<br />
Though we never thought that we would grasp, degrees of freedom,<br />
We can calculate them now, in fact, we just subtract, Prof. Reifman,<br />
We can calculate them now, in fact, we just subtract, Prof. Reifman,<br />
<br />
When we're old and grey, Prof. Reifman,<br />
And for many years we haven't played in your "rock band,"<br />
We'll still hear the strums, Prof. Reifman,<br />
And we'll recall learning AMOS, like Emeril, can go "Bam!",<br />
And we'll still call point-0-0-0 "Paula Abdul significance,"<br />
<br />
There was something in the air that day,<br />
The fog went away, Prof. Reifman,<br />
He was talking about SEM,<br />
And our heads didn't swim, Prof. Reifman,<br />
<br />
Though we never thought that we would grasp, under-identification,<br />
We now see it's when a model's flown, with too much unknown, Prof. Reifman,<br />
<br />
There was something in the air that day,<br />
The fog went away, Prof. Reifman,<br />
He was talking about SEM,<br />
And our heads didn't swim, Prof. Reifman,<br />
<br />
Though we never thought that we would grasp, degrees of freedom,<br />
We can calculate them now, in fact, we just subtract, Prof. Reifman,<br />
We can calculate them now, in fact, we just subtract, Prof. Reifman,alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-45760546572114516772010-04-08T17:17:00.000-07:002017-06-10T16:15:14.866-07:00Graphics Programs for Drawing SEM DiagramsOn the SEMNET discussion listserv around April 3-4, 2010, someone asked about graphics programs for drawing structural-equation-model diagrams, and other participants sent in suggestions. I, personally, find AMOS and PowerPoint to be good. However, if anyone wants to examine additional programs, here are some:<br />
<br />
<a href="http://web.missouri.edu/~kolenikovs/graphviz_sem.html">GraphViz</a><br />
<br />
<a href="http://www.benitoarias.com/tutoriales/sem_conceptdraw/sem_conceptdraw.html">Concept Draw</a><br />
<br />
<a href="http://cmap.ihmc.us/conceptmap.html">Concept Map</a> (perhaps more appropriate for illustrating theory construction than actual SEM drawing)<br />
<br />
<a href="http://www.eazydraw.com/">Easy Draw</a> (seems like a very general graphic-arts program)<br />
<br />
<b><span style="color: red;">ADDED 6/10/2017:</span></b> <a href="http://blog.marjoleinfokkema.nl/2015/08/creating-nice-sem-diagrams-in.html">PowerPoint Tips</a> for Displaying SEM Modelsalanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-83119543401180194972010-03-07T17:41:00.000-08:002017-04-29T21:27:04.410-07:00Further Issues for Full Structural ModelsNow that we've learned the basics of full structural models, we'll be taking up the following topics in the coming weeks:<br />
<br />
<a href="http://reifman-sem.blogspot.com/2008/02/today-lets-take-some-time-to-talk-about.html">Maximum Likelihood Estimation</a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2008/03/well-next-consider-what-is-known-as.html">Equivalent models</a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2012/04/we-recently-discussed-how-to-handle.html">Handling single-indicator variables </a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2007/03/as-you-begin-work-on-your-next.html">Negative variances</a> (Heywood Cases)<br />
<br />
<a href="http://reifman-sem.blogspot.com/2016/04/mediational-models.html">Mediation</a><br />
<br />
<a href="http://reifman-sem.blogspot.com/2008/03/this-week-well-learn-about-equality.html">Equality constraints</a> (these lecture notes also touch briefly on longitudinal models and multiple-group analyses)<br />
<br />
<a href="http://reifman-sem.blogspot.com/2007/04/our-next-topic-is-longitudinal-sem.html">Longitudinal (panel) models</a><br />
<br />
<b>Multiple-Group Modeling</b> (see notes on equality constraints above; Kyle Gillett dissertation in links section to the right; and this <a href="http://www.oberlin.edu/faculty/ndarling/lab/eara.pdf">article</a>, which we'll revisit from when we learned about measurement and structural models)<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiaeMn_gTLRE5fRTrnfyMK-A84hLcdjeu2Dy5epw4550woS7joJzXFQvO5_oxOQwc1Y9oh_KTeQo2mh1Nsk5sQdivdGKLL0TbCN3f3Jf4kL74UROJLfWvUvHdjIPR8zlHopnlO5nbQ_v54/s1600/partially+constrained+model+(SEM).jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="278" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiaeMn_gTLRE5fRTrnfyMK-A84hLcdjeu2Dy5epw4550woS7joJzXFQvO5_oxOQwc1Y9oh_KTeQo2mh1Nsk5sQdivdGKLL0TbCN3f3Jf4kL74UROJLfWvUvHdjIPR8zlHopnlO5nbQ_v54/s1600/partially+constrained+model+(SEM).jpg" width="400" /></a></div>
<br />
<br />
<a href="http://reifman-sem.blogspot.com/2009/04/today-we-will-take-up-topic-of-dyadic.html">Dyadic analysis in SEM</a> (Actor-Partner Interdependence Model)<br />
<br />
Running an AMOS model off of a published <a href="http://reifman-sem.blogspot.com/2009/03/as-i-alluded-to-recently-if-one-wanted.html">correlation/covariance matrix</a> from the literature<br />
<br />
Software comparison: <a href="http://reifman-sem.blogspot.com/2009/04/yoona-chin-newly-minted-ph.html">AMOS vs. Mplus</a> <br />
<br />
<i>Advanced Applications</i><br />
<br />
Latent Growth Modeling (<a href="http://oregonstate.edu/dept/hdfs/papers/lgcgeneral.pdf">here</a> and <a href="http://www.unc.edu/~curran/example.htm">here</a>)<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEisF8xzBaRmXz5tOoGFqYYi_yHz_fzf5hyuYDfWHYALnXqx8kA6Me0RRinXca5VyXt_DTlpfI-RihJc3GwAYekv9FO-x3hcoJ1hr_q4IMUht4JnrmPsSOvcLUVgVg1aoWOkr4rEdf0HWD4/s1600/growth+curve+graphic.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="331" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEisF8xzBaRmXz5tOoGFqYYi_yHz_fzf5hyuYDfWHYALnXqx8kA6Me0RRinXca5VyXt_DTlpfI-RihJc3GwAYekv9FO-x3hcoJ1hr_q4IMUht4JnrmPsSOvcLUVgVg1aoWOkr4rEdf0HWD4/s400/growth+curve+graphic.jpg" width="400" /></a></div>
<br />
Compared to the more piecemeal/incremental cross-lagged panel models for longitudinal analysis, latent growth models test for predictors and correlates of respondents' long-term growth trajectories (see <a href="http://runrev.com/newsletter/october/issue35/newsletter3.php">cannon-ball analogy</a>). Thanks to Tim and Xiaohui for photographing the board after class on May 1, 2012. Some illustrative references on latent growth modeling are: <br />
<br />
<span style="color: #cc0000;">Barnes, G. M., Reifman, A. S., Farrell, M. P., & Dintcheff, B. A. (2000). The effects of parenting on the development of adolescent alcohol misuse: A six-wave latent growth model. <i>Journal of Marriage and the Family, 62</i>, 175-186. </span><br />
<br />
<div style="color: #cc0000;">
Wampler, R. S., Munsch, J., & Adams, M. (2002). Ethnic
differences in grade trajectories during the transition to junior high.
<i>Journal of School Psychology, 40</i>, 213-237.<br />
<br />
<span style="color: black;"><a href="http://reifman-sem.blogspot.com/2017/04/partial-least-squares-small-sample.html">Partial Least Squares</a> (Alternative to Conventional SEM for Small Samples)</span></div>
alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-77506186735180781052010-03-04T09:27:00.000-08:002014-04-18T13:56:53.384-07:00University-Quality AssignmentThe following is the model for the new assignment. You will run the model twice, once without the three red-dashed paths and once with them. We will learn about <a href="http://reifman-sem.blogspot.com/2007/03/as-weve-discussed-part-of-latest.html">comparative model testing</a>.<br />
<br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhR0b-Y5c3HGyqis_tYby25-Nxst9C1fvSZqrqnzLujjd-NoMsw4Cvj0SEVP1Fzq-nlD22mc78iLhS64W23Fi_0pCr1efPwPIlhZOp8Bjp7jd1I_sJgyJ_6RfX0v2wx-pxIvxj7_qAYZP0/s1600-h/univ+2006+model.jpg"><img alt="" border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhR0b-Y5c3HGyqis_tYby25-Nxst9C1fvSZqrqnzLujjd-NoMsw4Cvj0SEVP1Fzq-nlD22mc78iLhS64W23Fi_0pCr1efPwPIlhZOp8Bjp7jd1I_sJgyJ_6RfX0v2wx-pxIvxj7_qAYZP0/s400/univ+2006+model.jpg" id="BLOGGER_PHOTO_ID_5444831853612871778" style="cursor: hand; display: block; height: 400px; margin: 0px auto 10px; text-align: center; width: 367px;" /></a><br />
<br />
Here's a direct link to the figure we recently looked at regarding <a href="http://reifman-sem.blogspot.com/2007/02/now-that-were-beginning-to-learn-how-to.html">where variances are located</a> in full structural models, as well as how degrees of freedom are <a href="http://reifman-sem.blogspot.com/2007/04/some-of-students-wanted-review-of.html">determined</a> in a full structural model.<br />
<br />
The model also makes salient the issue of outliers, in particular that Harvard's endowment (and to a lesser extent those of a few other institutions) are so much larger than most others. Harvard and these other elite universities have endowments in the billions, whereas many other schools have endowments well under 1 billion. This <a href="http://www.ohio.edu/plantbio/staff/mccarthy/quantmet/lectures/Error&Power.pdf">document</a> discusses approaches to handling outliers; in the past we've used winsorizing, but this year, we'll use a square-root transformation (already implemented in the data set).<br />
<br />
There is some <a href="http://www.nytimes.com/2009/10/09/education/09harvard.html">evidence</a> that the depressed economy may be "winsorizing" Harvard's endowment, but this didn't occur early enough to be reflected in the data set.<br />
<br />
<b><span style="color: blue;">UPDATE (2014):</span></b> Here's an illustration of the terminology you should use, regarding <i>measurement</i> and <i>structural</i> portions of the model, and <i>factor loadings</i> vs. <i>structural paths</i>. (Thanks to Satabdi for the photo.)<br />
<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhW7VtbB0E2-uTVb1YQ4rT5ZIWwq9WLpGwRWJ6pUqmAHslKJG_baivK1LMq7Tfcx70CXgC2xsoiQxuLv2LDokKVfgmV-EXaKZH1I2Dvl74TbJ-KtJb2xb_Nw614tbUxEAc2iHiYs45BM64/s1600/meas+&+struc.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhW7VtbB0E2-uTVb1YQ4rT5ZIWwq9WLpGwRWJ6pUqmAHslKJG_baivK1LMq7Tfcx70CXgC2xsoiQxuLv2LDokKVfgmV-EXaKZH1I2Dvl74TbJ-KtJb2xb_Nw614tbUxEAc2iHiYs45BM64/s1600/meas+&+struc.jpg" height="257" width="400" /></a></div>
<br />alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-47219974876795500742010-02-09T09:54:00.000-08:002013-02-06T14:53:50.863-08:00Creating Subscales: Exact Factor Scores vs. Unit WeightingHere's a graphic display I created to illustrate the difference between using exact factor scores to make subscales vs. unit weighting. I also found a good <a href="http://pareonline.net/pdf/v14n20.pdf">online article</a> by DiStefano et al. and a thorough <a href="http://core.ecu.edu/psyc/wuenschk/MV/FA/FA-spss.pptx">PowerPoint show</a> by Wuensch on the subject. You can click on the graphics below to enlarge them (<i>note that there are TWO slides to click on, one on top of the other</i>).<br />
<br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjsx5tCwvXlPXPO70PX-E39UKUUpbcN5oxCV1klHxKJymQUys9bt_7qKIhwTnxO3J2d0MtnK6wYPJJNStMyDzuB8bS_PJGlM9070uNAgjgC4A1AzTiRu9oiJDO7wWlcmvYZyDjBrS2vbfQ/s1600-h/factor+scores+1.jpg"><img alt="" border="0" id="BLOGGER_PHOTO_ID_5436771492047026946" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjsx5tCwvXlPXPO70PX-E39UKUUpbcN5oxCV1klHxKJymQUys9bt_7qKIhwTnxO3J2d0MtnK6wYPJJNStMyDzuB8bS_PJGlM9070uNAgjgC4A1AzTiRu9oiJDO7wWlcmvYZyDjBrS2vbfQ/s400/factor+scores+1.jpg" style="cursor: hand; display: block; height: 300px; margin: 0px auto 10px; text-align: center; width: 400px;" /></a><br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiPkNxQcvearmM_gqeTuhiOCQ3009P1lQDXqig3Gg8qqNn28KZ3ijnPVAcvBHDKtsHKBdq6fNebVzASUJHxAKafDEzz09Kflp8zEo3HUs2acWo6kOayq6Wr0euYfXi0hwFMnVjAtojEewk/s1600-h/factor+scores+2.jpg"><img alt="" border="0" id="BLOGGER_PHOTO_ID_5436771427757826194" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiPkNxQcvearmM_gqeTuhiOCQ3009P1lQDXqig3Gg8qqNn28KZ3ijnPVAcvBHDKtsHKBdq6fNebVzASUJHxAKafDEzz09Kflp8zEo3HUs2acWo6kOayq6Wr0euYfXi0hwFMnVjAtojEewk/s400/factor+scores+2.jpg" style="cursor: hand; display: block; height: 300px; margin: 0px auto 10px; text-align: center; width: 400px;" /></a>alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-7278913987225723132009-04-20T23:22:00.001-07:002015-04-23T19:25:19.930-07:00SEM The Musical 3Here are some newly written songs (including one by a student) for <b>SEM The Musical 3</b>, which we'll perform on Wednesday. We'll also do some "oldies" from the <a href="http://reifman-sem.blogspot.com/2007/04/for-roughly-last-half-hour-of-period-on.html">first</a> and <a href="http://reifman-sem.blogspot.com/2008/04/on-wednesday-april-23-we-will-present.html">second</a> annual musicals.<br />
<br />
<b>The SEM Way</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="https://www.youtube.com/watch?v=hnFZsrs32Co">Let’s Live for Today</a>,” Mogol/Shapiro/Julien, popularized by the Grass Roots)<br />
<br />
You’ve got your sets of measures, some constructs they could form,<br />
Plus, indices of fitness, to compare to a norm,<br />
You draw yourself a model, with circles, squares, and paths,<br />
The AMOS program handles, the complicated math,<br />
If you get too many errors, you can express your wrath,<br />
<br />
1-2-3-4<br />
<br />
Analyze your work, the SEM way,<br />
Analyze your work, the SEM way,<br />
Don’t forget to, check your RM-SEA,<br />
Analyze your work, the SEM way…<br />
<br />
<b>AMOS is Ideal</b><br />
Lyrics by Susan Murray<br />
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=lydBPm2KRaU">Jesus, Take the Wheel</a>,” James/Lindsey/Sampson, popularized by Carrie Underwood) <br />
<br />
She was working last Friday on her laptop battery, <br />
On homework to achieve,<br />
Her constructs were getting muddy, with her model nowhere near complete, <br />
Fifty specs to go and she was running low on patience and caffeine,<br />
<br />
It was complex and unclear, <br />
She had a constraint and the software caused the tension, <br />
But they say SAS is unsurpassed, <br />
Before she knew it she was closing down that darn software SAS, <br />
<br />
She saw the latent variables flash before her eyes, <br />
She didn’t even have time to imply, <br />
She was so impaired, <br />
She suddenly was aware,<br />
<br />
AMOS is ideal, <br />
Take causation from my hands, <br />
‘Cause I can’t do this on my own,<br />
I’m letting go,<br />
<br />
No coding song and dance, <br />
To learn about the population, <br />
AMOS is ideal,<br />
<br />
A cross-lagged panel model she pulled out of the folder,<br />
And like Emeril she went BAM! nonstop, <br />
She cried like a baby when she saw the RMSEA drop,<br />
Her hypothesis and all the parameters, <br />
She now could weigh,<br />
She could model all day,<br />
In a Paula Abdul light,<br />
Software to exchange,<br />
Arbuckle already did the fight,<br />
<br />
AMOS is ideal, <br />
Take causation from my hands, <br />
‘Cause I can’t do this on my own,<br />
I’m letting go,<br />
<br />
No coding song and dance, <br />
SAS I won’t depend upon,<br />
Oh, AMOS is ideal,<br />
SAS, I’m saying no, <br />
<br />
No coding song and dance <br />
SAS I won’t depend upon,<br />
My loyalty is withdrawn,<br />
AMOS is ideal,<br />
<br />
Oh, don’t you take it from me, <br />
Find my μ <br />
<br />
<b>Equal</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “<a href="http://www.youtube.com/watch?v=y7ZEVA5dy-Y">Mercy</a>,” Duffy/Booker)<br />
<br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgbN3L7Qd0JxBo4Xdh6wUCajb33DbeqhBPf9UrqIEWXyyT1_ZDR4_kwCy0zSiPxvKX1hGhTzA3D7UHEVTbN5npLB3K5JYZCT0r9kxVu3wrYxJidSch_AFkOjz5o0Gukoel2yu1B7ENXu9g/s1600-h/multi-group+model+example.gif"><img alt="" border="0" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgbN3L7Qd0JxBo4Xdh6wUCajb33DbeqhBPf9UrqIEWXyyT1_ZDR4_kwCy0zSiPxvKX1hGhTzA3D7UHEVTbN5npLB3K5JYZCT0r9kxVu3wrYxJidSch_AFkOjz5o0Gukoel2yu1B7ENXu9g/s400/multi-group+model+example.gif" id="BLOGGER_PHOTO_ID_5184102781029530562" style="cursor: hand; display: block; margin: 0px auto 10px; text-align: center;" /></a><br />
A-A-A, B-B-B, C-C-C, D-D-D<br />
<br />
I’ve got two, paths that you can view,<br />
I think their strength might be the same, and that’s the frame,<br />
You’ve got, to see through,<br />
<br />
Let’s run the model free, paths can be any sized,<br />
Then run equalized,<br />
<br />
We need a way to choose, a way to compare,<br />
Test delta chi-square,<br />
<br />
You’ve got me constrained to be equal,<br />
Why won’t you release me?<br />
You’ve got me constrained to be equal,<br />
Why won’t you release me?<br />
Can’t you rele-e-e-e-ase me?<br />
<br />
Lower chi-square, will always be there,<br />
When you let the paths go free, but you must see,<br />
True sig-nif-i-cance,<br />
<br />
When the chi-square’s non-sig, the one to retain,<br />
Is where you constrain,<br />
<br />
But if the change is big, p’s under oh-five,<br />
Free paths shall survive,<br />
<br />
You’ve got me constrained to be equal,<br />
Why won’t you release me?<br />
You’ve got me constrained to be equal,<br />
Why won’t you release me?<br />
Can’t you rele-e-e-e-ase me?<br />
<br />
<b>It’s Still SEM to Me</b><br />
Lyrics by Alan Reifman<br />
(May be sung to the tune of “It’s Still Rock and Roll to Me,” Billy Joel)<br />
<br />
It’s a way to show, inter-relations,<br />
In a set of, latent constructs,<br />
It gives you, some global fit statistics,<br />
Does your model, really stack up?<br />
Squares and circles, now you’re off and you’re running,<br />
Will your results, be routine or be stunning?<br />
<br />
LISREL, M-PLUS, EQS, or AMOS,<br />
It’s still SEM to me…..alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-90316104081403094472009-04-17T10:49:00.000-07:002017-04-28T23:49:16.147-07:00AMOS vs. Mplus ProgramsYoona Chin, a newly minted (2009) Ph.D. recipient from our department, will be giving a guest presentation today on differences between <b>AMOS</b>, the program we've been using in class, and <b><a href="http://www.statmodel.com/">Mplus</a></b>, an increasingly popular program. She's put together a very elaborate PowerPoint slide show, a few key graphics I wanted to put online (with Yoona's permission). You may click on the following images to enlarge them. <br />
<br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEh8tDOxoKd-v1VfE3M_-mDurgKrDc0ufN-c88ng8jgTrCx_evbNRpknmwwxbEwRR5IODjbmYZpyz6-rHQFjk79UCR0li7VaetZE-QEo9rVh8UkH2b2G_iLF54ztIxW4T8Y7_MLvxAPrBFA/s1600-h/yoona+1.jpg"><img alt="" border="0" id="BLOGGER_PHOTO_ID_5325719428991892226" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEh8tDOxoKd-v1VfE3M_-mDurgKrDc0ufN-c88ng8jgTrCx_evbNRpknmwwxbEwRR5IODjbmYZpyz6-rHQFjk79UCR0li7VaetZE-QEo9rVh8UkH2b2G_iLF54ztIxW4T8Y7_MLvxAPrBFA/s400/yoona+1.jpg" style="cursor: hand; cursor: pointer; display: block; height: 274px; margin: 0px auto 10px; text-align: center; width: 400px;" /></a><br />
<br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgnIh-8wxzWQmzl1-EG96W6A93bdHJE2sp6wZctVTsScs1k_SuVCUZH4hzPYn0xHvxMcCLU0YN5dbFrJ4JG3_6b0blAA38cTdQ8ou8l1Y00I0Lu2FrZnfrSYPu1blAencvVSmRnZhZs8wE/s1600-h/yoona+2.jpg"><img alt="" border="0" id="BLOGGER_PHOTO_ID_5325719328238001298" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgnIh-8wxzWQmzl1-EG96W6A93bdHJE2sp6wZctVTsScs1k_SuVCUZH4hzPYn0xHvxMcCLU0YN5dbFrJ4JG3_6b0blAA38cTdQ8ou8l1Y00I0Lu2FrZnfrSYPu1blAencvVSmRnZhZs8wE/s400/yoona+2.jpg" style="cursor: hand; cursor: pointer; display: block; height: 300px; margin: 0px auto 10px; text-align: center; width: 400px;" /></a><br />
<br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhDkC1nn0VS9qfSu8NZ2PTqT6NkX62v278JDmDIiq-lb69Wmt_R22rA19rLNoSWHQuTGrdpoQQJ8yVdkFg21D2AAqjXE8H0zmEm3AMN1hgkhCgnkhXMNh3oqXSpn5-2SGW3kl_IZGgYmZc/s1600-h/yoona+3.jpg"><img alt="" border="0" id="BLOGGER_PHOTO_ID_5325719211260935442" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhDkC1nn0VS9qfSu8NZ2PTqT6NkX62v278JDmDIiq-lb69Wmt_R22rA19rLNoSWHQuTGrdpoQQJ8yVdkFg21D2AAqjXE8H0zmEm3AMN1hgkhCgnkhXMNh3oqXSpn5-2SGW3kl_IZGgYmZc/s400/yoona+3.jpg" style="cursor: hand; cursor: pointer; display: block; height: 300px; margin: 0px auto 10px; text-align: center; width: 400px;" /></a><br />
<b><span style="color: blue;">UPDATE (4/28/2017):</span></b> Now that we're using Mplus as our main program for the course, I wanted to go over some of the advanced applications of this software. I find Mplus more cumbersome to use than AMOS or ONYX, but people claim Mplus makes up for this disadvantage through its versatility in offering numerous advanced techniques. According to this <a href="http://sites.education.uky.edu/apslab/files/2015/11/IntroToMplus2016.pptx">University of Kentucky PowerPoint slideshow</a>, Mplus offers the following techniques (beyond basic SEM). I have added further links on the specific techniques:<br />
<ul>
<li><i>Multilevel Modeling </i>(<a href="https://www.youtube.com/watch?v=kiC71FRItFM">YouTube</a>)</li>
<li><i>Item Response Theory</i> (<a href="https://familycenter.byu.edu/Documents/A%20Different%20Way%20to%20Think%20About%20Measurement%20Development%20Olson%20Busby%20and%20Chiu%20IRT.pptx">PowerPoint</a>)</li>
<li><i>Bayesian Analysis </i>(<a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4158865/">Article</a>)</li>
<li><i>Latent Class Analysis</i> (<a href="http://www.restore.ac.uk/latentvariablemodels/workshopfiles/Latent%20Class%20Analaysis_Orla%20McBride/Course%20Presentation/Powerpoint%20slides%20-%20LCA%20workshop.pdf">PDF of PPT presentation</a>)<i> </i></li>
<li><i>Mixture Modeling </i>(<a href="http://r2ed.unl.edu/presentations/2011/RMS/012111_Kupzyk/012111_Kupzyk.pdf">PDF of PPT presentation</a>)</li>
<li><i>Monte Carlo Simulations</i> (<a href="http://www.statpower.net/Content/319SEM/Lecture%20Notes/MonteCarlo.pdf">PDF of PPT presentation</a>)<i> </i></li>
<li><i>Multiple Imputation for Missing Data</i> (<a href="https://www.youtube.com/watch?v=BS_bXe1dd7U">YouTube</a>)</li>
</ul>
alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-26226781768802416012009-04-15T10:31:00.000-07:002016-04-29T20:58:53.666-07:00Dyadic AnalysisToday we will take up the topic of dyadic analysis in SEM, particularly something known as the Actor-Partner Interdependence Model (APIM). We will draw upon the following article, which is available via the TTU library.<br />
<br />
Popp, D., Laursen, B., Burk, W. J., Kerr, M., & Stattin, H. (2008). Modeling homophily over time with an Actor-Partner Interdependence Model. <i>Developmental Psychology, 44</i>, 1028-1039.<br />
<br />
The <a href="http://courses.ttu.edu/hdfs3390-reifman/data.htm">notes</a> from my Methods class on <i>unit of analysis</i> may be helpful for this topic.<br />
<br />
An important issue is whether the two partners in a dyad are distinguishable (i.e., non-exchangeable), as opposed to being indistinguishable (exchangeable). See David Kenny's webpage on <a href="http://davidakenny.net/dyad.htm">dyadic analysis</a> (particularly Topic 3) and the <a href="http://davidakenny.net/webinars/powerpoints/Dyad/Standard/APIM.ppt">slides</a> from a talk he gave. As Kenny, Kashy, and Cook (2006) state in their book <i><a href="http://www.guilford.com/cgi-bin/cartscript.cgi?page=pr/kenny2.htm&dir=research/res_quant&cart_id=995064.26460">Dyadic Data Analysis</a></i>:<br />
<br />
<i>When dyad members are distinguishable, we estimate the path model or CFA model for each of the two members combined in a single model... However, when members are indistinguishable, it is less clear exactly how to do the analysis. The use of SEM with indistinguishable or exchangeable dyad members has generally been viewed pessimistically...</i> (p. 111).<br />
<br />
A suggested reference in this regard is:<br />
<br />
Olsen, J. A., & Kenny, D. A. (2006). Structural equation modeling with interchangeable dyads. <i>Psychological Methods, 11</i>, 127-141.<br />
<br />
In honor of David Kenny's contribution to dyadic analysis, I've written a <a href="http://socialpsychlyrics.blogspot.com/2008/01/at-recent-social-psychologists-in-texas.html">song</a>.<br />
<br />
As another example of an APIM-type model, see Hye-Sun Ro's dissertation in the online collection to the right.<br />
<br />
<b style="color: #38761d;">UPDATE 1 (4/24/12): </b>As we discussed in class, data from both members of a couple (or from a repeated-measures/panel design of individual participants) can be organized two ways. I have created the following graphic to illustrate (you may click on it to enlarge). The "elbow" arrows are meant to convey that, in moving from the "long" to the "wide" format, the second line of couple data (in this case, the wife's) is raised to join the husband's data on the first line. In the longitudinal/panel example, each participant's Time 2 and Time 3 data are moved up to the first line, to join the Time 1 data.<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgscnRRQW0xeRgMuqOt6wFpJQ4aP_ouA4yLfIVqTVPKSHhgnZSBB4sqgLPaVmSWXV6gzBgFCRqUT1oYOOfmObDf3btBWApKzYA3poWujhS9Zr4Bz2PM9S2AbOQBOOqd2vvYBidRMui5sjo/s1600/data+restructuring.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="240" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgscnRRQW0xeRgMuqOt6wFpJQ4aP_ouA4yLfIVqTVPKSHhgnZSBB4sqgLPaVmSWXV6gzBgFCRqUT1oYOOfmObDf3btBWApKzYA3poWujhS9Zr4Bz2PM9S2AbOQBOOqd2vvYBidRMui5sjo/s400/data+restructuring.jpg" width="400" /></a></div>
<br />
Here's a <a href="http://www.youtube.com/watch?v=-GNRFeHAVy0">YouTube video</a> I discovered on SPSS's <i>restructuring</i> technique, which can be used to convert between the above two formats.<br />
<br />
<b style="color: #38761d;"><span style="color: blue;">UPDATE 2 (4/28/16):</span></b><b style="color: #38761d;"> </b>A potentially confusing situation can arise when one wants to compare the paths of men and women. Scenarios can exist for using multiple-group modeling or dyadic/APIM analysis. As I wrote on the board (shown below), multiple-group modeling is done when the groups (such as men and women) are independent of each other. Dyadic analysis, on the other hand, is done when there is a connection or interdependence between paired individuals, such as a man and woman in a heterosexual marriage (thanks to OR for taking the picture).<br />
<br />
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiuwbJvAVUC-Ojap9cM_qbzoc4-vk8WD73zRO9jORS8YCOEvBbFtEICKD7JBPY0R94scDG8ALP3hxW0YCagSSvYyDS6SqnnTrszV_gXE7Wb_SBmmnusHsAFepSnIVxRZy923BkF7SQ96lM/s1600/MGM+vs+APIM.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" height="214" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiuwbJvAVUC-Ojap9cM_qbzoc4-vk8WD73zRO9jORS8YCOEvBbFtEICKD7JBPY0R94scDG8ALP3hxW0YCagSSvYyDS6SqnnTrszV_gXE7Wb_SBmmnusHsAFepSnIVxRZy923BkF7SQ96lM/s320/MGM+vs+APIM.jpg" width="320" /></a></div>
<br />alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.comtag:blogger.com,1999:blog-7292073149242506593.post-56056393401240443792009-03-18T19:03:00.000-07:002013-01-15T18:52:04.932-08:00Presenting Model Fit StatisticsOne of our students this semester, Susan Murray, came up with a way to present model fit statistics that Kristina (the TA) and I both thought was very effective. With Susan's permission, here is the tabular format she came up with. The listed criteria for desirable values come from the Garson and Kenny documents in the links section to the right.<br />
<br />
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgh5PvePn-_esuLPmGzrhERV3cKqkjusAnkZGzPkC-iYvkfcHDg9UFVpZuy5mldWnXRfnqe174IeBFkS5DO21ifpTgUi5ji49nb3jjWoopjFTO8h1wkioVjcw9aaRX4foeB6IL7fOqgYPY/s1600-h/murray+table.jpg"><img alt="" border="0" id="BLOGGER_PHOTO_ID_5314715769262726162" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgh5PvePn-_esuLPmGzrhERV3cKqkjusAnkZGzPkC-iYvkfcHDg9UFVpZuy5mldWnXRfnqe174IeBFkS5DO21ifpTgUi5ji49nb3jjWoopjFTO8h1wkioVjcw9aaRX4foeB6IL7fOqgYPY/s400/murray+table.jpg" style="cursor: hand; cursor: pointer; display: block; height: 135px; margin: 0px auto 10px; text-align: center; width: 400px;" /></a>alanhttp://www.blogger.com/profile/08047057328265529252noreply@blogger.com