|Subject:||Re: Factor analysis|
|Date:||Wed, 6 Jan 2016 16:49:52 -0600|
|User-agent:||Mozilla/5.0 (Windows NT 6.1; WOW64; rv:38.0) Gecko/20100101 Thunderbird/38.5.0|
The regression procedure does contain an option to save the scores, but it's not that hard to just create a compute statement:
compute yhat = -2.4567 + X1 * 0.2348 + X2 * (-2.4500).
I don't deny that factor scores can be useful and my understanding is that they are more complex to compute but I have not ever used factor scores because I almost always want to know what a real score is, not a factor score. That is, if I have a measure and find that items 1, 2, 3, all measure factor 1, I DO NOT want the F1 score (usually), I want the score one those three items, which is also trivial to create:
* assume items 1 and 3 are revrse-coded. All variables are on a 5-point likert.
compute V1R = 6 - V1.
compute V3R = 6 - V3.
compute F1 = mean(V1R, V2, V3R)*3.
So, your statement that EFA is worthless unless you can save the factor scores is incorrect.
On 1/6/2016 4:35 PM, news wrote:
Saving factor scores would be a very welcomed innovation. What is the sense of factor analysis if you can't reuse the scores ?
-- Alan D. Mead, Ph.D. President, Talent Algorithms Inc. science + technology = better workers +815.588.3846 (Office) +267.334.4143 (Mobile) http://www.alanmead.org Announcing the Journal of Computerized Adaptive Testing (JCAT), a peer-reviewed electronic journal designed to advance the science and practice of computerized adaptive testing: http://www.iacat.org/jcat
|[Prev in Thread]||Current Thread||[Next in Thread]|