Timeline for When is it defendable to use manifest instead of latent variables in psychological research?
Current License: CC BY-SA 3.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 20, 2022 at 12:48 | answer | added | Ben porter | timeline score: 0 | |
| Jul 23, 2013 at 12:42 | comment | added | Gala | let us continue this discussion in chat | |
| Jul 23, 2013 at 12:40 | comment | added | Jens Kouros | This is getting a bit off-topic, I think. My question was a sincere one, I wanted to get answers from people on this site on what they think might be psychological questions that can/should be investigated using manifest variables. However, I did not want to put anyone down, or downplay experimental psychology, which seems to me is what you are accusimg me of. I am a big believer in empirical science. I don't think doing experiments or learning about techniques is easy, either. And in fact, I agree with most of your arguments (e.g. sample size) and get the feeling that we have similar views. | |
| Jul 23, 2013 at 12:25 | comment | added | Gala | How many experimental psychologists do you know? Are you reading experimental psychology journals? You make it sound as if it was not much but knowing/learning about the techniques at all is already a big effort. Dealing with the fact that you are interested in participants' responses to different stimuli (as opposed to traits) is not easy either and makes much of the literature/introductory material on latent variables basically irrelevant. I would also think that getting reasonable results would require larger samples than typical psychology experiments (many have N=10 to 40). | |
| Jul 23, 2013 at 12:09 | comment | added | Jens Kouros | @GaëlLaurans: I don't think that from an experimental psychology perspective, there is that much additional effort required in taking a latent variable approach. You might need an additional indicator, but I think that's pretty much it. The difference is in how you analyze the data. | |
| Jul 23, 2013 at 10:24 | comment | added | Jens Kouros | @GaëlLaurans: It is true that I did approach this question from a psychometrics perspective. But my question at the end was meant in exactly the way that you suggest: What are the cases where a latent variable approach is not that useful? I realize that this means that I assume in most cases it is useful. (thanks for the hint with the paper, I will look for it.) | |
| Jul 23, 2013 at 9:45 | answer | added | Gala | timeline score: 8 | |
| Jul 23, 2013 at 9:32 | comment | added | Gala | Btw, the way you frame your question seems to suggest you are looking at this from a psychometrics/personality psychology perspective. Looking at it from an experimental psychology perspective, you could just as well ask: Is all the technical baggage, data collection, etc. required for latent variable really worth it? | |
| Jul 23, 2013 at 9:28 | comment | added | Gala | You might be interested in Lee Cronbach classic paper “The Two Disciplines of Scientific Psychology”. I believe Dennis Borsboom also wrote something about that. | |
| Jul 23, 2013 at 5:15 | history | edited | Jeromy Anglim | CC BY-SA 3.0 |
added 3 characters in body; edited title
|
| Jul 23, 2013 at 4:33 | answer | added | Jeromy Anglim | timeline score: 6 | |
| Jul 23, 2013 at 2:35 | history | tweeted | twitter.com/#!/StackCogSci/status/359502014532689922 | ||
| Jul 22, 2013 at 12:43 | history | asked | Jens Kouros | CC BY-SA 3.0 |