Millisecond Forums

Multiple IAT administrations for each subject

https://forums.millisecond.com/Topic11271.aspx

By briancjo - 9/18/2013

I did multiple administrations of the IAT per subject. 4 to be exact. I was trying to see if I could get more accurate results after multiple trials. I also used within-subject and between-subject counterbalancing. For example, Subject 1 did 4 trials; order 1, order 2, order 1, order 2. Subject 2 did 4 trials; order 2, order1, order2, order 1. to get around the automatic counterbalancing of Inquisit. I labeled subject 1-order 1, 101, subject 1-order2, 102, subject 2-order2, 202, subject2-order1, 201. This way I maintained an entire counterbalanced sample based on the order of the subjects' initial IAT as well as counterbalancing within to see if there was difference within or if i could get a mean or aggregate score that is more representative of each subject.


Anyway, I recently tried to use SPSS syntax for my IAT data. The problem i encountered is I am not sure what SPSS did with the multiple administrations because there each subject number, eg 101, which should have two scores, only has one score. I'm wondering if there is a way to change the syntax to reflect the multiple administrations, or like i have though, have to relabel my subjects to reflect each individual trial in some sane way.

By Dave - 9/18/2013

You could (a) compute the results for each IAT separately and later merge the summary files in SPSS or (b) introduce another variable to specify the different IATs and /break on that. As for how the SPSS syntax works, you could generate a set of sample data and step through the execution while at the same time studying the syntax. For general guidance on using SPSS, reading and writing SPSS syntax, I recommend you consult the SPSS manuals as well as other SPSS-specific support resources (books, courses, tutorials, etc.).