Item validiation (IAT)

Back

Item validiation (IAT)

Author	Message
nbUOS	nbUOS posted 8 Years Ago Topic Details Share Topic
Group: Forum Members Posts: 11, Visits: 24	hello there, i am sorry f this might be the wrong forum, let me know if another one might be better suited (and which one ;-)) but here i go: i am trying to validate my IAT items (if all pictures i used/presented generate about the same respnse time and or bias) at first i thought some kind of reliability analysis might do the trick but unfortunately: not the solution i have since been trying this with a simple latency-average comparison but i guess i need the congruent/incongruent factor of the item presentation and response of the participant with it to get useful information. to do this i tried "editing" the IAT syntax provided by greenwald & collegues but that doesnt quite work can anyone help me figure out how - or where - i can get an answer to this? Tags IAT analysis items validation SPSS
	Reply Like 59
Dave	Dave posted 8 Years Ago ANSWER Post Details Share Post
Group: Administrators Posts: 12K, Visits: 98K	It sounds to me like you are looking to do a "by-item" as opposed to a "by-subject" analysis. To do this, you will have to aggregate your data in a completely different way, i.e., the existing (Greenwald) SPSS syntax will not help you much. See e.g. http://www.talkstats.com/showthread.php/20913-by-subjects-versus-by-items?s=df455cddce3351535f919c3fb33ec00d&p=63578&viewfull=1#post63578 and https://psychologicalstatistics.blogspot.com/2012/06/stimuli-as-fixed-effect-fallacy.html for a more general introduction. I'm not sure whether that will tell you what you want to know (I'm not clear on what exactly you are looking for), because it will only indirectly touch on questions of reliability and/or validity. The question a "by-item" analysis is supposed to answer is whether results generalize beyond the particular stimuli / exemplars used, i.e., to the larger population of all possible stimuli. If you're only concerned with whether your stimuli are (approximately) equally easy to categorize, I would turn to the data for the attribute and target practice blocks (not the combined blocks). I should note that -- ideally -- you would have validated your stimuli in a separate study prior to the IAT. See Lane et al. (2007; page 86 ff.; http://faculty.washington.edu/agg/pdf/Lane%20et%20al.UUIAT4.2007.pdf ) regarding the importance of stimulus choice.
	Reply Like 57