Open ben-domingue opened 3 months ago
This is a little bit hard dataset for me. So, 1) Do we need to exclude demographic variables if they are in likert scale? (e.g., "To what extent do you identify as Arab-Muslim" and then later asks the actual item "On average, how much do you like Arab Muslims?"). 2) It has session ID (those are unique) and then it has participant ID (those are not unique as participants can participate multiple times). Do we need to keep both of them?
i'd be inclined to just take the first observation of each uniqueid. so just duplicate_id!=1 thoughts about that? we can then make the uniqueid the 'id'
Results IAT_study2.xlsx IAT_study3.xlsx
we ended up with large volumes of NAs here:
id item resp
1 11492565 meatlikemeat NA
2 11513951 meatlikemeat NA
3 11548646 meatlikemeat NA
4 11541671 meatlikemeat NA
5 11507119 meatlikemeat NA
6 11458988 meatlikemeat 5
7 11521640 meatlikemeat NA
8 11570527 meatlikemeat NA
9 11448415 meatlikemeat NA
10 11501922 meatlikemeat NA
11 11520335 meatlikemeat NA
not a huge deal as i'll just cut them out but i wanted to check first. and what is the difference between study2 and study3?
Yes. If you look at this supplement material page 22 is for the study-2 and page 25 is for study-3. (In the article also it says study-2 compares only best and worst performing words, and study-3 tests variability and relevance of attribute words to the attribute categories)
Files in csv with many NAs
https://osf.io/ezj5t/?view_only=1781c05b04d54b829fd2eff67e0d429c paper. https://link.springer.com/article/10.3758/s13428-021-01592-8#author-information