Hi I need to get the similarity between groups of married couples in term of big five personality (extraversion, agreeableness, conscientiousness, neuroticism and openness) so that i can run pearson correlation of that similarity with the overall result i get from dyadic adjustment scale (DAS). I have computed the sum of each of the big 5 dimensions for both side of each couple pair n the total sum of each pair for the DAS. But i am at lost on how to measure the similarity of each of the personality similarity. Please help!
Perhaps these will help:
http://cs.tsu.edu/ghemri/CS497/ClassNotes/ML/Similarity%20Measures.pdf
https://www.umass.edu/landeco/teaching/multivariate/readings/McCune.and.Grace.2002.chapter6.pdf
If your hypothesis is couples who are X are more similar than couples who are Y, you could calculate the correlation within group X and separately for group Y. Then do a Fischer Z-score comparison to test if there’s a stronger correlation among X than Y. It’s more challenging if your trying to create a single similarity score for each couple and use it as one of many independent variables. Maybe you could find the geometric mean for each couple (i.e., square root of the sum of the squares of the differences between each couple members scores on each of the 5 dimensions). Scores closer to zero identify more similar couples and all non-zero values are positive. Especially if the standard deviations of the 5 dimensions differ, you might first z-score each dimension (so each of the 5 dimensions effectively gets equal weight). Hope this helps some, Jenny! ~ Kevin
- Badges
- Science topic