# Converting Big 5 questionnaire data to Big 5 scores using R

I'll preface this with I have no knowledge of psychology, but my goal is to produce some graphics dealing with correlations, and one example that I had thought of is how the Big 5 personality traits correlate with each other. I came across the R dataset qgraph" which has the results of a 240 item questionnaire (48 questions per Big 5 trait), but this only gives me the responses to questions. I want to be able to aggregate the questionnaire data in such a way that I can accurately assign Big 5 trait percentile scores to each respondent; however, it is not at all clear to me how this should be done. What is the process of taking questionnaire data and getting percentile values for each of the Big 5 traits for the respondents? Alternatively, if there's a dataset that already include percentile scores for the respondents, I'd be interested in that.

• Are you sure this question, which seems to be pretty agnostic to the dataset, wouldn't be better suited for stats.stackexchange.com? Dec 6 '17 at 14:44
• @Seanny123 you could be right. I thought it would be more appropriate here, because the dataset sued is largely irrelevant (I just included it for context). My question is essentially, what is the process of getting Big5 percentile values from questionnaire data? I imagine that I would get better answers about Big5 questions here than at stats.stackexchange.com Dec 6 '17 at 17:43
• Editing the question to emphasize this focus would be much appreciated. Dec 6 '17 at 19:07
• @Seanny123 done. Dec 6 '17 at 22:34

Here are some resources that will show you how to score the Big 5 using R.

The main function that I like is called scoreItems in the psych package. It handles item reversal, imputation, and multiple scales. It will also give you things like alphas and item total correlations.

William Revelle has this tutorial: http://personality-project.org/r/psych/HowTo/scoring.pdf

I have this video ( which has an associated github repository): https://www.youtube.com/watch?v=iFFW5sK3Bhk

See in particular, this bit of the code: https://github.com/jeromyanglim/score-tests-with-r-exercise/blob/master/finished-analysis/munge/01-munge.R

sc <- scoreItems(meta.bfi[,v$scales], ccases[,meta.bfi$name])
ccases[,colnames(sc$scores)] <- sc$scores