Dear all,
I am trying to do my data screening according to Curran (2016) and Meade & Craig (2012). However, I encounter some technical problems.
First, I am not sure how to calculate the Mahalanobis distance in SPSS. I found a tutorial video on YouTube (only 8 minutes, https://youtu.be/AXLAX6r5JgE), in which the tutor completes the calculation with the Linear Regression function in SPSS. Is it right?
Another question related to the Mahalanobis distance is about the cutoff value. Both the papers do not tell us a respondent with how large value of MD can be judged as a careless respondent. But they just mention that it can be judged by the probability value. If so, what is recommended level of significance coming out from the Chi-Square test of MD? 0.001, 0.01 or 0.05?
Second, both of the odd-even consistency and antonyms/synonyms consistency approaches are designed to create a with-person correlation coefficient. However, I am confused about how we can compute and save such a correlation coefficient for each person. Because all I know before is that correlation coefficients can be computed at between-person or sample level and cannot be saved as a column of variable in the dataset. Could you please tell me how I can make it in popular statistical software such as SPSS or Stata?
I would be grateful if anyone could solve my puzzles or share some reference materials with me.
Here is the paper I refer to:
Meade, A. W., & Craig, S. B. (2012). Identifying careless responses in survey data. Psychological Methods, 17(3), 437-455.
Curran, P. G. (2016). Methods for the detection of carelessly invalid responses in survey data. Journal of Experimental Social Psychology, 66, 4-19.
Best,
Chuding
Locked
