Determining Cutoffs for the Psychometric Synonym Analysis to Detect IER
Tyler Barnes · OhioLink ETD Center (Ohio Library and Information Network) · 2018
Determining Cutoffs for the Psychometric Synonym Analysis to Detect IER.The validity of individual responses is required for valid inferences drawn from data.Insufficient Effort Responding (IER; Huang, Curran, Keeney, Poposki, & DeShon, 2012) is one possible threat to individual response validity.There are many methods to detect IER, but the Psychometric Synonyms Index, despite its practical utility, is understudied.The purpose of this study is to provide recommendations for its use that are empirically grounded.Using a simulation, I found that the strength of the within-pair correlations used for inclusion into the index, the number of pairs, the type of random responding, the correlation between the pairs, the skewness of the data, and IER severity with an individual case have an impact on the psychometric index and by extension the cut-off one should use for classifying cases as IER or careful.Recommendations for the index depend on the situation.vii 10.Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .1 and dependence of 0.5.......…………………………………………………....57 11.Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .2 and dependence of 0.5.......…………………………………………………....59 12. Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .3 and dependence of 0.5........…………………………………………………...61 13.Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .4 and dependence of 0.5........…………………………………………………...63 14.Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .1 and dependence of 0.75........………………………………………………….65 15.Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .2 and dependence of 0.75........………………………………………………….67 16.Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .3 and dependence of 0.75........………………………………………………….69 17.Mean power, error, and mean Psychometric Synonym index in normally distributed data with an empirical cutoff of .4 and dependence of 0.75........………………………………………………….71 18. Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .1 and dependence of 0……........…………………………………………………..73 19.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .2 and dependence of 0........…………………………………………………..........75 viii 20.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .3 and dependence of 0..........…………………………………………………..........77 21.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .4 and dependence of 0. ..........…………………………………………………........79 22. Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .1 and dependence of 0.25...........…………………………………………………...81 23.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .2 and dependence of 0.25...........…………………………………………………...83 24.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .3 and dependence of 0.25...........…………………………………………………....85 25.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .4 and dependence of 0.25...........…………………………………………………....87 26.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .1 and dependence of 0.5...........…………………………………………………......89 27.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .2 and dependence of 0.5...........…………………………………………………......91 28.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .3 and dependence of 0.5...........…………………………………………………......93 29.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .4 and dependence of 0.5...........…………………………………………………......95 ix 30.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .1 and dependence of 0.75. ...........…………………………………………………......97 31.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .2 and dependence of 0.75. ...........…………………………………………………......99 32.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .3 and dependence of 0.75. ...........…………………………………………………....101 33.Mean power, error, and mean Psychometric Synonym index in negatively skewed data with an empirical cutoff of .4 and dependence of 0.75. ...........…………………………………………………....103 34.Supplemental analysis with psychometric synonyms with within-pair correlation strength of 0.7, 32 pairs, and normally distributed random responding..…………………..……..……….………10535.Summary of recommended cut-off indices for the use of Psychological Synonyms in normally distributed data…………….……….………10636.Summary of maximum recommended cut-off indices for the use of Psychological Synonyms in negatively skewed data……………………………...…………………..……..……….………107