Appendix A: Standardization, Reliability, and Validity Tables

Table A.34. Frequencies and Descriptive Statistics of Inconsistency Index Scores in EQ-i 2.0 Normative and Random Samples

The following table displays the proportion of participants in the EQ-i 2.0 normative sample and the sample of computer-generated random EQ-i 2.0 responses at each score on the EQ-i 2.0 Inconsistency Index (IncX). The mean and standard deviation of the IncX for each sample is also provided. d provides an effect size for evaluating the difference between the IncX scores of the two groups as small, medium, or large. The validity of the IncX scale is supported by the meaningfully higher scores (i.e., d ≥ .20) found in the random sample relative to the normative sample. The validity of the choice of IncX scale cutoff (i.e., 3) for identifying participants as potentially invalid would also be supported if it classifies a high proportion of random respondents as potentially invalid, as found in the EQ-i 2.0.

Inconsistency Index Score

Normative Sample Random Sample
N % N %
10 0 0.0 1 0.0
9 or higher 0 0.0 28 0.7
8 or higher 1 0.0 152 3.8
7 or higher 3 0.1 543 13.6
6 or higher 5 0.1 1,293 32.3
5 or higher 23 0.6 2,287 57.2
4 or higher 60 1.5 3,172 79.3
3 or higher 140 3.5 3,733 93.3
2 or higher 455 11.4 3,939 98.5
1 or higher 1,449 36.2 3,994 99.9
0 or higher 4,000 100.0 4,000 100.0
M (SD) 0.5 (0.9) 4.8 (1.6)
Cohen’s d 3.36

Note. An IncX score of 3 is used to indicate a potentially inconsistent response style. Positive Cohen's d values represent higher means in the random sample. Guidelines for evaluating |d| are .20 = small, .50 = medium, .80 = large.