Skip to main content
. Author manuscript; available in PMC: 2015 Jul 1.
Published in final edited form as: Drug Alcohol Depend. 2014 Apr 13;140:213–216. doi: 10.1016/j.drugalcdep.2014.03.034

TABLE 2.

Chi-square (χ2) test of independence, Cohen's Kappa (κ), symmetric Somers' d, and Cramer's V measures of association for DSM-5 vs. DSM-IV crosstabulations for nicotine, alcohol, and cannabis use (N=525)

χ 2 κ Symmetric Somers' d Cramer's V
Nicotine 248.8 ---- 0.96 0.90
Alcohol 344.1 0.81 0.99 0.70
Cannabis 925.1 0.84 0.98 0.77

Notes: Degrees of freedom (df) for likelihood ratio χ2 test is 6 for nicotine and 9 for alcohol and cannabis; all p values < 0.001. Cohen's κ is a measure of nominal scale agreement between two categorizations of responses or ratings; it is restricted to square tables where row and column values represent the same scaling. Thus, there is no κ value for nicotine. Symmetric Somers' d is an ordinal measure of association between categorizations and indicates the proportional reduction in error from predicting the ranking of respondents in one categorization from the ranking in the other categorization. Cramer's V is a measure of association for nominal variables for use with tables greater than 2 × 2 in size.