Table 2 shows average scores on the EPQ-R subscales separately for smokers and non-smokers.
An analysis of covariance (ANCOVA) on the average log k values with the extraversion, neuroticism and psychoticism subscales of the EPQ-R as covariates found that the effect of smoking status was significant, F(1,821) = 17.
Alternatively, a different combination of items from the EPQ-R might be more strongly associated with delay discounting.
These seven items were summed (with #35 reversed) to make a "DD subscale" that represented the EPQ-R items that were most strongly predictive of delay discounting.
Interestingly, the effect size was larger than for other EPQ-R subscales.
There are indications that we succeeded in this effort, since the alphas of the CPS scales are high and the EPQ-R and the CPS scales appeared to be sufficiently related.
01 Table 3 Means, Standard Deviations, Reliabilities and Correlations of the P, E, and N Scales of the EPQ-R and the CPS College students (N=206) M SD [alpha] R EPQ-P 10.
A comparison of the parolee sample with the norms for EPQ-R (N = 508) and the BASIS-A (N = 1,083) was conducted using multiple t- tests with a Bonferroni correction to control for Type I Errors.
A comparison of the parolee group with the EPQ-R norms yielded significant differences (p < .
01) between the high antisocial group and the EPQ-R norms on the P (t = 11.
The high antisocial group was significantly higher than the normative sample for the EPQ-R on the P and N Scales.
The EPQ-R indicates deviant P trait levels in the sample as a whole.