A-241 Five Point Test: Reliability, Validity, and Test Performance in Cognitively Intact Adults

Abstract

Abstract Objective Reliability and validity evidence related to the Five Point Test (FPT) scores is severely limited. The primary purpose of this study was to examine psychometric evidence related to two commonly used FPT scores (number of unique designs (UD), percentage of repetitions (PR)) using one-week test–retest reliability, correlations with demographic and neuropsychological variables, and convergent validity in line with a regression-based explanation-focused view of validity. Methods The sample consisted of 86 cognitively intact, non-depressed adult men and women ages 21–82 years (M = 52.7, SD = 17.7) with 7–21 years of education (M = 14.2, SD = 3.13) recruited from the general community and tested individually. Results UD ranged from 8–60 (M = 35.4) and PR ranged from 0–45% (M = 6.9%). Test–retest coefficients were .83 for UD but only .43 for PR. Age was significantly correlated with UD (r = −.59) and PR (r = .23). Education was significantly correlated with UD (r = .26) but not PR (r = −.10). There were no gender differences. UD showed significant bivariate correlations with WAIS-III Block Design, Trail-Making Test (TMT) A, TMT-B, Bicycle Drawing Test, and FAS Verbal Fluency but, together in a regression, only age and TMT-B remained significant. PR scores did not correlate significantly with any neuropsychological variables. Conclusion UD showed strong test–retest reliability. UD performance tends to be poorer with older age and less education. The meaning and interpretation of UD performance using a regression-based explanation-focused view of validity will be discussed. PR reliability is poor even over a short interval and attenuates subsequent statistical findings. Use of PR is not recommended in research or practice.