Find in Library

Neuropsychological tests are increasingly used as part of clinical assessment in multiple sclerosis (MS) and temporal lobe epilepsy (TLE), the most prevalent chronic disabling neurological conditions in adulthood. Heterogeneity of clinical samples may be one important reason for null hypothesis statistical testing (NHST) being frequently nonsignificant for group comparisons, but limitations in routine statistical methods may also be implicated. The present study examined the hypothesis that NHST may be an inadequate analysis strategy when comparing the performance of neurological patients with that of controls. Twenty-six TLE patients, 113 MS patients, and 117 normal controls of similar sociodemographic characteristics were compared on 11 instruments (35 scores). NHST resulted nonsignificant for 18 scores comparisons according to Bonferroni criteria, while effect size (d) and power estimates were adequate for 27 comparisons. Effect size estimations should be considered as an important analytical tool in selecting appropriate neuropsychological measures for clinical use.