ASSESSMENT OF VERBAL MEMORY IN TEMPORAL-LOBE EPILEPSY USING THE SELECTIVE REMINDING TEST - EQUIVALENCE AND RELIABILITY OF ALTERNATE FORMS

Each of four forms of the verbal Selective Reminding Test (vSRT) were administered in counterbalanced order to 24 patients with epilepsy. Results of repeated-measures analysis of variance indicated that the forms were equivalent in difficulty. Furthermore, no practice effect was detected, in that performance on the first administration did not differ significantly from the second, third or fourth administrations, regardless of form. Although the four forms were found to be equivalent in difficulty and without significant practice effect, intercorrelation of forms was modest in magnitude. The measures derived from standard scoring of the vSRT (Total Recall, Long-Term Storage, Long-Term Retrieval, and Consistent Long-Term Retrieval) were highly intercorrelated. Factor analysis of these indices yielded a single factor solution for each of the four forms. We recommend using a single composite measure and obtaining multiple baseline estimates. This may provide a more reliable and stable estimate of verbal memory impairment, resulting in improved identification of patients with dominant temporal lobe epilepsy.