Comparison of paper-and-pencil vs. computerized administration of the Leistungsprufsystem (LPS)

The increasing use of computerized diagnostic procedures inevitably questions the equivalence of conventional paper-and-pencil versions and corresponding computer transpositions. To investigate the equivalence of the computerized version of the Leistungsprufsystem (LPS), as implemented in the Hogrefe Testsystem, and the conventional paper-and-pencil version, 131 participants, were tested applying a counter-balanced design with repeated measures. Heterogeneous results of the paper-and-pencil and the compterized version clearly show that one cannot proceed on the assumption that both versions are equivalent, and the results, emphasize the necessity of systematic I equivalence checks. An additionally performed investigation of retest reliability-of the computerized LPS version resulted in reliability coefficients ranging from r(n) = 0.55 to r(n) = 0.94 for a 2-week retest interval. Potential reasons for the non-equivalence of both LPS versions are discussed.