Establishing equivalence of different list conditions.

Experimental designs frequently involve different lists across conditions, resulting in a treatment-list confounding. If treatment effects are obtained, it is then impossible to assess the contribution of list characteristics to this effect. A simple technique was presented whereby experimenters who must use different lists for different experimental treatments can postexperimentally establish equivalence of list difficulty. The technique involves inclusion of one or more constant words across lists. Data from a previously published experiment (Glidden, 1977) were analyzed according to this technique. Results indicated that the effect obtained was not an artifact of differential list difficulty.

PMID: 525661 [PubMed - indexed for MEDLINE]