Stem-completion tests are commonly used to assess memory. In Brazil, however, there are no publications that propose adequate stimuli for use in this type of task. Our objective was to select adequate word-stems to maximize memory assessment using this task in Brazil. We obtained chance-completion profiles of 137 word-stems that completed at least 10 common words in Portuguese from Brazil and manipulated factors that could influence memory estimates using these stems. We found that depending on the chance-completion of the most frequently used word to complete a stem it is not possible to show memory effects; that is, completion with familiar stimuli (seen before) is not distinguishable from baseline. Guidelines are proposed for the selection of adequate stems and stimuli to maximize the acquisition of memory estimates using indirect stem-completion tasks.
memory; stem-completion; direct tests; indirect tests
