Interference in episodic memory: retrieval-induced forgetting of unknown words

The items were categorized according to three different principles manipulated between participants. In a first condition, the categories were the first letter of the items. In a second condition, the items were linked to meaningful labels and did not correspond in their first letters per category. In a third condition, different meaningful labels were used that started with the same letter as all items linked to the label. All labels were categories that participants probably only had very limited knowledge about (see Appendix). In the initial letters condition, instructions stated that so-called non-words, letter strings that were not words, were going to be presented. The initial letter was presented additionally to the left of the non-word. In the two conditions with meaningful labels, instructions stated that word pairs were going to be presented in which the left word was a category name and the right word an exemplar of that category. It was emphasized that only exemplars had been chosen that the participants probably would not know. Retrieval-practice success did not differ between categorizing conditions, F < 1. With regard to the test phase, categorizing condition did not moderate RIF or Rp+ enhancement, all Fs < 1.68, all ps > 0.194.