Research has shown that the effects of prior knowledge in studies using recall have been large, yet small or nonexistent in studies using recognition. Several memory researchers argue that there are at least two processing components involved in recognition: one is said to be influenced by the retrieval of specific, meaningful information about a studied item (i.e. recollection) and the other is influenced by an assessment of the global similarity between an item at test and information stored in memory (i.e. familiarity). So, the question is whether prior knowledge has different effects on the two memory processes.

Experiment 1 examines recognition performance among high- and low- knowledge readers about Star Trek.

Method: Participants – 96 undergraduates took a Star Trek knowledge test, consisting of two parts: a character test and a life-form test. The scores of the knowledge test ranked participants according to their performance, highest to lowest. Through the ranking, top to bottom, participants were placed into groups of four. Each member of the group was randomly assigned to a text set (2 different sets): one Star Trek story and one psychology chapter. After reading the texts, they were asked whether they had read either text before and then they received the recognition test. Recognition items were presented in two blocks: one block of Star Trek items (50 sentences consisting of 25 items from each story) and the other block of psychology items (50 sentences consisting of 25 items from each chapter). Target and distractor items within a block were presented randomly. Participants were asked (for both stories) to decide whether or not each item had appeared in the story. If the item was not in the story, they were asked to press a key labeled no. If the item was in the story, they were asked to decide whether they had a conscious awareness of having read the sentence; if they did, then they pressed a key labeled R (for remember). If they did not, but they still believed that the sentence had been in the story, they pressed a key labeled K (for know).

Results: The prediction that experts would provide more remember responses than would novices, only in response to Star Trek items, was revealed. The two groups did not differ in response to psychology items though. The analysis of know judgments to old items mirrored the remember responses; experts made fewer know responses than did novices, but only for Star Trek items. The overall results are consistent with previous research that knowledge effects do not impact overall recognition accuracy. In the analyses of the familiarity estimates it is appears that knowledge has a large effect on recollection but little effect on familiarity.

Experiment 2 examines the effect of knowledge on recollection and familiarity.

Method: Participants – 72 undergraduates. The same texts and recognition items were used from Experiment 1 (25 Star Trek items, 25 Psychology items). A third Star Trek story and a third psychology chapter were added. The texts were paired to create six text sets. After completing the Star Trek knowledge test, participants were randomly assigned to a text set: either two Star Trek stories or two psychology chapters. There were four study-test conditions: in two conditions the participants read the text and took the recognition test in counterbalanced instructional blocks. First block – respond "yes" if the sentence appeared in the first text they read or if they recognized the sentence but could not remember in which text it appeared. They were told to respond "no" to any new item. Second block – respond "yes" only if the sentence was in the second text. They were told to respond "no" if they recognized the item from the first text or if the item was new. In the other two conditions, the instructions were reversed; respond "yes" if the sentence appeared in the second text.

Results: Experts had higher estimates of recollection than did novices, but only in the Star Trek condition. The two groups had similar recollection estimates in the psychology condition. Domain knowledge appears to have a reliable and robust effect on recollection. Experts were more accurate when asked to recognize items that had been in either text (inclusion instructions), and when linking an item with the specific text in which it appeared (exclusion instructions).

General Discussion

Recollection plays a significant role in the recognition performance of high-knowledge individuals. In Experiment 1 Star Trek experts had more remember responses, and in Experiment 2 experts had a larger estimate of recollection than did novices. The findings are only in response to Star Trek items, meaning that expertise has a significant influence on the recollective component of recognition. Robust knowledge effects emerge when recognition is separated into familiarity and recollective components. The involvement of two processing components in recognition explains why it is difficult to find knowledge differences in recognition memory.