How is the research question operationalized? What is the abstract constructs being

studied, the concrete way these are being observed or measured including the IV and DV.

Research article

METHOD Participants

A total of 288 Stony Brook undergraduates volunteered for partial fulfilment of a course requirement. Of these, 192 completed a group norming study and 96 participated individually in the experiment.

Design

Information accuracy (correct, misleading), and story status (read, not read) were manipulated within participants. Delay (immediate, 48hours) and repeated testing (delayed test: single, repeated) were manipulated between participants. Half of the participants took the immediate test, and all returned for the second session. To pinpoint the role of delay, performance on the immediate test was compared to the delayed test performance of participants who had not previously been tested. To pinpoint the role of repeated testing, performance on the single delayed test was again examined, but instead compared to performance observed on the repeated delayed test. The dependent measures were the proportion of general knowledge questions answered correctly versus with misinformation, and the classification of answers as Remember, Familiar, or Just Know.

Materials

In the norming study participants answered 36 general knowledge questions related to facts from two stories from Marsh (2004; "Summer Star Search" and "The Inventor"). There was no penalty for guessing and participants were encouraged to answer even if unsure. In all, 32 questions were selected to span a range of difficulty (the hardest question was correctly answered by 0.5% of students while the easiest was correctly answered by 81.2% of students).

We matched the difficulty of the questions across the two stories. Each story contained two pairs of facts also equated for difficulty: one pair was presented in correct format (e.g., "The Pacific is the largest ocean") and the other was incorrect or misleading (e.g., "The Atlantic is the largest ocean"). Across participants, both stories were read equally often and each fact appeared equally often in correct and misleading formats. Finally, we adapted the story "The Art Thief" (Marsh, 2004) for use as a distractor story. All critical facts were removed from this story to prevent unintended interference.

The general knowledge test contained 48 questions, 32 of which corresponded to the critical studied and non-studied facts; 16 questions were fillers and also spanned a wide range of difficulty (Nelson & Narens, 1980; mean difficulty=46.1, SD=27.49). Each question was in cued-recall format with a line for recording one's answer and three small boxes labelled "R", "F", and "JK" for Remember, Familiar, and Just Know.

Procedure

The experiment had six phases: story reading, distractor activity, cued-recall test #1, delay, cued-recall test #2, and final questionnaires.

In Phase 1, story reading, participants were allowed up to 5minutes (with a time warning at 2.5minutes) to read one of the critical stories. They were instructed to read the story only once and to cover it when finished. Afterwards, all participants answered a brief reading comprehension questionnaire, which began with "Did you finish reading the story?" All participants answered affirmatively.

In Phase 2, distractor activity, participants read the filler story. This was included to increase the plausibility of the cover story (that the experiment was about reading comprehension). Participants again received 5minutes for reading, and answered comprehension questions afterwards.

Phase 3, cued-recall test #1, was only completed by the participants in the Repeated Test condition. Participants were instructed to answer the questions in sequential order and to only answer each question if they were reasonably sure of the answer (i.e., not to guess). Before taking this test, participants were instructed about the Remember, Familiar, and Just Know classifications. These instructions were adapted and combined from Rajaram (1993, 1996) and Conway et al. (1997). Remember responses were described as those for which the participant had a vivid or conscious recollection of an answer being in the story (e.g., could remember which character said it, where it had occurred in the story, what the participant was thinking while reading that piece, etc.). Familiar responses were described as responses for which the participant knew the answer had been in the story, but did not have a vivid or conscious recollection of it being in the story. Thus, Familiar responses here parallel the "Know" instructions in most previous studies. Finally, Just Know responses were described as answers that were "just known" and could be answered based on one's general world knowledge (i.e., answers known prior to the beginning of the experiment).1

It was made clear to participants that while both the Remember and Just Know boxes could be checked for a given answer, or both the Familiar and Just Know boxes, it was impossible to check both the Remember and Familiar boxes. It was also emphasised that although more than one box could be checked, it was equally acceptable to check only one. As each participant was tested alone, the experimenter ensured that all were able to correctly explain (in their own words) the distinctions between the three responses before beginning the test. Participants received as long as they needed to complete this phase, and took approximately 20minutes.

Phase 4, delay, involved a 48-hour delay. Participants received no instructions regarding the second day of the experiment beyond the fact that they would be fully debriefed at the end of the study.

In Phase 5, cued-recall test #2, each participant took the same cued-recall test as was used in Phase #3 (including the Remember, Familiar, and Just Know instructions). Thus this was a repeated test for half of the participants (delayed test: repeated) and the first test for the other half (delayed test: single).

Finally, in Phase 6 participants were probed about their beliefs about the experiment's purpose. They were explicitly asked whether they noticed errors within the story and whether or not they had looked up any of the facts during the 48-hour delay. Finally, all participants received the correct facts as part of the debriefing. Each was thanked, and asked not to discuss this experiment with others.