Summary presented by Jamie Huffman

Landau, J., Marsh, R. (1997) Monitoring source in an unconscious plagiarism paradigm. Psychonomic Bulletin & Review, 4, 265-270.

Summary presented by Jamie Huffman

(Style somewhat consciously plagiarized from Megan)

Previous research:

Two paradigms concerning unconscious plagiarism:

Brown and Murphy's (1989) paradigm used three tasks : initial generation, recall-own task, and generate-new task. The initial generation task required participants in a group setting to one-at-a-time generate members of a category. For example, name things that are red. In the recall-own task, the participants were to recall the answers that they had given. In the generate-new task, participants were to give new answers that had not been previously given by themselves or other participant.

Marsh's paradigm takes Brown and Murphy's tasks and applies them to the game Boggle (4X4 matrix with 16 letters).

Source monitoring:

Simply, source monitoring is the ability to recall where particular information was received from. The authors contend that the recall-own task is essentially a source monitoring decision because it requires one to differentiate between old items and choose a source. Thus, if this is true, the authors argue that the recall-own task will be affected more by source manipulations than either initial generation or generate-new task.

Experiment One:

Goal: To increase the likelihood of plagiarism by making the tasks more confusable.

Participants were 42 undergraduates receiving course credit.

Method:

Between subjects design.

Conditions and Procedure:

Intact- read: participants were shown the computer's answer intact.

Generate-reveal: participants were shown letter by letter the answer being generated by the computer. After each letter was shown, the participants were asked if they wanted to guess the word. The task in this condition,, according to the authors, is very similar to generating your own answers. Thus, plagiarism should increase.

Basic Procedure:

Initial Generation -> Recall Test

Participants given 4 puzzles. Participants were shown each puzzle and asked to recall 4 answers that they had given(recall own) and 2 that had not been given (generate new).
For each puzzle, the computer first gives solutions, the participant then gives 1 solution. This continues until the computer has given 12 and the participant has given 4.

Results:

Generate-reveal did indeed increase the recall-own plagiarisms, as hypothesized.
In fact, those in either condition were more likely to plagiarize than give a new answer.

Experiment 2

Goal: To decrease the likelihood of plagiarism by making the tasks more distinguishable. The tasks were made more distinguishable by creating two conditions. All participants completed the same procedure as the read intact condition of Experiment 1, except half of the participants did so with the computer and the other half did so with a human partner. The human partner condition was hypothesized to decrease plagiarism because of an increase in perceptual and contextual details.

Participant were 40 undergraduate students receiving course credit.

Method and Procedure:

Design: Between Subjects

Materials and Procedure is same as read-intact condition of Experiment 1.

Conditions:

1. Computer partner- Participants were shown the computer's answers to the puzzle intact.

2. Human Partner- Participants were given a paper version of the puzzle and human partner functioned as the computer.

Results:

Human partner condition decreased recall-own plagiarisms.
There were no significant differences in the recall-own errors.
There were no significant differences in the generate-new plagiarisms.

University of Arkansas

Department of Psychology

Lampinen Lab

False Memory Reading Group

False Memory Reading Group Spring 1999