Overview

The present study explores the influences of social conformity on source attribution. It cites past research, such as Asch (1951), in which social influences in the realm of conformity was supported.

To apply these ideas to source memory, subjects in the present experiments saw pictures of some objects and imagined other objects. They then took a source monitoring test in which they had to indicate if objects had been seen, imagined, or neither. On the test they also were given information about the response of a high credibility source or a low credibility source. The hypothesis for these experiments was that the response of the source would directly affect the participant’s source attribution.

In addition, another goal of the present experiment was to determine if confederate credibility affected the level of conformity. Specifically, that low credibility source would produce less conformity (experiments 1A and 1B).

Experiment two was altered in the amount of time between presentation of pictures and the subsequent recall test (20 min. interval vs. 24 hour interval). The idea here was that subjects should be more influenced by the source if their own memory for the event is sparse.

Method

Participants were shown color photos of objects on a slide projector with an audiotape playing at the same time saying the name of the object shown. Some slides were blank, but the audiotape still said a word of an object. During these blank slides, participants were asked to imagine a picture of the object in their head. A cover task (participant’s estimated length of time an artist would take to draw the object) was given during the presentation. Then a 20-min. filler task (completing word puzzles) was implemented.

Participants were tested using a source-monitoring test (perceived, imagined, or new) after a 48-hour delay (20-min. vs. 24 hours interval for experiment 2). Before responding, participants were allowed to view the response of a confederate (high credibility vs. low credibility in experiment 1A & 1B) or lack thereof (base-line condition).

After the participant responded, the participant was then asked to rate, in percent, their level of confidence in their response- 0% (completely guessing), 20%, 40%, 60%, 80%, 100% (completely sure).

¨ ª © § It is interesting to note that the original presentation of objects is visual with auditory cues, while the recall test was words (instead of pictures) of the visual objects presented or not presented.

Experiment 1A

~ 72 objects (chosen from category exemplars described by Battig and Montague, 1969)

~ 8 categories: fruit, birds, land animals, tools, kitchen items, vehicles, musical instruments, and sea creatures.

~ 9 exemplars from each of the 8 categories

~ 12 critical percieved items and 12 critical imagined items presented to subjects in the study phase

~ 48 filler trials, 24 perceived items and 24 imagined items, presented after the critical items to decrease retrieval accuracy. The same filler items were shown in the same order to all the subjects.

¨ ª © § Primacy effects??

I.V. = A. Instructions (high credibility vs. low credibility of the confederate)

(High credibility=graduate student from University of Washington)

(Low credibility=computer picking responses in a random manner)

*Between subjects factor

B. Trial type (congruent, incongruent, or baseline)

*Repeated measure/6 within-subjects experimental condition

Conditions for the new items:

1. Confederate responds perceived, test item is new (incongruent)
2. Confederate responds imagined, test item is new (incongruent)
3. Confederate responds new, test item is new (congruent)
4. Confederate does not respond, test item is new (baseline)

1. Confederate responds perceived, test item was perceived at study
2. Confederate responds imagined, test item was imagined at study

D.V. measure = A. Reality monitoring test (total of 72 test items)

~ 24 critical items studied earlier

~ 48 new test items not encountered before in the experiment.

~ Presentation of test items randomized with restriction that no more than 3 items of the same class (e.g. imagined items) could be presented in a row.

~ 72 exemplars divided into 6 subsets of 12 items each. Mean ranking for prototypicality (Battig & Montague, 1969) was equated for each sub-set.

~ 6 subsets listed above under Trail type.

~ Each item appeared in each of the subsets approximately equally as often.

B. Confidence Rating (circle after each response of type of test item)

~ 0% (completely guessing), 20%, 40%, 60%, 80%, 100% (completely sure).

Results

Accuracy of responses:

~ ­ Credibility helped confidence (marginally significant)

~ Item type helped confidence (congruent)

Purpose:

~ The correct response for baseline items equally likely to be OLD or NEW.

Change from 1A:

~ 12 additional study objects

~ 7^th subset added-confederate does not respond, test item is old (baseline)

Results

Accuracy of responses:

Same except ­ credibility did NOT boast confidence rating

Experiment 2

Purpose:

~ Only high confederate credibility used

~ Interval of testing time (20-min. vs. 24 hours)

Results

~ Accuracy decreased over time in baseline condition

~ No signigicant difference in accuracy on congruent trials vs. baseline

~ Significant difference b/w lower performance in incongruent trials vs. baseline

~No interaction b/w delay and confederate’s response found!

~ Short delay group more confident in their responses (not necessarily correct responses)

****Showed no evidence that social influence was stronger for the 24-h delay group

Which goes against social literature and predictions based on misinformation paradigm.

Discussion

~These experiments supported the work done by Johnson et al. (1993) which also showed that the qualities associated with the retrieved memory at time of retrieval allow people to make inferences as to the source origin

~The present results allow us to see a new dimension that contributes to reality-monitoring decisions: social influence, which increases with higher levels of source credibility.

~Experiment 1A & 1B showed:

(1) The impact of erroneous social influence on reality-monitoring decisions

(2) That undermining the credibility of the confederate reduced conformity

~Experiment 2 showed:

*Interrogative suggestibility (type of obtrusive social pressure)