Degree Name

Master of Arts (MA)

Semester of Degree Completion


Thesis Director

Bruce C. Wheatley


Researchers and scholars in the fear appeal area have been unable to determine what factor or factors make fear appeals effective. Two studies, Hewgill and Miller (1965) and Powell (1965), suggest that the personal involvement of the audience with who or what is threatened is the factor which causes fear appeals to be effective. If ego-involvement is the personal involvement factor indicated by these studies, then it would appear to be the key to fear appeal effectiveness. Therefore, the present investigation was designed to experimentally compare the effects of ego-involvement and fear appeals upon task performance.

Sixty nine undergraduates enrolled in the basic speech course at Eastern Illinois University served as subjects. They were divided randomly into four groups, each of which took a two part test on basic speech fundamentals. The dependent variable was each subject’s performance (the number of questions answered during the ten minute time period) on the test. The independent variables, fear appeal and ego-involvement, were introduced through the instructions for parts “B” and “A” of the test respectively.

Part “A” of the test consisted of 79 questions. On part “A” of the test the ego-involvement variable consisted of informing the subjects in two of the experimental sections that the score they made on the test would count as an hourly examination grade in their basic speech class. The two non ego-involved sections were told that their scores would not affect their grades in basic speech class. The subjects in all sections were given 10 minutes to work on test “A”.

At the end of the first testing period the subjects were given test “B”. Included in the instructions for test “B” was the fear appeal variable. The two fear appeal sections were told that in order to pass test “A”, the national average indicated that they would have had to answer 37 questions. The experimenter had already determined that the majority of the subjects would be unable to answer this number of questions in the time allowed. The no fear appeal sections were told nothing about how they were doing on the test.

The results from the two tests were tabulated and the difference between each subject’s performance on each test was calculated. An analysis of variance was completed on this data. The results were not statistically significant.

In order to verify the independent variables, a posttest battery was given. The posttest battery consisted of six semantic differential scales, three ego-involvement positional choice scales, and finally, three ego-latitude scales. The results of this data was recorded in tabular form. An analysis of variance were completed on this data. The results of this analysis were not statistically significant.

The results indicated the following: (1) the subject of grades did not act as an ego-involved topic with the subjects tested; (2) performing up to the national average did not act as a fear appeal with the subjects tested; and (3) after checking all possible design errors, the unlikely option that ego-involvement and fear appeal have no predictable effect on communication seems less unlikely.