RETURN TO
PRINT BY SECTIONS

 


Analyzing Credibility:  A Study Examining
Demographic Factors and Personality Traits
that Influence Military Public Affairs' Credibility


Instruments

       To measure credibility, we use the McCroskey’s (1966) source credibility scale because it has been proven by many researchers to have face and criterion-related validity.  The article that first introduced this scale has been referenced at least 100 times in previous research.  This scale has been used to confirm presumed credible speakers will actually be perceived as credible (Rubin, 1981).  The format of the scale has emerged as the “predominant method of scaling” (Rubin, 1981, p. 335) credibility.
       The communicative competence scale created by Wiemann (1977) will be used to measure competence of the PAO.  This scale has been used in research numerous times and found to be effective with alphas between .85 and .91 (Rubin, 1981).  The specific scale used for the PAOs is an adapted self-report format employed by Cupach and Spitzberg (1983). 
       Many methods are available to test for assertiveness and assertiveness training is offered by many different organizations.  The World Wide Web offers thousands of self-administered assertiveness tests. For the purpose of this study, we are employing the Rathus (1973) Assertiveness Schedule which Lorr and More (1980) claim as “one of the better known self-report measures” (p. 128).
       Cegala’s (1981) scale measures the degree a person possesses for high or low interaction involvement.  Researchers use this scale today because of its track record of reliability and validity.  “Test-retest reliability for the trait Interaction Involvement Scale appears to be very good” (Rubin, 1981, p. 187).  Tests performed by Cegala et al. (1981) reported their test-retest with reliable alpha levels in the .80s.  Other tests by researchers reported test-retest results with alpha levels in the .60s.  “The Interaction Involvement Scale also appears to be internally consistent” (Rubin, 1981, p. 187).  Validity tests performed on this scale yielded evidence to concur strong validity.

Statistical Analysis

       Once the three year survey is complete, the data collected could be analyzed using one of two quantitative methods.  The first possible method is to perform an analysis of variance (ANOVA) on the results.  The other approach is multiple regression.
       The ANOVA can be used since this study examines at least two or more categorical independent variables, each with at least two levels (i.e. personality trait high and low levels).  This factoral design will analyze the participants’ self-reported results of the three trait independent variables which will be demonstrated by 2 x 2 x 2 = F (credibility).  A median will be found for each trait exposing the high and low levels of that trait.  This test will show the effect of each variable independently on the dependent variable and the interaction of the variables together.  A Pearson’s correlation will be run between each demographic variable and the dependent variable to examine the strength of the relationship between the variables.
       Table 1 is a factorial ANOVA visual demonstration of this credibility study.

Table 1.

First                             Second                        Third                            Fourth
Independent               Independent                 Independent               Independent
Variable                      Variable                       Variable                     Variable
_________________________________________________________________

Demographics:           Communication             Assertiveness:        Interaction
Years of service          Competency:                                                   Involvement:
Training                        Hi/Low                            Hi/Low                      Hi/Low

       Multiple regression could also be used to analyze the results.  Each subject will have an average score for each predictor variable (personality trait).  The three independent variables will be utilized to predict individual’s credibility levels.  Credibility results will be collected from the supervisor’s source-credibility survey. 

RETURN TO
PRINT BY SECTIONS