Method

Collection of PSAs

The Armed Forces Information Service has approximately 4,000 radio PSAs in circulation at all times.  A convenience sample of three audiocassette tapes was obtained containing 189 radio PSAs covering a wide range of subjects.  Although this collection of PSAs is not exhaustive and only represents 4.7 percent of the total AFRTS collection, it is felt to be representative of what is being heard by military families overseas.

From the audiocassettes, the three-person coding team determined 51.3 percent of the PSAs to be health and safety related (n=97).  Because this information was drawn from a convenience sample, it is impossible to know whether this percentage is generalizable to the entire inventory.

Coding Procedure

     Three coders analyzed the manifest content of the 189 PSAs included in this study, with each coder responsible for one audiocassette, or an average of 63 PSAs each.  Because of concerns about inter-coder reliability, the study did not attempt to measure latent content of the spots.  Each coder was a member of the Department of Defense Joint Course in Communication at the University of Oklahoma.  The coding instrument (See Appendix B) was developed jointly by the four members of the research team.  The coders received two hours of training on the coding instrument and had access to the codebook (See Appendix C) during the analysis.  Inter-coder reliability, used by most researchers when conducting a content analysis, is the process of analyzing how two or more coders code the same data while working independently using the same coding sheet (Kaid & Wadsworth, 1989).  Having all three coders code one of the audiocassettes, or 35 percent of the PSAs, tested inter-coder reliability.  An inter-coder reliability of .86 was obtained.

Description of Variables Used in the Analysis

            Health or safety.  When listening to PSAs, coders were asked to first determine whether or not the PSA was health or safety related.  PSAs of this type could include subjects such as quitting smoking, eye exams, avoiding terrorism, using seat belts, and others.

            Fear appeal.  Coders then decided which health and safety PSAs used fear appeal.  A fear appeal was defined for the coders as a “threat meant to scare the listener about possible consequences of some action or of not taking some action.”  Fear appeal was coded as “Yes” or “No,” depending upon the coder’s judgment as to whether the variable was present.

            High level threat vs. low level threat.  After determining that a PSA was using fear appeal, coders were asked to determine whether the threat was high level or low level.  A high level threat was defined for the coders as “a direct verbal threat that suggests physical harm to the listener or his or her family members and/or contains frightening sound effects or images.”  A low-level threat was defined for the coders as “a threat to society or career that does not suggest physical harm to the viewer and does not contain frightening sound effects or images.”

     Each PSA determined to be both health and safety related and to use fear appeal was then coded for the presence or absence of the following five characteristics:

      Severity.  Severity refers to the seriousness of the threat.  A specific claim that the threat is “serious” would obviously meet the criteria for severity.  A threat that would cause serious physical harm, such as, but not limited to, death, was coded as severe (e.g., “Glaucoma is a serious threat.  If untreated, it can cause blindness.”)  Severity was coded as “Yes” or “No,” depending upon the coder’s judgment as to whether the message included info on the severity of the specific health behavior.

    Susceptibility.  Susceptibility refers to whether or not the PSA conveyed a sense that the listener has a good chance of experiencing this threat.  Telling the listener that “military members are especially at risk for glaucoma and the number of cases grows every day” would convey a high sense of susceptibility. Susceptibility was coded as “Yes” or “No,” depending upon the coder’s judgment as to whether the variable was present.

     Response efficacy.  Response efficacy refers to the cues or action steps offered by the PSA.  Specifically, does the message indicate that there is something the listener can do to avoid the threat?  (e.g., “Fortunately, there’s a new medical procedure that can stop glaucoma before it starts.”)  Response efficacy was coded as “Yes” or “No,” depending upon the coder’s judgment as to whether the variable was present.

     Self-efficacy.  Self-efficacy refers to the ability of a person to effectively perform the action step.  Does the message make it clear that the solution is something the listener can do?  (e.g., “And this medical procedure is available now, on this base.”)  Self-efficacy was also coded as “Yes” or “No.”

     Ease of self-efficacy.  Coders were asked to determine whether the PSA indicated that the action step would be convenient and easy.  (e.g., “All it takes is a quick trip down to your base health clinic for a free, 15-minute eye exam.”)  Once again, ease of self-efficacy was coded as “Yes” or “No.”