Persuasion

Measures in Persuasion Research

Attitudes are the most common outcome measure of persuasion research, though there are many applied contexts in which actual behavior is also of interest. Many types of attitude measures exist but the two most commonly used are Likert and semantic differential items. Researchers most often measure attitudes directly after message exposure. They rely on random assignment to ensure that differences in attitudes are the result of the different messages to which they might expose an audience. It is also possible to measure attitudes both before and after message exposure to gain a more precise estimate of the extent to which the message changes the audience’s attitude. Researchers also sometimes measure attitudes after a particular period of time has elapsed since message exposure to determine if the attitude change associated with the message remains or if it decays and the audience’s attitudes return to pre-exposure positions. On the contrary, other researchers measure changes in beliefs related to attitudes every 77 milliseconds to examine how such changes occur dynamically during and after message exposure.

Another important aspect of attitudes concerns the strength of the attitudes. Attitude strength has been defined in many ways but it generally refers to how easy it would be to change an individual’s attitude toward some object. Some researchers use self-report measures such as Likert scales with statements indicating that their attitudes are firmly held, intensely supported, important, associated with strong emotions, or difficult to change. Others prefer to measure how quickly the subject can report her or his attitude (also called attitude accessibility). Ongoing research is attempting to determine which of these methods is the most valid.

A number of processes have been proposed as mediators of various message features and attitude change. One of the earliest to emerge was the extent to which the audience of the message fully comprehended the persuasive message’s intended meaning and arguments. It was thought that messages that are not understood cannot persuade. Comprehension has usually been measured by asking the audience to report back what the message said through a free-recall test. It is also sometimes measured by closed-ended recognition tests of message content.

The popularity of dual-process models of persuasion focused persuasion researchers on the extent to which the audience carefully cognitively processed the message. The elaboration likelihood model of Richard Petty and John Cacioppo in particular has provided many new measures and method to persuasion research.[Page 1225] Such models posit a continuum of cognitive processing ranging from the audience carefully considering all of the message arguments to an almost complete lack of cognitive processing devoted to the message. Both predictors of cognitive processing and the processing itself have been measured to varying degrees of success. Involvement in the message is considered a predictor of processing depth and has been measured with many ad hoc, single-item measures. A validated multi-item self-report measure of the various types of involvement was developed by Hyunyi Cho and Frank Boster in 2005. Others produced measures of personality traits such as need for cognition that were thought to predict general dispositions to process messages carefully. Another self-report measure, the message elaboration measure, which was developed by Ronald A. Reynolds, was designed to directly measure the extent to which the audience believed they carefully processed the message.

...