METHODS: Advantages of an Online Survey
Anonymous Online Surveys Reduce Social Pressures to Provide Normative Responses
All responses to the Death Penalty Study questionnaire were completely anonymous. That means there is no way that researchers (or any outside party) can link any completed questionnaire to any individual who participated in the study. Despite the limitations of online survey research, there are several advantages. This is especially true when dealing with a sensitive topic (executions) where answers to questions are likely influenced by social pressures.
Questions about sensitive topics often generate what social scientists call normative responses. A normative response is an answer that a person considers socially acceptable, even if that answer does not really match that person's true feelings. In the case of the Death Penalty Study, people were asked questions that, in other contexts, would tap into a general moral stance that it's wrong to kill people. People were asks if they would attend an execution (Question #16) and, if so, would they be willing to actually execute (kill) the person they previously convicted (Question #17). Elsewhere in the questionnaire, people were asked if they think a person sentenced to death should suffer during his or her execution. They were asked if they approve of executing people in the U.S. today by cutting off their heads or stoning them to death. Many people said yes. These are not normative responses.
Using random digit dialing (RDD), it is possible to conduct a telephone survey that, in principle, can include over 95% of U.S. households. However, mobile phones and caller ID have greatly reduced the effectiveness of RDD at reaching a truly random sample of U.S. residents.
Further, an interview over the phone introduces a social interaction with another person. Researchers in the Death Penalty Study felt that people would give more honest answers to questions if no direct human communication was involved. Data from the study seemed to support this view. For example, 18% of people studied indicated a willingness to kill another human being. This is probably a higher percentage than telephone interviews would have revealed. The Death Penalty Study used a non-probability sample. As such, the strength of influences in the sample are more informative than statistical significance. Learn more about how to interpret the findings using effect size and Eta Squared.