FOOD:

This is an important experiment for my study. I would like to study how the number of candidates affects voters’ preferences for those candidates, so here I create some fake candidates so I can manipulate the number of candidates that the voter learns about. I am interested in seeing whether people are more likely to choose the candidate seen as more moderate or more viable. If you have not already discussed my project with me, you may enjoy reading http://www.duke.edu/~rml5/dis/PoliTheory1.doc . I still welcome comments on that theory chapter…

On Friday (or whenever we meet), I would like to discuss the following files:

http://www.duke.edu/~rml5/dis/localq.doc and http://www.duke.edu/~rml5/dis/candidateprofiles.xls. This email is http://www.duke.edu/~rml5/dis/FOOD3-13.htm

This experiment is designed to include two candidates that present a difficult, but stark, contrast. Respondents who learn about only two candidates (Condition 1) are expected to have a difficult decision choosing between two. The presence of a more moderate candidate makes the decision less difficult (less anxiety inducing, easier to justify, etc). So, condition #2 adds a candidate who espouses views more moderate than the two base candidates. Condition #3 adds a candidate more extreme so that one of the base candidates appears more moderate (a range-frequency effect). Condition #4 adds a candidate that adds a new dimension to the campaign.

To make this work, I need to create credible candidates and present the candidates in what would appear to be a realistic “voter’s guide.” So, the candidates I present here are modeled on real candidates who ran in this area in the last couple of elections and the information provided is in a format as near as possible as the format used in the newspapers’ voters guides. The most important information is in the priority field. The other information is included to give a more realistic feel to the voter guide.

I appreciate any and all feedback, but I especially am curious about the following:

Do you think my candidates and voters guide is realistic? Please consider the tone, length and language of profiles: Is it appropriate for respondents?

            Please note my use of “We need…” phrases, instead of “I will…” or “I have…” phrases most commonly seen in the profiles of successful candidates. What do you think of my use of “We need…” phrasing?

The other information can be randomly rotated to each of the candidates. Do you think all of the other pieces of information leaves the candidates as equal –other than their “priority”?

What do you think about the use of the most important question at the beginning? Do you think that it is a good idea to prime the respondents to consider a broad (comprehensive?) range of salient local issues? Do you think it is a bad idea to prime them, so it would be best if I use an open ended format or drop the question altogether?

Should I include explicit instructions to not turn back the page?

The other use of the MIP question is to see whether the respondent is predisposed towards a particular dimension, especially for the fourth condition. If crime is the most important problem, then the “new” candidate in condition #4 should be the easiest and most likely choice. If not, the respondent should be facing an even more difficult decision than she might have had if she had been in Condition #1. In such a scenario, the crime might be the “easiest” decision since the respondent is not reminded of any tough policy trade-off.

What is the best way of wording the scale question (question #6, on page 4). Option A, option B or ??

What is the best way of labeling the options on the scale?

What do you think of the question ordering?