Appendix A_questions for field

Appendix A
Research Questions for Field Experiments

APPENDIX A
RESEARCH QUESTIONS FOR FIELD EXPERIMENTS

Table A-1. Research questions for field experiments
Research question
What are the differences in response distributions by
telephone and mail?

Outcome measures
Means, response distributions, and variances.

Are there differences by mode once controlling for
differences in nonresponse and coverage?

Multivariate models.

Do mode effects differ by sociodemographic
characteristics of the respondent?
Are there differences related to the type of question that
is asked (e.g., behavioral, recall, attitudinal, and
sensitive information)?

Means, response distributions, and variances.
Multivariate models.

Are there differences in the format of the question?
- Open-ended questions
- Number of response categories
- Likert scales
Does the mode affect the correlation among scale and/or
related items?

Correlations and cross-tabulations among selected items.

What are the processes associated with the mail survey?

Debriefing information.

-

-

Do mail respondents complete the survey in the
presence of other respondents or
nonrespondents?
Do respondents engage in other activities while
filling out the questionnaire?
How often do respondents stop and start the mail
interview?

Are mode effects related to the processes associated with
the mail survey?

Mulitvariate models.

How often do respondents interact with interviewers?
Are these interactions related to questions that exhibit
mode effects?

Recording of interviews.

Are mode effects related to questions where respondents
are likely to interact with an interviewer?

Multivariate models.

A-1