Guidelines for Social Science Research

Wake Forest Research Fellowship Program

The ultimate goal of social science research is to understand, explain, and make inferences about human behavior. Empirical research emphasizes direct observation and experimentation as a way to answer questions. Everyday observations are not always made carefully and systematically, nor do they control for factors that may influence the events being observed. Research requires an hypothesis or question to be tested, a methodology that addresses the hypothesis, and analyses that test how well the data answer the research question. Two research paradigms are generally used in the sorts of studies conducted by Research Fellowship students: Observational studies and Questionnaire studies.

Observational Studies. Scientific observations are made under clearly defined conditions, in a systematic, objective manner, and with careful recording. Naturalistic observations are made by unobtrusive observers in the location(s) where the behaviors of interest normally occur (e.g., children's aggressive interactions on a playground). Narrative records provide faithful descriptions of all ongoing behaviors as they happen, without focusing on any particular behavior. Narrative records may be written, tape recorded, or video recorded.

Analysis of narrative records requires data reduction through careful coding of behaviors according to specific predefined criteria (e.g., kicking, slapping). The frequencies or durations of specific behaviors are calculated and compared statistically. In order to control for observer biases, interobserver reliability should be established by examining the coding agreement of two independent observers who review the same sample of behavior (e.g., the same videotapes). Personal biases and subjective impressions are not appropriate in social science research.

Questionnaire Studies. Questionnaire studies focus upon one behavior, or a limited set of behaviors, and responses are collected in a standardized fashion from every study participant. Surveys or questionnaires are composed of a set of predetermined questions that are asked in the same way for all respondents. Oral or written responses constitute the principal data obtained from a survey, and when properly conducted, surveys are an excellent way to describe the attitudes and opinions of a group. Questionnaires must be constructed prior to data collection, and the way a question is asked is critical to the validity of the responses given (e.g., "How often do you beat your wife?" vs. "Do you ever physically fight with your wife?"). Good survey questions should: contain simple, direct vocabulary that is familiar to all respondents; be clear and specific; not leading or loaded to obtain a desired response; be as short as possible; not contain offensive language or concepts; be readable; and use appropriate response formats.

Questionnaires should be pilot tested with a small group of participants similar to the actual study participants. Pilot testing insures that participants' responses reflect the information of interest to the researcher and not bias due to wording or question format. Pilot testing is essential when questionnaires are given to different cultural groups who may interpret questions differently. Appropriate modifications in questions, based on the data from the pilot study participants, should be made prior to administering the questionnaire to the actual study participants.

Evidence for both the reliability and validity of a questionnaire is essential for collecting meaningful data. As with observational data, questionnaire data requires data reduction through coding and statistical analysis.

Samples. For both observational and questionnaire studies, the sample of individuals who will participate in a study should be identified well before the study begins. Study participants should be a clearly defined group (e.g., African-American men over age 50) selected because of their relevance to the study hypothesis and their representativeness of the population of interest. They should not be a heterogeneous mixture of whoever is available.

If a study is to be conducted in another country, long before leaving the USA, the researcher should make contact with colleagues who will help locate individuals who are willing to participate in the study. Do not expect to get off the plane and find potential research participant waiting to take part in your study.

Many individuals are reluctant to answer questions or participate in research because they are unsure of what will be done with the information collected. Many individuals are generally suspicious of researchers or of Americans, and in some cultures "giving away" information about oneself is considered inappropriate or dangerous. It is vitally important that every detail of the study's methodology be planned before going to the study location.

Institutional Review Board (IRB). Wake Forest University must comply with federal guidelines governing research with human subjects. Therefore, any study conducted with human participants must be submitted to the IRB. Should one's project call for IRB approval, the WFURF Committee strongly recommends that said approval be sought and gained PRIOR to the submission of your proposal on or before March 17, 2008. The IRB, which is composed of faculty members and community representatives, will examine your study description to ensure that subjects are treated safely and respectfully. IRB application forms are available from the Office of Research and Sponsored Programs, Room 117E, Reynolda Hall or on the web at: http://www.wfu.edu/rsp/irb/forms.html.

The IRB is moving to eIRB, an electronic submission and review system, this fall. Access to the eIRB training schedule, training materials and the eIRB submission site through the Reynolda Campus IRB website. Some procedures are changing with the advent of eIRB. Researchers are advised to review these changes on the eIRB FAQs accessible on the IRB web.

Applications that qualify for exempt or expedited review may be submitted any time. For purposes of planning, the IRB coordinator has provided the following schedule for researchers submitting protocols that must be reviewed by the full IRB: