Survey: Questionnaire

Item Selection

The items in a questionnaire come from the questions researchers are trying to address in their study. For example, if a study is examining the links between family communication regarding divorce and family outcomes, the questionnaire should include items that are tied to those constructs. These items are one of two types: closed-ended questions or open-ended questions. Closed-ended questions consist of preselected choices for the participants, such as the choices “Yes” or “No” to the question, “Has your family ever been divorced?” Open-ended questions allow the participants to answer in their own words, such as “How did your family generally communicate about the topic of divorce?”

In terms of an individual study, a complete questionnaire consists of the operational definitions of all constructs in the study. Researchers can distinguish these operational definitions in one of two main ways. First, researchers commonly use already validated scales that consist of several items to address each specific construct in the study. In the current example, this means that the researcher might select scales that have already been created regarding general family communication patterns, specific communication about divorce, and family outcomes such as closeness or satisfaction. In each case, the researcher must decide that the scale is a valid operationalization of the construct. Second, the researcher might decide to create new items for the questionnaire. In this case, the researcher will have to argue in his or her analysis how the new item is a valid operationalization of the construct (e.g., face validity). In addition to the items, a questionnaire normally includes some demographic questions regarding the participant, including sex, ethnicity, age, and socioeconomic status.

...