|
Of course, your particular study may not actually be an experiment. To determine whether your study is experimental, ask yourself whether any of the variables that look like IVs have involved the random allocation of participants to conditions or of orders of conditions to participants. If not, then your study is not experimental and you should not describe it as one. (You can find more about variables that are not true IVs in Section 13.8 of the book and also Section F of this Web site.)
A strict approach would bar you from referring to IVs and DVs in the of such a study. Instead, you might use terms such as predictor variables and criterion variables. (You can find more about other alternative terms you could use in Section 13.8 of the book.) However, you will come across published papers of nonexperimental studies that refer to independent variables so there is no consensus on this.
The main thing is that you do not describe your study as experimental when it is not. If you elect to include a section, describe the type of study clearly. Here is a very broad template that can help you with this:
| |
Design
The study employed a nonexperimental approach to investigate the relationship between XXX and YYY. Principal questions/items used to measure XXX were [specify]. Scales used were [describe typical scales]. Exceptions included [describe any exceptions]. Principal questions/items used to measure YYY were [specify]. Scales used were [describe typical scales]. Exceptions included [describe any exceptions]. |
Here is an example using the above template:
| |
Design
The study employed a nonexperimental approach to investigate the relationship between variables in Taylor’s (1983) theory of cognitive adaptation (TCA) and response to contracting venous thrombosis. Items measured the core constructs of the TCA: meaning, mastery, self-esteem and optimism. Outcome variables were measures of anxiety, depression, thrombosis worries and quality of life. |
Precise details of the measures developed or used to assess the variables in the TCA and to measure the outcomes would then appear in the or section. You can see how this has been done in the above study if you read the original reference:
Moore, T. J., Norman, P., & Harris, P. R. (2006). Cognitive appraisals as predictors of psychological distress following venous thromboembolic disease: An application of the theory of cognitive adaptation. Social Science and Medicine,63, 2395-2406.
Depending on the nature of the study you may nevertheless be able to report it using the typical sections of the report, with a lengthier and shorter than is the norm for laboratory-based experiments and perhaps with the section on questionnaire development described in Section E3 of this Web site. However, these labels may not be the most appropriate for the material you need to describe, in which case you should consider suitable alternatives, such as or instead of and instead of . The important thing is to make sure you report the material clearly and accurately.
With a questionnaire used as a survey, you will need to concern yourself with the details of how you went about sampling and from what population rather more than is the norm for laboratory-based experiments. External validity is central to surveys. (To find out more about external validity see Section 10.8 of the book.) You might consider sections labelled and respectively.
|
