Cook & Cook [Chris]
Cook, B. G., & Cook, L. (2008). Nonexperimental quantitative research and its role in guiding instruction. Intervention in School and Clinic, 44(2), 98-104.
Abstract by authors:
Different research designs answer different questions. Educators cannot use nonexperimental quantitative research designs, such as descriptive surveys and correlational research, to determine definitively that an intervention causes improved student outcomes and is an evidence-based practice. However, such research can (a) inform educators about a number of issues related to the education of students with disabilities and (b) guide experimental research efforts that can more definitively determine whether a practice is evidence based. Additionally, in the absence of relevant, high-quality experimental research, special educators can use some types of correlational research as the best available evidence to guide their instructional and curricular decision making.
Experimental design rarely used in educational research
Experiments require a control condition in which the instructional intervention is not used, for comparison with the experimental intervention, among other elements
Reviews surveys, comparing them with qualitative research
Notes that surveys provide a wide view of a phenomenon, whereas qualitative research describes in depth the phenomenon
Surveys used to measure attitudes and perceptions of stakeholders
Not necessarily appropriate to measure perceptions, as the stakeholders – e.g. teachers – may not adequately perceive whether an intervention has a desired effect on student outcomes
Three types of error needing to be controlled for in order to ensure valid survey results: sampling error, measurement error, and nonresponse error
Discover relationships and report on their direction – that is, do they co-exist or not in groups of people
Finding relationships between two variables can spur further research, including seeking evidence of causation
Experimental research is difficult to design given the many variables which may influence student outcomes; however, correlational research results which are repeatedly and independently found, using proper research methods, and which are connected to theory, can be useful to educators planning instruction.