Quantitative: Causal-Comparative Research Like experimental and quasi-experimental research, causal-comparative research (sometimes called ex post facto research) attempts to establish cause-and-effect relationships. However, the researcher has much less control over the independent variable and cannot use randomness in selection and assignment. For example, a researcher might use causal-comparative research to attempt to establish a causal relationship between students’ home environments and differentiated instruction. The methodology should answer the following specific questions: 1. What subtype will be used? Four subtypes are frequently used:between-group, time series, path model, and archival time series. 2. If a between-group design is used, how will the homogeneity of groups be established? If a path model, what theory will be used as the basis? For archival time series, how will you control for history, maturation, and mortality? 3. In the interpretation of results, how will you avoid the post hoc fallacy—assuming that because one factor preceded another, it must have caused the other? #researchmethodology #dissertation #research #thesis #researchdesign #CausalComparativeResearch

Quantitative: Correlational Research Correlational studies attempt to understand patterns of relationships among variables. Though such studies cannot prove causation, they are useful in predicting one variable from another or building a theory about a complex phenomenon. Correlational research would be used to answer this question: “How are smartphone usage for required mathematics assignments and assignment submission related?” In detailing the methodology, respond to the following issues: 1. What subtype will be used: bivariate correlations, multiple regression, discriminant analysis, partial correlations, or factor analysis? 2. How will the variables be measured? How will the validity and reliability of the measures be established? Why were these variables selected? 3. How will data be displayed? Will scatter plots or correlation matrices be used? 4. What correlation statistics will be used? Why were they chosen? #researchmethodology #dissertation #research #thesis #researchdesign #CorrelationalResearch

Quantitative: Descriptive Research Descriptive research is used to describe the characteristics of a population by directly examining samples of that population. Descriptive studies make primary use of surveys, interviews, and observations. They are usually undertaken in the early stages of a phenomenon. Thus, a researcher might decide to use descriptive research to answer this question: “What are the distinguishing characteristics of teachers who use peer assessment as an evaluative technique?” These are the special questions to address, in addition to the general design issues previously noted: 1. How will the sample be selected—convenience, volunteer, quotas, or probability sampling? 2. What variables will be studied? Why were they selected? 3. If a survey will be used, what response rate will be required? How will you deal with the problem of nonresponse bias? 4. What statistics will be used to report results? How will data be displayed? #researchmethodology #dissertation #research #thesis #researchdesign #DescriptiveResearch

Evaluation Research Evaluation research makes judgments about the merit or worth of educational programs, products, and organisations, usually undertaken to assist administrators in making professional decisions. Although evaluation is considered a type of research here, many research experts distinguish between research and evaluation. For example, Cohen et al. (2011) note the features that they believe distinguish evaluation from research: Evaluation is parochial, in the sense that it focuses on a single entity; it examines multiple aspects of the unit being studied; it often originates with a client, not the researcher; and it is often undertaken when a decision must be made. Of all these distinguishing features, the parochialism of evaluation seems to be the major obstacle in justifying it as research. Thus, a study is not likely to make a significant contribution to knowledge if it reaches a conclusion of this sort: “Instructional strategy is not effective.” However, an evaluation study can escape the limits of parochialism if it accomplishes one or more of the following purposes: • Provides an early test of a new approach or model of evaluation. • Develops an instrument that can be used in other studies. • Evaluates a program that is widely used but has had little systematic evaluation. • Documents how the evaluation results were used by groups of stakeholders. An evaluation research methodology should address the following specific issues: 1. Who are the clients? Who initiated the request for evaluation? 2. What evaluation model or theory will be used? 3. Who are the stakeholders? Who has an interest in the results? 4. What criteria will be used? How were they derived? 5. What standards will be used to make judgments? #researchmethodology #dissertation #research #thesis #researchdesign #Evaluationresearch

@Hamza above suggests that 'Most experimental studies are intended to establish cause-and-effect relationships'.

In any educational setting such relationships will usually be affected by a range of visible and less visible context-specific variables.

Unless these variables are clearly identified/taken into account, the links that we make are unlikely to be valid. Where the context of the setting is recognised, causal relationships that we think we have discovered will also be context-specific.

Does this 'limit' the value of such studies in education?

Martin Wedell Yes, it does limit the generalizability of experimental studies in education, but not their value. The context limits general claims but doesn't diminish the value of the experimental studies. Instead, it urges us to carefully interpret the findings and focus on how and why something works, not just if it works.