Research Methods and Ethics

Research design fundamentals

A research design is the overall plan that links a research question to data collection and analysis, and its main purpose is to make the conclusions credible. Before choosing a design, a researcher must know how the key terms relate to one another. A construct such as intelligence must be distinguished from the specific abilities proposed within it. For example, Gardner's theory of multiple intelligences includes linguistic, logical-mathematical and spatial intelligences, whereas insight is a problem-solving concept from the Gestalt tradition and is not one of Gardner's intelligences. Reliability refers to the consistency of results. When a study repeated under the same or similar conditions yields consistent findings, the results are reproducible, and replication is the standard way to establish this reliability. Validity is different. It asks whether a measure or study actually captures the intended construct, and it requires evidence from design and measurement rather than repetition alone. In an experiment, the independent variable is the factor the researcher deliberately manipulates, and the dependent variable is the outcome that is measured. If room temperature is varied to see its effect on aggression, temperature is the independent variable and aggression is the dependent variable. A hypothesis is a tentative answer to a research problem, stated as a clear relationship between the independent and dependent variables. The alternate hypothesis states that a relationship or difference exists, while the null hypothesis states that none exists. The nil hypothesis is an older label for the null form, whereas 'phi-hypothesis' is not a standard term in research methodology. The independent variable is also called the explanatory variable, since it accounts for variation in the outcome. In regression and structural equation modelling, an exogenous variable is one whose value is determined outside the model and is often used as the predictor, so it is treated as a synonym for an independent variable in those contexts. Blinding is used to limit expectancy and observer bias. In a single blind study, participants do not know their group allocation while the experimenter does. In a double blind study, neither the participants nor the experimenter knows who is in the experimental or control group. Random experimentation, by contrast, refers to random assignment, which is a different procedure. Confounding occurs when an uncontrolled variable makes it uncertain whether the dependent variable was caused by the independent variable. High internal validity allows clear causal conclusions from an experiment, while high external validity means that the results generalise to other settings and populations. The placebo effect is an improvement produced by a participant's expectation of treatment rather than by the treatment itself. Mixed-method research combines qualitative and quantitative components, but its designs are not limited to quantitative-qualitative combinations, since qualitative-qualitative mixing is also recognised. Mixed-method approaches draw on pragmatism and other paradigms and do not require the positivist paradigm.

Qualitative and descriptive research methods

Qualitative and descriptive methods aim to describe people, groups and events in depth rather than to test causal hypotheses. The narrative method collects data in the form of cohesive stories about individuals' lives and experiences. Ethnography describes the cultural characteristics of a group in detail, usually through prolonged immersion in the community being studied. A case study examines a person, event or experience in detail, often using multiple sources of information. Content analysis is a systematic method of coding text or other materials and extracting themes from them. A classic illustration is psychoanalysis, which Sigmund Freud built largely from intensive case studies of individual patients, including their free associations and dreams, rather than from surveys or introspection alone. Grounded theory is an inductive approach in which theoretical concepts and categories emerge from the data through systematic coding and constant comparison. The theory is therefore generated by the data rather than assumed before the data are collected. Intensive methods collect detailed information from a small number of people. A depth interview is an open-ended interview with a few participants, and a case study examines one or a few individuals in depth. By contrast, surveys gather standardised responses from large samples, and correlational research measures relationships across generally large samples. Neither yields detailed information on a few persons. Thick description is a concept from ethnography introduced by Clifford Geertz in 1973. It records the layered meaning structures that a culture-sharing group uses to interpret its actions, and in interpretive ethnography these meanings are treated as co-constructed between the researcher and participants during fieldwork. Sigmund Freud, however, is associated with psychoanalysis rather than with the ethnographic method.