Cognitive Connie
Understanding questionnaires, interviews & observations
Experiments are the gold standard for establishing causation, but they are not the only — or even the primary — tool in psychological research. Many of the most important questions in psychology cannot be answered by assigning people to conditions in a laboratory: what is it like to live with psychosis? How do children actually negotiate conflict in playgrounds? What values drive political behaviour?
Key figures
Rensis Likert
1903–1981American organisational psychologist who developed the Likert scale in 1932 — the ubiquitous response format in which participants rate agreement with statements on a 5- or 7-point scale from "strongly disagree" to "strongly agree." Likert's innovation was to sum responses across multiple items measuring a single attitude, producing a more reliable composite score than any single item alone. The Likert scale became one of the most widely used data collection tools in social science history.
Erving Goffman
1922–1982Canadian-American sociologist who used participant observation in his fieldwork at a Shetland Island hotel (The Presentation of Self in Everyday Life, 1959) and covert participant observation in a psychiatric institution (Asylums, 1961) to develop dramaturgical theory — the idea that social life is a performance. Goffman's work exemplifies both the power of participant observation to reveal hidden social structures and the ethical tensions of covert methods.
Elizabeth Loftus
1944–Cognitive psychologist whose research on the misinformation effect and leading questions (from Loftus & Palmer, 1974, onward) fundamentally altered our understanding of memory, eyewitness testimony, and questionnaire design. Loftus demonstrated that the wording of a question can alter the memory of an event — not merely the verbal report. Her work has influenced legal procedures around eyewitness testimony and is the most cited evidence for why question wording must be carefully controlled.
Key concepts
Closed questions
Questions with a fixed set of response options: Likert scales, yes/no items, multiple-choice options. Generate quantitative data suitable for statistical analysis and comparison across large samples. Advantages: easy to score, comparable across respondents, amenable to inferential statistics. Disadvantages: restrict what participants can say, may not capture nuance or unexpected responses, and responses may reflect the available options rather than participants' genuine views.
Open questions
Questions that invite participants to respond freely in their own words, generating qualitative data. Advantages: rich in detail, sensitive to individual variation, can surface unexpected themes. Disadvantages: harder to quantify, time-consuming to analyse, and difficult to compare systematically across respondents. Open questions are typically analysed through thematic analysis or content analysis.
Leading questions
Questions whose wording suggests an expected or desirable answer, biasing responses away from genuine attitudes or experiences. Demonstrated powerfully by Loftus and Palmer (1974): "How fast were the cars going when they smashed into each other?" elicited higher speed estimates than "when they contacted each other." In questionnaire design, avoiding leading questions requires careful neutral wording and piloting with a representative sample.
Structured interview
Every participant is asked the same predetermined questions in the same order, with no deviation. Produces data directly comparable across participants and enables statistical analysis. Advantages: high reliability, standardised across interviewers, easy to analyse. Disadvantage: inflexible — cannot follow up unexpected but potentially important responses.
Semi-structured interview
A topic guide provides predetermined themes and key questions, but the interviewer is free to probe, rephrase, and follow unexpected directions within those topics. The most commonly used interview format in qualitative psychology. Balances comparability across interviews with flexibility to explore individual experience in depth. Analysed typically through thematic analysis or interpretative phenomenological analysis (IPA).
Unstructured interview
A conversational approach following the participant's lead with minimal predetermined structure. Maximises flexibility and participant direction — valuable for exploratory research or phenomena where the researcher does not yet know what questions to ask. Difficult to analyse systematically or compare across participants. Risk of interviewer bias shaping the conversation.
Naturalistic observation
Observing behaviour in the participant's natural setting without manipulation or intervention. Advantages: high ecological validity — the behaviour observed reflects what people actually do in natural contexts. Disadvantages: no control over extraneous variables, cannot establish causation, observer effects may alter behaviour if participants know they are being watched.
Overt vs covert observation
In overt observation, participants know they are being observed (consent obtained; risk of Hawthorne effect — altering behaviour because of observation awareness). In covert observation, participants are unaware — yielding more natural behaviour but raising ethical concerns about consent. Covert observation in public settings (where no reasonable expectation of privacy exists) is less ethically problematic than in private spaces, but still requires ethical justification.
Participant vs non-participant observation
In non-participant observation, the researcher remains external to the group being observed. In participant observation, the researcher joins the group — taking on a role within it — to observe from the inside (ethnographic approach). Provides access to tacit, informal, and context-dependent knowledge invisible to an outside observer. Risk: going native — over-identifying with the group and losing analytical objectivity.
Inter-observer reliability in observations
Two independent trained observers code the same behaviour and their ratings are compared. Quantified by percentage agreement or Cohen's kappa (κ). High inter-observer reliability (κ ≥ 0.7) indicates that the behavioural categories are clearly defined and consistently applied. Low reliability indicates ambiguous categories or insufficient observer training, both of which undermine the validity of the observational data.
Test your knowledge
Frequently asked questions
When should you use a questionnaire versus an interview?+
Questionnaires are preferable when: you need to reach large numbers of participants efficiently, your questions are sufficiently well-defined to be answered without clarification, you want quantitative data for statistical comparison, and anonymity is important for honest responses on sensitive topics. Interviews are preferable when: you need depth and nuance rather than breadth, the topic is complex or poorly understood and you want to follow unexpected directions, participants' experiences and meanings are the object of study, or the questions are too complex for written self-report. Semi-structured interviews allow both structure and flexibility and are appropriate for most qualitative research questions.
What is the Hawthorne effect and how does it affect observational research?+
The Hawthorne effect refers to participants modifying their behaviour because they know they are being observed — named after a series of studies at the Western Electric Hawthorne plant in the 1920s–30s. In observational research, overt observation risks this: workers (or participants in any setting) who know a researcher is present may perform better, be more compliant, or suppress behaviours they consider socially undesirable. Covert observation avoids this problem but raises ethical concerns. Habituation — allowing participants to become accustomed to the observer's presence before data collection begins — is a practical strategy for reducing Hawthorne effects in overt designs.
What is thematic analysis and when is it used?+
Thematic analysis (Braun & Clarke, 2006) is the most widely used method for analysing qualitative data in psychology. It involves: familiarising yourself with the data, generating initial codes (recurring ideas, concepts), clustering codes into candidate themes, reviewing and refining themes, defining and naming final themes, and writing up the analysis. It is appropriate for interview transcripts, open-ended survey responses, focus group data, and any text where you want to identify patterns of meaning across participants. It can be used inductively (themes driven by the data) or deductively (themes guided by theory), and is compatible with a range of epistemological positions.
Sources
Last reviewed July 2025- 1.
Braun V. & Clarke V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101.
+About this source
Foundational paper defining the thematic analysis method and providing the six-phase guide widely used in qualitative psychology research.
- 2.
Creswell J.W. (2014). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches (4th ed.). Sage.
+About this source
Comprehensive textbook covering quantitative, qualitative, and mixed methods approaches to research design.