Design a study question Flashcards
How do you design an experiment?
• Experimental method (what type, not quasi
• Variables: Operationalised (clearly defined and measurable) IV and DV.
• Experimental design
• Extraneous variables: Identification, effect and control of these; linked to the experimental design repeated measures, independent groups or matched pairs.
• Materials and apparatus: How specific resources could be used.
• Pilot study: How specific aspects of the design could be checked.
• Ethical issues: Especially consent, harm, deception; linked to the experimental method.
How do you design an observation?
Behavioural categories: How will you categorise behaviour?
-Operationalise these behaviours: as a type (e.g. verbal aggression, physical aggression); degree (e.g. mildly aggressive, very aggressive) or rating scale
-Avoid observer bias: How are you going to avoid observer bias? Consider using more than one observer to assess inter-observer reliability, standardised procedure/materials and a pilot study to check their consistent implementation.
• Sampling technique: How are you going to sample behaviour? How often and for how long you are going to observe the participants?
o Event sampling: Record specific behaviours whenever they occur (e.g. physically aggressive acts shown by children during morning break).
Time sampling: Record target behaviour at set time intervals (e.B. every 30 seconds during a 10-minute observation).
• Ethical issues: Especially privacy and informed consent with naturalistic observations.
Recording data: How? Quantitative data might involving tallying on a chart how many times a participant performs a target behaviour; qualitative data might involves making written notes of the behaviour observed; audio or video recording would create a permanent record of what was observed and enable re-watching to ensure no behaviours are missed.
• Pilot study: How specific aspects of the design could be checked.
what can self report methods be used as ?
as a research method or as a research technique used to gather data as part of another research method, such as an experiment
how would you design a questionnaire as a research method?
-The aim of the study may be to find out about smoking habits in young people. -the researcher would design a questionnaire to collect data about what people do and why
- the questionnaire is the research method.
How would you design questionnaire as a research technique?
-aims of a study might be to see if children who are exposed to an anti smoking educational programme have different attitudes towards smoking than children not exposed to such a programme.
-The researcher would use a questionnaire to collect data about attitudes but the analysis would involve a comparison between the two groups of children.
-The questionnaire a research technique used as part of a quasi-experimental method.
What is the design of a questionnaire?
• Kinds of questions: What types of data do you wish to collect? If you want quantitative data use closed questions obtaining numerical data; if you want qualitative data use open questions; for a combination of quantitative and qualitative data use open and closed questions.
• Write good questions: Ensure questions are not leading, ambiguous, complex or double-barrelled.
• Constructing good questionnaires: Consider the number of questions - keep it relatively short - no more than 20 questions; order of questions - start with easy ones saving questions that might make someone feel defensive until the respondent has relaxed; demand characteristics - are filler questions needed to reduce demand characteristics?
• Ethical issues: Especially confidentiality and informed consent.
• Pilot study: How specific aspects of the design could be checked.
• Sampling: Selecting participants to take part.
How would you design an interview?
• Structure of interviews: What types of data do you wish to collect? If you want
quantitative data use structured interviews, if you want qualitative data use
unstructured interviews; you want a combination of qualitative and quantitative data use
semi-structured interviews
• Write good questions: Ensure questions are not leading, ambiguous, complex or doublebarrelled.
• Conducting good interviews: Consider investigator effects – how the behaviour or
appearance of the interviewer could influence how the participants act; investigator bias
– consider more than one interviewer to check reliability; order of questions – start with
easy ones saving questions that might make someone feel defensive until the respondent
has relaxed; demand characteristics – are filler questions needed to reduce demand
characteristics?
• Ethical issues: Especially confidentiality and write to withdraw.
• Pilot study: How specific aspects of the design could be checked.
• Sampling: Selecting participants to take part.
How would you design a correlational analysis?
•Correlational method: Justification of choice over an experimental method.
• Co-variables: Operationalised (clearly defined and measurable).
• Materials and apparatus: How specific resources could be used.
• Pilot study: How specific aspects of the design could be checked.
• Ethical issues: Especially informed consent, deception
