Psychology - Methods

True experiment

True experiment

• Participants are randomly allocated to conditions

• Indépendant variable is manipulated and the dependant variable is measured

• Data is statistically analysed and the p-value is calculated

• It is easy to reduce extraneous variables leading to high internal validity

Focus Group

• A group interview that acts as a purposive sample of 8-12 parti

• Researcher acts as the facilitator of the group and keeps the discussion focused

• Data is qualitative and must be interpreted through content analysis

• Participants may prompt responses from other participants

• Maybe be biased due to social desirability bias

Case Study

• Based on method triangulation

• Data can be quantitative and qualitative

• Normally longitudinal in nature

• Focused and detailed look at one person, group or organisation

Naturalistic Observation

• Can be overt or covert

• Occurs in a natural environment so high external validity

• Hard to rule out extraneous variables

• Can use an observational checklist to obtain quantitative data

• Can use field notes to obtain qualitative data, these are often difficult to analyse and content analysis is used to interpret the data

• Typically multiple researchers are used to eliminate bias

Semi Structured Interview

• Follows the flow of the conversation but has predetermined questions, follow up questions are asked

• Face to face so may have interviewer effect and social desirability bias

• Results must be interpreted through content analysis

Natural Experiment

• Participants are not randomly allocated to conditions

• The independent variable is naturally occurring and the dependant variable is measured but causality cannot be proved

• Not possible to control extraneous variables

• Data is statistically analysed and the p-value is calculated

Correlations Study

• Often pre-existing data sets are used but they may also be collected through surveys or standardised tests

• No manipulation of an independent variable, instead data sets are compared to determine a level of correlation

• Produces qualitative data which must be statistically analysed for significance

Questionnaire/Survey

• Not face to face so interviewer effects are avoided

• Allow participants to remain anonymous

• Questions may be misinterpreted by participants

• Data may be qualitative or quantitative, qualitative data must be interpreted through content analysis

Quasi Experiment

• Although an independent variable may be changed and a dependable variable measured, causality cannot be determined

• Attempt to control extraneous variables, increasing internal validity

• Participants are not randomly assigned to conditions that are grouped based on a participant variable

• Data is statistically analysed and a p-value calculated

Unstructured Interview

• No set questions only topics with the goal of avoiding leading questions

• Interview style is highly naturalistic

• Face to face so may have interviewer effect and social desirability bias

• Requires more training, can be difficult to keep conversation on track without a script

• Data must be interpreted through content analysis