Reliability and validity
- Created by: jessicawarren
- Created on: 26-04-16 10:19
View mindmap
- Reliability and Validity
- Reliability
- Refers to consistency of measurement i.e get the same result every time
- Internal: How consistent a method is inside the study e.g same method for everyone, standardised methods
- External: Consistency over time and on different occasions e.g you look at it again/someone else looks at it and finds same results
- Experiments more reliable because they are easy to replicate- simple procedure, change one variable, control all other factors
- Observations not reliable: subjective/observer bias. Improved by operationalizing observations by using behaviour checklist, split half method, test-retest method
- Split half method: One half of test compared to the other to check if scores are consistent
- Test-retest method: Test carried out on same/similar participants several times and similarity of results recorded.
- Higher correlation coefficient= higher reliability
- Validity
- Whether a study measures exactly what is meant to measure and whether behaviour truly represents what the researcher is investigating
- Internal: Measuring the right variable/ control e.g controlling extraneous variables
- External: What goes on outside experiment, generalising to other people/ situations (ecological/ population validity)
- Lab experiment valid: Extraneous variables controlled, internal validity. However, lacks EV peoples behaviour may be different in artificial environment
- Self report data not valid: Social desirability decreases accuracy, low internal validity
- Pilot study: First study to test any potential issues and operationalize all variables
- Reliability
Comments
No comments have yet been made