RM - Reliability Flashcards
What is reliability?
Refers to how consistent the findings from an investigation or measuring devise are.
How can reliability be assessed?
- Test-retest
- Inter-observer reliability
How does test-retest assess reliability?
This involves administering the same test or questionnaire to the same person/ppl on diff. occasions. If it is reliable the results collected should be consistent.
Must be sufficient time between tests to ensure they don’t remember answers but not so long that views have changed. The sets of data are correlated and if the corre. is sig. and positive reliability is good.
When and why is inter-observer used?
Used during observational studies. A researchers interpretation of events may differ widely from someone else’s. This can result in subjectivity bias.
What does inter-observer reliability involve?
Researchers should conduct their observations in teams of at least 2. May run a pilot study to check that the beh. categories are being applied in the same way by each observer. They must record their data independently. Data correlated to assess reliability.
What other tests are used for reliability?
Content analysis - inter-rater reliability
Interviews - inter-interviewer reliability
How can we improve reliability of questionnaires?
Reliability of Qs can be measured over time by test-retest. Q that produces low test-retest r. may need some items to be ‘deselected’ or rewritten.
One solution is to replace open Qs w/ closed Qs which are fixed choice alternatives which are less ambiguous and avoid misinterpretation.
How can the reliability of interviews be improved?
Use the same interviewer each time. If this is not possible, all interviewers must be properly trained - avoid Qs that are leading or too ambiguous. This more easily avoided in structured interviews, interviewers beh. more controlled by fixed Qs.
Unstruc. and free-flowing interviews less likely to be reliable.
How can the reliability of experiments be improved?
Lab exp.s often reliable due to strict control over variables like instructions given to ps and the conditions they are tested in. Reliability in exp.s is more abt. replication of a particular method than reliability of findings.
How can the reliability of observations be improved in terms of beh. categories?
Can be improved by making sure that beh. categories have been properly operationalised and they are measurable and self-evident. Categories should not overlap and all possible beh.s should be covered on checklist.
If not operationalised, overlapping or absent, observers may make own judgments of what to record - differing and inconsistent results.
What is internal reliability?
Different parts of the test should give consistent results. Can be assessed using the split-half method.
What is external reliability?
The test should produce consistent results regardless of when its used. E.g. if u took the IQ test on 2 separate days u should get the same results.
