Validity And Reliability: Self-reports Reliability Flashcards
What are the types of reliability relevant to self reports
- internal reliability
- external reliability
What is internal reliability
- A measure of the extent to which something is consistent within itself
- e.g. all the Qs on a questionnaire should be measuring the same thing
what is external reliability
- A measure of consistency over time
- e.g. if the same interview/questionnaire was conducted one day then a week later, the outcome should be the same
What is the method for assessing internal reliability
- Split half method
1. Randomly select half the test items and placing them on form A and placing the other half on form B
2. Therefore you have 2 forms of the same test
3. To have good reliability, the scores from these tests should at least be 80% in agreement
What is the method for assessing external reliability
- Test-retest method
1. Give someone same questionnaire / interview to the same person on 2 separate occasions
2. With a sufficient gap for the chance to forget
3. If both yield same results = reliable
How do we improve low in internal reliability
Remove Qs that don’t appear to be getting consistent results
How do we improve low external reliability
- rewrite Qs so they aren’t ambiguous or leading
- train interviewer to deliver Qs the same way each time (participants might respond in certain ways depending on how it’s asked)
What are the ways of assessing internal reliability
Face / content / concurrent / construct / predictive validity
What are the ways of assessing external validity
Ecological / population / historical validity
What is the type of reliability associated with observations, how is it assessed
- inter-rater reliability
- check for high positive correlation (+0.8)
What types reliabilities are associated with self-reports, and how are they assessed
- Internal reliability: split-half method
- external reliability: test-retest method
How do we deal with low reliability in observations
- it would be low inter-rather reliability
Dealt:
—> pilot study (help clearly operationalise)
—> train the observers
What are the ways of dealing with low internal reliability
- control extraneous variables
- by: random allocation, single blind, etc….
What are the ways of dealing with low external reliability
- more realistic setting/task (different research location method)
- A more representative sample (different sampling method)
- replicating research in current day