reliability and validity Flashcards
what is reliability ?
how consistent the findings from a study or measuring device is
internal reliability?
refers to the consistency of the items within the test
external reliability?
refers to the extent of which a measure varies from one use to another. if a measured a persons social media today would it be the same next week?
issues of reliability ?
- operationalisation of variables = not clearly stating how you’re defining and measuring your variables making replication impossible
- lack of standardised procedures = could affect consistency of research
how to deal with reliability issues ?
- operationalisation of variables = ensure test is reliable through pilot test, take any questions that may confuse people
- lack of standardise procedures = use standardise instructions that are written down and strictly follow them, treat every PT exactly the same way
3 ways to assess or check reliability ?
- split half reliability
- test retest reliability
- inter observer reliability assessment
split half reliability ?
tests internal reliability by finding a correlation between scores on the first half of the test and the second half. if theres good internal reliability then there shoddy be a strong positive relationship. >0.8
test retest reliability ?
tests external validity reliability finding if theres a correlation between scores of the same PT on a test when done on 2 separate occasions, scores would be added to a separate spreadsheet. >0.8 is strong
inter observer reliability assessment ?
to check the consistency of observers we find if theres a correlation between the results of 2 observers who have observed the same behaviour. 2 observers to record behaviour using event sampling.
validity ?
does the study measure what it intends to, is it true and accurate ?
internal validity ?
accuracy of the measurement tool whether the IV has caused DV to change rather than other variables
external validity ?
whether the findings can be generalised outside of the research study
examples of external validity?
population and ecological validity
issues with with validity?
- researcher bias
- demand characteristics
- social desirability bias
how to deal with issues of validity ?
- researcher bias = follow standardise procedures, double blind procedure
- demand characteristics = use single blind procedures, use red herrings questions
- social desirable bias = ensure confidentiality and don’t ask for names
5 ways of assessing or check validity?
- face validity
- content validity
- concurrent validity
- predictive validity
- construct validity
face validity ?
ask PT in the filed to rate each question on the social media question and rank how relevant
content validity ?
ask an expert in the filed to examine your measure to see if they think its measuring what its supposed to
concurrent validity ?
assessing validity by comparing results with another measure
predictive validity?
does the results match the predicted scores that were expected, can the test predict behaviour?
construct validity ?
the extent to which a test measures what its supposed to, to assess we operationalise.
