reliability- research methods Flashcards
reliability
measure of consistency
if measurement made twice and produces the same result then measurement described as being reliable
test retest
administering same test or questionnaire to the same person or different occasions
if test reliable results should be the same or similar each time theyre adminstered
sufficient time between test and retest to ensure that pp cant recall their answers but not so long that their attitudes opinions or abilites may have changed
inter observer reliability
one observers interpretation may differ from someone elses intro subjectivity bias and unreliability into the data collection process
observers should not do it alone but conduct their observations in teams of at east two
involve pilot study to check that observers applying behavioural categories in the same way or comparison reported at the end of study
need to watch the same event or sequence of events but record their data independently
dtaa collected should be correlated to assess reliability
measuring reliability
correlation coefficient- +.80 for reliability
questionnaires (imroving reliability)
should use test retest
comparing data should produce correlation exceeds +.80
low test retest reuire some items to be deselected or rewritten
may be interpreted differently by the same person on different occasions
relace some of the open questions with close fixed choice alternatives which may be less ambiguous
interviews (improving reliability)
use same interviewer each time
all interviewers properly trained so one interviewer not asking leading questions or ambiguous
more avoided in structured interviews where interviewers behaviour controlld by the fixed questions
unstructured are more free flowing and less likely to be reliable
observations (improving reliabiity)
behavioural categories have been operationalised and that they are measurable and self evident
categories should not overlap and all behaviours should be covered on the checklist
if categories not operationalised well or overlapping or absent differnet observers have to make their own judgemetns of hat to record where and end up with differing inconsistent records
observers may need further ftraining in using behavioural categories and may wish discuss their decisions with each other so they can apply their categories more consistently
experiments (imprving reliability)
procedures focus of reliability
compare performance of different pps the procedures must be the same every time
experimenter concerned about standardised procedures