How consistency a method measures over time when repeated
Issues of internal reliability
More than one researcher:
inconsistency in how they carry out the research or interpret behaviour
More than one measure:
if there is morethanonetest or measure, are they consistent in the difficulty
Within the measure:
consistency across questions,tasks or words in a memory test
Dealing with issues of internal reliability
More than one researcher:
operationalise variables - less room for interpretations
standardised procedures
detailed training of researchers
Multiple measures:
standardised measures
Within measure:
standardised each question or item
Issues of external reliability
Behaviour only measured once:
test may not measure behaviour consistently at another time
Dealing with issues of external reliability
ensure tests are designed in relation to measuring behaviours, attitudes, etc beyond one setting or one-time
test-retest the same participants
Assessing internal reliability
More than one researcher = inter-rater/inter-observer
observations by having two people conduct an observation independently using the same measures and compare results
should be an 80% agreement
Within measure = split-half
comparing the results if half the items in a test with the other half of the items in a test
Assessing external reliability
Are findings consistent at other times? = test-retest
comparing the results of the research conducted on one occasion with the results of the research conducted on a later occasion
uses the same participants
cannot be used if participants did not know they were taking part or the situation is unique and cannot be replicated as same participants cannot be found
