Reliability

Question 1

reliability

Answer 1

how consistent the findings from an investigation are

Question 2

importance of reliability

Answer 2

ensures DV is measured accurately
ensures the outcome is the same over time
ensures all conclusions are made accurate

Question 3

valid results =

Answer 3

reliable results
but reliable results doesn’t equal valid results

Question 4

internal reliability

Answer 4

describes how consistent something is within itself

Question 5

external reliability

Answer 5

when consistent results are produced regardless of when the investigation is used/ who administers it

Question 6

assessing the reliability of observations

Answer 6

inter-observer reliability (2+ observers)
intra-observer reliability (same observer)
then compare and analyse against a correlation coefficient of 0.8

Question 7

improving reliability of observations

Answer 7

if the correlation coefficient is below 0.8
operationalise behavioural categories - make them unambiguous, don’t overlap
researchers give the observers a briefing/training/practice - to ensure the observers are giving sufficient observations

Question 8

assessing the reliability of questionnaires

Answer 8

test-retest method - researcher gives the same test to the same pp twice after short interval, questionnaire is reliable if correlation coefficient is 0.8

Question 9

improving the reliability of questionnaires

Answer 9

make questions closed (not open) so there is no range for interpretation
make questions unambiguous, clear, understandable

Question 10

assessing reliability of interviews

Answer 10

inter-interviewer reliability (2+ interviewers)
intra-interviewer reliability (same interviewer)
then compare and analyse against a correlation coefficient of 0.8

Question 11

improving the reliability of interviews

Answer 11

make questions unambiguous, clear, understandable
interviewer should be formal, calm, unresponsive, consistent language, unopinionised, well-presented (body language), consistent tone

Question 12

assessing the reliability of experiments

Answer 12

split half method: randomly selected half of group test one condition, other half test the other condition
then compare and analyse against correlation coefficient of 0.8

Question 13

improving reliability of experiments

Answer 13

standardisation - using strict control over many aspects of the procedure (instructions)