Reliability

Question 1

Define reliability

Answer 1

Refers to how consistent the findings are from an investigation or measuring device

Question 2

How do u asses reliability?

Answer 2

Test-retest method and Inter-observer/rater reliability

Question 3

Test-retest

Answer 3

-most common (questionnaires, surveys, interviews)
-same participants take test on different occasions
-high correlation between test scored indicates test scores r similar + has good rel (rho + r tests)

Question 4

What tests can be used for correlation in test retest?

Answer 4

Spearman’s rho + Pearson’s

Question 5

Why is the timing of the test-retest crucial?

Answer 5

There must be sufficient time between to ensure p’s cannot recall answers to qs in survey but not long that thoughts/feelings/attitudes may have changed

Question 6

Why is inter-obs an issue?

Answer 6

Consistency between different researchers working on a study is important for reasearch as subjectivity bias isn’t wanted

Question 7

Can you have high rel but not validity?

Answer 7

Yes

Question 8

Can u have validity without reliability?

Answer 8

No

Question 9

How do u improve reliability? For questionnaires?

Answer 9

Replace open qs with closed ones = fixed choice which may be less ambiguous (clear)

Question 10

Improve reliability - interviews

Answer 10

Use same interviewer each time
All interviewers must be trained well
Unstructured = free flowing which reduced reliability

Question 11

Improve reliability experiments

Answer 11

Ensure procedures r consistent every time
Standardised procedures

Question 12

Improve reliability observations

Answer 12

Ensure behavioural categories have been operationalised/measurable
Categories shouldn’t overlap