Reliability

Question 1

What is reliability?

Answer 1

A measure of consistency.

Question 2

What is internal reliability?

Answer 2

Assesses the consistency of results across items within a test / experiment.

Question 3

What is external reliability?

Answer 3

The extent to which a measure varies from one use to another.

Question 4

What are 2 methods of testing reliability?

Answer 4

Test-retest method

Split-half method

Question 5

What is the test-retest method?

Answer 5

The same pps are asked to undertake the research measure, E.G. a
questionnaire, on different occasions.

Question 6

What are some cautions that need to be considered with the test-retest method?

Answer 6

= Be aware of any potential demand characteristics =
E.G.
If same task/measure given twice in 1 day, there’s a strong chance that pps will be able to recall responses they gave in the first test, and so psychologists could be testing their memory rather than the reliability of their measure.

= Time frame =
Make sure that there’s not too much time between each test.
E.G.
If measuring depression, and you question the participants a year apart, they may have recovered in that time and give completely different responses rather than the questionnaire being unreliable.

Question 7

How are the scores of a measure correlated?

Answer 7

1 is perfect correlation.

A score of +0.8 correlation is significant and reliable.
A score of below 0.8 is unreliable.

The closer to 1, the more reliable.

Question 8

What is inter-observer reliability?

Answer 8

When 2 or more observers are recording behaviour in a consistent way and collect similar results.

Reduces subjectivity / bias.
Also ensures behavioural categories are being used correctly.

Question 9

What are some issues of reliability?

Answer 9

Lack of standardised procedures =
Investigator, task, environment and measuring tool must all be the same or the results may show inconsistencies. Being in a different environment with a different interviewer may change the pps feelings / opinions which causes results to differ from the original.

Measuring tool may be unreliable =
The test itself may be unreliable which leads to inconsistent results.

Extraneous variables =
EVs such as the temperature of the room, noise levels…etc must be controlled as they might affect pps behaviours and feelings in the moment and lead to inconsistencies.

Question 10

What is the split-half method?

Answer 10

Test is split in two and the scores for each half of test is compared with one another to measure consistency.

If test is consistent it leads researcher to believe that it’s all measuring the same thing.

Question 11

How can you improve the reliability using questionnaires?

Answer 11

Questionnaires =

• Test-retest method - shows whether questions are reliable if similar results obtained.

(when failed test-retest, you adapt and change questions to make them clearer)

Question 12

How can you improve the reliability of interviews?

Answer 12

Interviews =

• Having the same interviewer for each participants.
• Using a structured interview can also improve reliability.

Question 13

How can you improve the reliability of experiments?

Answer 13

Use lab experiments =

• Have good control over the procedures to ensure the reliability (e.g. same testing room/researcher etc)

Question 14

How can you improve the reliability of observations?

Answer 14

Use controlled / structured observations =

• Categories should be measurable and clear / operationalised (e.g. pushing is clearer than aggressive)
• Have more than 1 observer and do an inter-observer reliability check.