Reliability

Question 1

Reliability

Answer 1

-Saying that results are reliable is another way of saying the results are consistent
-If the researcher replicated the exact experiment (e.g at a different time or with different participants) the results would be the same

Question 2

External reliability

Answer 2

-External reliability is how consistent a measure is when repeated (e.g is the results stay consistent when conducting an exact replication of a study with different participants and/or at a different time)

Question 3

Internal reliability

Answer 3

-Internal reliability is the extent to which different measures within a study are consistent with themselves (e.g If all the questions in a questionnaire are measuring the same thing)

Question 4

Assessing internal reliability (split -half method)

Answer 4

1.Researchers use a questionnaire and split the test in half
2.They have participants complete both halves of the test
3.They then test the strength of the correlation between the two measures using a test of correlation (e.g spearman’s rho or Pearson’s)
4. A strong correlation indicates internal reliability

Question 5

Assessing external reliability (test- retest)

Answer 5

-Test retest tests external reliability by conducting an exact replication of a study with the same variables (at a different time and/or) with different participants
-They then use a test of correlation, a strong correlation indicates high external reliability

Question 6

Assessing external reliability (inter rater reliability)

Answer 6

-Inter-rater reliability is where two trained observers independently observe behaviors using the same behavioural categories and then compare their independently produced results by using a correlational test (e.g spearman’s rho or Pearson’s)
-A strong correlation (0.8 or stronger) indicates a high external reliability

Question 7

How to improve reliability in observations

Answer 7

-Improve the training of the observe will increase their accuracy when assigning particular behaviours to the correct behavioural categories
-Additionally, pilot studies help to identify poorly defined behavioural categories

Question 8

How to improve reliability in interviews

Answer 8

-Use structured interviews rather than unstructured interviews as this provides the researcher with a set script to follow during the interview
-This means that each participant has a similar experience and provides answers that can easily be compared with answers from other participants

Question 9

How to improve the reliability of questionnaires

Answer 9

-Use closed questions to limit the amount of participant responses
-If there is an established questionnaire that tests for what the researcher is choosing to measure, the researcher should use this test

Question 10

How to improve the reliability of experiments

Answer 10

-Researchers should ensure that the same standardised procedures are used, to give the participants all similar experiences (e.g speaking the same way to each participant and keeping environmental conditions the same)