Reliability

Question 1

Define Reliability

Answer 1

• The extent to which a method of measurement/research study produces constant findings across a situation/overtime

Question 2

What are the Types of Reliability?

Answer 2

• Inter-Rater/Observer/Interviewer Reliability
• Internal Reliability
• External Reliability

Question 3

Describe Inter-Rater/Observer/Interviewer Reliability

Answer 3

• If researchers are observing the same thing

- Consistency across researchers

Question 4

How do you Assess Inter-Rater/Observer/Interviewer Reliability?

Answer 4

• Observers record their individually
• Sets of data correlated
• Inter-Rater/Observer/Interviewer reliability achieved if highly significant positive correlations are obtained

Question 5

How can you Improve Inter-Rater/Observer/Interviewer Reliability?

Answer 5

• Pre-training of observers
• Ensuring variables involved are operationalised + observers fully understand them
• Using same video - researchers see same behavior

Question 6

What is Internal Reliability?

Answer 6

• Consistency within the test

Question 7

How do you Assess Internal Reliability?

Answer 7

• Split Half Method
• Compares one half of the test with the other half to check whether the scores are consistent
• Splitting can be done by comparing odd and even numbered questions OR comparing the first half to the second half
• Two sets are correlated
• Highly significant positive correlation indicates internal reliability/consistency

Question 8

What is External Reliability?

Answer 8

• Ability to produce the same results over time

Question 9

How do you Assess External Reliability?

Answer 9

• Test-Retest Method
• Same participant doing the same test/questionnaire on different occasions with no feedback after first test
• The time interval shouldn’t be:
  Too Short = Remember previous answers
  Too Long = Participants may have changed in some way relevant to the test
• If there is a highly significant positive correlation between scores then it will indicate that the test is reliable

Question 10

Describe Pilot Studies

Answer 10

• Used to checks method of measurement works properly
• Altering improve correlation
• Test-Retest correlation is low - reliability of test is question - researchers alter test to improve reliability - look at test itself + identify sections that don’t correlate well on 2 occasions - remove components of test that weaken its reliability + replace with alternative questions - new test again - process is repeated until a high level of reliability is shown
• take an extra measurement from each participant