Reliability

Question 1

What is reliability?

Answer 1

If the research is consistent and able to be repeated
e.g measure the height of a chair one day and expecting the same result the next day due to the ruler being consistent and if there was any change then it should be because of the chair

Question 2

What is validity?

Answer 2

Validity is whether the observed effect is a genuine one

Question 3

Reliability of observations

Answer 3

• in observation research it should always be consistent and two observers watching the same behaviour should produce exactly the same results and we refer to the extent of the way they agree as inter-rater reliability
• measured by correlating the observations of two or more observers - if there is more than 80% agreement then the data has inter observer reliability

Question 4

How to improve reliability in observations

Answer 4

• observers should be trained in the use of a coding system/behaviour checklist

Question 5

Internal reliability

Answer 5

A measure of the extent to which something is consistent within itself
e.g all the questions in an IQ test should be measuring the same thing
- may not be relevant to all questionnaires because internal consistency is not important

Question 6

External reliability

Answer 6

The extent to which a measure varies from one occasion to another
e.g if an interviewer conducted an interview and then conducted the same interview with the same interviewee a week later then the outcome should be the same otherwise the interview is not reliable

Question 7

What is used to test external and internal reliability?

Answer 7

Split-half method- used to assess internal reliability
- one group of participants is given a test once
- the participants answers are then divided in half and compared
- the individuals scores on both halves should be very similar

Test-retest method- used to assess external reliability
- a group of participants is given a test or questionnaire or interview once
- then once the participants have had a chance to forget it they do it again
- the answers can be compared and should be the same
- same test must be given

Question 8

How to deal with low reliability

Answer 8

• internal — questions can be removed to see if the split half test returns a high reliability score
• external — may have poorly written questions that are causing confusion and need to be rewritten
• if an interview has low reliability then the interviewer may need to be retrained in order to be more consistent