Reliability and Validity

Question 1

Define reliability

Answer 1

The extent to which a test produces consistent results. Reliability can be internal(within itself) or external (over several ocassions)

Question 2

What are five issues which can affect reliability and how can they be overcome?

Answer 2

1. Only one researcher=Involve more than one
2. Research was only conducted once=Repeat the same study using the same methodology
3. Instructions not given in the same way=Standardise the procedure
4. Variables not clearly defined=Operationalised
5. P’s not asked Q’s in the same way=Standardise

Question 3

What are the three ways of measuring reliability?

Give a basic description of each

Answer 3

• Test-retest=P’s are give the same test twice
• Split half=split the test in half and compare the scores
• Inter-rater=Two or more researchers interpret the same behaviour and compare the results

Question 4

Define validity

Answer 4

The extent to which a test accurately meauses what it claims to measure. This can be internal (within the study) or external (outside the study)

Question 5

What are the three validity issues?

Give a basic description of each

Answer 5

• Researcher Bias=Researcher directly or inderctly influences the results of a study
• Demand Characteristics=Participants unconsciously work out the aim and act differently
• S.D.B=participants change their behaviour in a particular way which will show them in the best positive light

Question 6

How can validity issues be overcome?

Answer 6

• Double blind procedure (research and particpant don’t know who’s in each condition)=Resarcher bias
• Give particpants anonymity and confidentiality=S.D.B
• Make the aim difficult to guess and double blind=demand characteristics

Question 7

What are the five different types of ways assessing validity?

Answer 7

• Predicitive
• Content
• Face
• Construct
• Concurrent

Question 8

What is predictive and content validity?

How can each be assured to be high

Answer 8

• Predictive=Degree to which a test can accurately be used to forecast future behaviour (compare results with other measures)
• Content=Does the method used actually seem to measure what it is intended to (a panel of experts assess)

Question 9

What is concurrent and face validity?

How can we assures each validity is high

Answer 9

• Concurrent=How well does the measure agree with existing measures (test the participant with both the new test and the establisged test)
• Face=On the face of it, is this a good measure (simply look at the test you are using and determine where or not it seems to measure what you want)

Question 10

What is construct validity?

How can we assure that it is high

Answer 10

Is the method actually measuing all the parts of what we are aiming to test (need to define what we are measuing and ensure that all the parts of the definition are being measured)