Reliability

Question 1

Two types of reliability:

Answer 1

• external
• inter-observer reliability

Question 2

Reliability (what is it?)

Answer 2

Reliability refers to whether something is consistent
- researchers should be able to measure or observe something time and time again and get similar (reliable) results

Question 3

External reliability (what is it?)

Answer 3

Consistency between tests - the ability to produce similar results between different occurences of a research

Question 4

How to measure and assess external reliability?

Answer 4

MEASURE: Test-retest method
- take the same group of ppts, make them repeat the same test twice

ASSESS: Statistical analysis to compare scores from first and second time.
- Spearman’s Rho statistical test - 0.8+ strong positive correlation or above -> shows the results are reliable

Question 5

Inter-observer Reliability (what is it?)

Answer 5

• consistency between observers
• the ability to produce similar results between different observers

Question 6

How to measure and assess Inter-observer Reliability?

Answer 6

• multiple observers need to observe the same ppts at the same time from the same place
• need the same behaviour catagories
• observation is done independantly from each other
• compare the tallies/frequencies across the different observers
• perform a correlation analysis like Spearman’s Rho. Gives us a gage on the similarity of observations
• look for a strong positive correlation of at least 0.8
• high-inter observer reliability (0.8-1)

Question 7

Improving reliability in lab experiments:

Answer 7

• standaridsed procedures
• standardised instructions
• high control over variables + EVs
• pilot study
• operationalise variables -> specific and easier to reproduce the study

Question 8

Improving reliability in interviews:

Answer 8

• consistency - keep the interviewer the same
• train interviewers on how to ask questions like tone of voice
• use structured questions

Question 9

Improving reliability in questionnaires:

Answer 9

• use closed questions for fixed responses
• avoid using ambiguous questions

Question 10

Improving reliability in observations:

Answer 10

• behaviour catagories - operationalised
• high control over variables
• standardised procedure
• prior to study - train observers on behaviour catagories and what would count as an example