3- reliability

Question 1

how can reliability be assessed in observations

Answer 1

1. test retest
   - repeat 2nd time with same participants and compare
   - should be similar and have correlation coefficient>0.8
2. inter observer reliability
   - use more then one observer to reduce bias
   - observer and record separately
   - should be similar and have correlation coefficient>0.8
3. pilot study
   - small trial run before main research
   - improves precision
   - may include standardised instructions, debriefing, and planning procedures properly to minis human error/variation
   - operationalise behavioural categories
   - make sure observers are trained to observe

Question 2

how to improve reliability in observations

Answer 2

1. if low inter observer reliability then operationalise behavioural categories and ensure observers understand categories and may need further training
2. if pilot study in unclear, then give more training to observers so they are familiar with categories and can respond quicker

Question 3

how can reliability be assessed in self-reports/questionnaires

Answer 3

1. test retest
   - self-report given to participants twice with interval of time between tests and compared
   - should be similar and have correlation coefficient>0.8
2. inter researcher reliability
   - more than one interviewer to reduce bias
   - record answers separately
   - need to act in similar/consistent ways
   - compared and should be similar and have correlation coefficient>0.8

Question 4

how to improve reliability in self-reports/questionnaires

Answer 4

1. make sure interview questions aren’t ambiguous and clear so the participants understand them and can give same answers if asked in the future (might need to rewrite or remove unclear questions)

Question 5

how can reliability be assessed in experiments

Answer 5

1. test retest
   - conduct experiment twice with interval of time between with same participants tested in the same way
   - compare results
   - should be similar and have correlation coefficient>0.8

Question 6

how to improve reliability in experiments

Answer 6

1. standardised instructions
   - procedures repeated twice so should be exactly the same for each participant
   - standardised instructions should be used and key concepts and variables should be operationalised