L8 - Survey Methods Flashcards
1
Q
What is a survey?
A
- Descriptive method where ppts are asked questions or given statements to rate
2
Q
What do surveys measure?
A
- Anything that can be observed, evaluated and reported accurately
- They are good for measuring things that are hard to induce or observe, like attitudes
3
Q
What methods are there for surveying?
A
- Interviews: face to face; clarification and follow up questions available
- Phone surveys: questions on phone, reduces cost
- Questionnaires: written questions via the internet, post or phone
4
Q
What issues are there with surveys?
A
- Correlation & Causation: can be fixed with longitudinal studies
- Third variable: can be addressed with stats techniques
5
Q
What biases are present in psychological research?
A
- Interviewer bias: interviewer’s behaviour influences responses
- Subject bias: behaviour of subjects is influenced by their beliefs of how they are supposed to behave.
- Social desirability bias: social desirability influences behaviours
6
Q
What are the three types of probability sampling?
A
- Simple random
- Stratified
- Cluster
7
Q
What are the types of nonprobability sampling?
A
- Convenience: requests volunteers from a group who meet the requirements of the study
- Snowball: Old ppts recruit new ones
- Purposive: researcher targets a particular group
- Quota: Like stratified, but not randomised.
8
Q
What are the cognitive functions for a participant in a questionnaire?
A
- Understanding questions
- Retrieving relevant info
- Forming judgement
- Formatting judgement to fit response alternatives
- Editing answer
9
Q
What are the rules for designing survey questions?
A
- Avoid ambiguity
- No leading questions
- Don’t ask 2 questions in one
- Aim for simplicity
- Use complete sentences
- Avoid slang
10
Q
What are cognitive interviews?
A
- Tools to evaluate survey questions
- Used for eyewitness testimonies for accurate memory
- Used for pilot studies
11
Q
What is factor analysis? What is it used for?
A
- Multivariate analysis where many variables are correlated
- Allows identification of which survey questions cluster together to form a scale
12
Q
How do you measure reliability of survey questions?
A
- Inter-rater reliability: correlation between 2 researchers
- Test-retest: correlation bvetween test scores in different times
- Split half: correlation between scores of first and second half of the test.
13
Q
How do you measure validity of survey questions?
A
- Convergent: correlation between accepted measures including Concurrent validity and Predictive validity
- Discriminant: No correlation with measure of some other construct.