13 - Descriptive Strategy Flashcards
(15 cards)
What is descriptive strategy?
Measuring a set of variables as they exist naturally
Useful as preliminary research
Includes observational, survey, case study research design
What does observational use?
Behavioural observation
Observer stays unnoticed or subjects habituate
Subjectivity
To avoid subjectivity
Well-defined categories of behaviour (behaviour categories); for clear observational definitions of constructs
Trained observers
Multiple observers for inter-rater reliability
How are scores used in behavioural observation?
- Frequency method: instances in given time
- Duration: how much time spent during given time
- Interval: dividing period into intervals then recording if behaviour occurs in each interval
- And 2. Can lead to distorted measurements
How are complex situations observed?
Sampling observation
- Record
- Or sample: 1. Time: observe, pause and write; 2. Event: observe one event in first interval, a second in second; 3. Individual: one indiv at a time per interval
What is content analysis?
Archival research?
- Using behavioural observation in books, tv, media
2. Measuring behaviours in historical records
Ethologist observation types
Nonhumans
1. Naturalistic or nonparticipant - Goodall - external validity - subjective interpretation
2. Participant - researcher participates in same activities - external validity - observer influence
3. Contrived or structured - arranged settings in lab - Piaget - bird house - less natural
Ethics - spying?
How are surveys put together?
- Questions developed
- Organize questions
- Participant selection
- How to distribute/administer
Advantages of open-ended questions
Flexibility in answer
Diff perspectives impossible to compare
Can’t use stats
Limited by willingness to express thoughts
Restricted questions
Rating-scale questions
- Limited and predetermined set of responses
- Likert five-point with equal spacing
A. Ppl avoid categories at extremes
B. More than 10 causes trouble
Ends are anchors and named
Bad: when all questions have same choices for response, ppl use same response for all (response set)
To fix: mix positive with negative statements
Good: can be treated as interval scales and used in stats
Easy to understand
Semantic differential scale
Bipolar adjectives
Good Survey design
Demographics at end
Sensitive in middle
Similar questions grouped
Appropriate language
Mail nonresponse bias
Indivs who return mail surveys are not representative of entire group who receives them
Fix with:
1. Cover letter
2. Gift
Phone interviewer bias
Tone of voice
Leading questions
Case study
Idiographic approach
Nomothetic approach
- Study of a single individual
May involve treatment
Non-treatment is case history - Study of indiv for clinical research
- Study of groups for clinical research
Good: Unusual cases, exceptions to rule, powerful
Bad: only describes, lacks internal/ext validity, selective and observational bias
