observational design- research methods

Question 1

ways of recording data

Answer 1

unstructured observation- write down everything they see and ends to produce accounts of behaviour irch in detail, appropriate when observations small in scale
structured observation- simplify target behaviours that will become main focus of investigation using behaviural categories

Question 2

behavioural categories

Answer 2

break target behaviour into set of behavioural categoies
target behaviours should be defined and made observable and measurable
researcher should ensure that they have included all; the ways in whcih target behaviour may occr within their behavioural checkist

Question 3

sampling methods

Answer 3

event sampling- counting the number of times particular behaviour occurs in target individual or group
time sampling- recording behaviour within pre established time frame

Question 4

structured eval

Answer 4

+ make recording of data easier and more systematic, data produced likely to be numerical which means analysing and comparing behaviour is more straightforward

Question 5

unstructured eval

Answer 5

• produce qualitative data which mya be more difficult to record and analyse
  +benefit from more richness and depth of detail
• greater risk of observer bias as objective behavioursal categories so only record those that catch their eye not the most important or useful

Question 6

behavioural categories eval

Answer 6

+ make data collection structurd and objective,
-cateogires must be clear and unambiguous as possible, observable measurable and self evident should not require further interpretation
- should ensre all possible forms of target behaviour included in the checklist, should not be dustbin category where different behaviours deposited
-+ should be exclusive and not overlap

Question 7

sampling methods eval

Answer 7

+ event sampling useful when target behaviour happens infrequently and could be missed
- if event too complex the observer may overlook important details if using event sampling
+ time sampling effective in reducing number of observations that have to be made
- behaviour sampled may be unrepresentative of the observation as a whole