Observations

Question 1

Naturalistic/Controlled

Answer 1

NATURALISTIC:
Carried out in natural environment where behaviour takes place so high eco validity and generalisable to real life. However, hard to replicate due to low variable control.

CONTROLLED:
Recording behaviour in controlled env with variable control, so easy to replicate. However, low eco validity and low mundane realism.

Question 2

Covert/Overt

Answer 2

COVERT:
Observers not visible and participants unaware being observed so no social desirability bias, however cannot get informed consent.

OVERT:
Observer visible and participants aware they’re being researched so ethical as can get informed consent, however participants may show social desirability bias and present positive behaviour.

Question 3

Participant/Non-Participant

Answer 3

PARTICIPANT:
Researcher becomes a member of the group they’re observing so can get in depth detail on behaviour as wont miss as many behaviours, however may affect objectivity.

NON-PARTICIPANT:
Researcher remains outside of group so more objective observations, however may miss behaviours and so gain less info.

Question 4

Structured/Unstructured

Answer 4

STRUCTURED:
Researcher uses systems to organise their observations and have set behaviours to record so makes recording data easier + replicable, however important info can be missed if it can’t be put into a set category

UNSTRUCTURED:
Researcher records all behaviour seen so get in depth, detailed observations, however risk of observer bias if researcher recording behaviours that fit their aims as there may be too many behaviours to record them all

Question 5

Time Sampling

Answer 5

Design behavioural categories of anticipated behaviours and agree on a schedule of when you will observe and for how long each time e.g. every two minutes for 10 seconds

ADVANTAGE:
Reduces number of observations made so data recording is easier

DISADVANTAGE:
May be unrepresentative of observations as whole as may miss behaviours when not recording

Question 6

Event Sampling

Answer 6

Design behavioural categories of anticipated behaviours and tally each time you see an instance of the behaviour]

ADVANTAGE:
Useful for recording infrequent behaviour that is likely to be missed in an interval

DISADVANTAGE:
May be difficult to accurately record observations if too many behaviours happen at once

Question 7

Behavioural Categories

Answer 7

In observations behaviour must be operationalised into different categories e.g. aggressive behaviour could be categorised into kicking, hitting etc…

Categories should:
-Not overlap
-Be observable
-Be precise
-Cover all possible behaviours

Question 8

Inter-Observer Reliability

Answer 8

Observations should be done by 2+ researchers to ensure behaviours not missed and give more objective and unbaised observations.

1. Researchers familiarise with behavioural categories
2. Pilot
3. Discuss differences and adjust if needed
4. Conduct observations and correlate (+0.8)
5. If have recorded same behaviours there is reliability

Question 9

Validity Issues

Answer 9

Observer bias - Only recording behaviours that fit aims of study, can be reduced using double blind technique or more than 1 observer.

Issues with coding system - Not all behaviours may be covered, can be reduced by conducting a pilot and making changes, or testing the concurrent validity by comparing to a similar study that exists