Research methods-
Observation- 1st method
Observational study=researcher watches/listens to participants- engages + records- observing
behaviour that is being studied
3 types.
Naturalistic- behaviour in its natural setting- observer making no attempt to influence the
behaviour.
e.g., Bullying.
- Participant- observer actively involved. Observe what they say and do- informal,
unstructured, and unplanned.
e.g. JS- SKIN COLOUR.
- Controlled- researcher intervenes and manipulates aspects; behaviour is environmental. -
aspects of the environments are controlled.
e.g., AB 1977- BBD experiment.
strengths and weaknesses.
Types of observation STRENGTH WEAKNESSES
Structured observation - Quantitative data - Events may not it in
researcher designs - More reliable categories
type of coding - Less time - Lack of validity
scheme to record consuming - Only useful studying
behaviour- small scale interactions.
behaviour broken
up into
*behavioural
categories-
quantitative data.
Unstructured observation - Rich qualitative - Too much too record
researcher records data - Behaviours recorded-
ALL behaviour they - Unexpected those visible to the eye-
can see. behaviours- still may not be relevant
included. behaviours.
- No BC – everything - Qualitative data- difficult
recorded. to analyse
- Hard to repeat- not that
reliable.
Covert observation - High validity- - Unethical- participants
participants are legitimacy- cannot give consent.
not aware that unaware natural - Stress- have to remain
they are being setting. under suspicion
studied- - Showing - Can’t take notes
, naturalistic
behaviour- free
from social
desirability bias
and demand
characteristics.
Overt observation - No ethical issue - Social desirability bias=
aware that they - Ask Q + make notes put themselves in a good
are being observed - Systematic light.
- Demand
characteristics= showing
behaviour that they think
researcher wants to see.
- Time consuming
Participant observation - Gain insight - Covert participation=
actively involved - Validity- legitimacy- deceiving people
rich qualitative data - Time consuming
- Access to suspects - Bias- get too involved-
- builds rapport= start to sympathise
more trust and - Loose
comfort- objectivity=interpretation
participants is biased- seeing only
behaving naturally. from participants view.
Non-participant - Open minded- - Covert- unethical
observation outsider - Less insight- separate
separate from - No sympathy- less - Hawthorne effect
people being likely to be bias (people act diff when
observed. - No ethical issues if being watched).
overt. - Time consuming
- Remain objective - Lack of trust with
- More reliable participants=researcher
- No observer bias has less insight-
participants do not
behave naturally.
operationalised
*Behavioural categories- dividing a target behaviour (e.g., stress or aggression) into a subset of
specific and behaviours.
Reliability –
- The consistency of research
- If results remain same or similar over many attempts- reliable
2 types
Internal- test/measure consistent within itself
External- extent to which a test produces consistent results over several occasions
, issues of reliability;
Issue How to deal
Lack of operationalism of the variable- BC
if you don’t clearly state- replication impossible
Order effects Counter balancing – splitting into 2 groups and
reversed order
Lack of standardised procedures – Use standardised procedure and scientific
inconsistency of measuring tools & measuring equipment- MRI scans
uncontrolled environment. Conducting research in lab
*BC- behavioural categories
*Operationalism- = ensuring variables are in a form that can be easily tested.
*Order effects- the order of the conditions having an effect on the participants behaviour- practice
perfect/ boredom.
Inter-rater reliability= the extent to which there is an agreement between two or more
observers involved in observations of a behaviour.
80%
Validity –
- Legitimacy
- Kelly (1927)- if a test is valid, it measures what it claims to measure.
2 types
Internal- the study measures what it intends to measure
External - whether the study paints a true picture of real-life behaviours (mundane realism)
& Would apply to different places, different times, or different people (population validity).
Issue How to deal
Researcher bias Double blind procedure
- Researcher expectations influence the Involves neither researcher nor participant
research Reduces validity knowing true aims of study
Demand characteristics - double blind procedure reduces
demand characteristics
single blind procedure
participants being unaware for the researcher
Observation- 1st method
Observational study=researcher watches/listens to participants- engages + records- observing
behaviour that is being studied
3 types.
Naturalistic- behaviour in its natural setting- observer making no attempt to influence the
behaviour.
e.g., Bullying.
- Participant- observer actively involved. Observe what they say and do- informal,
unstructured, and unplanned.
e.g. JS- SKIN COLOUR.
- Controlled- researcher intervenes and manipulates aspects; behaviour is environmental. -
aspects of the environments are controlled.
e.g., AB 1977- BBD experiment.
strengths and weaknesses.
Types of observation STRENGTH WEAKNESSES
Structured observation - Quantitative data - Events may not it in
researcher designs - More reliable categories
type of coding - Less time - Lack of validity
scheme to record consuming - Only useful studying
behaviour- small scale interactions.
behaviour broken
up into
*behavioural
categories-
quantitative data.
Unstructured observation - Rich qualitative - Too much too record
researcher records data - Behaviours recorded-
ALL behaviour they - Unexpected those visible to the eye-
can see. behaviours- still may not be relevant
included. behaviours.
- No BC – everything - Qualitative data- difficult
recorded. to analyse
- Hard to repeat- not that
reliable.
Covert observation - High validity- - Unethical- participants
participants are legitimacy- cannot give consent.
not aware that unaware natural - Stress- have to remain
they are being setting. under suspicion
studied- - Showing - Can’t take notes
, naturalistic
behaviour- free
from social
desirability bias
and demand
characteristics.
Overt observation - No ethical issue - Social desirability bias=
aware that they - Ask Q + make notes put themselves in a good
are being observed - Systematic light.
- Demand
characteristics= showing
behaviour that they think
researcher wants to see.
- Time consuming
Participant observation - Gain insight - Covert participation=
actively involved - Validity- legitimacy- deceiving people
rich qualitative data - Time consuming
- Access to suspects - Bias- get too involved-
- builds rapport= start to sympathise
more trust and - Loose
comfort- objectivity=interpretation
participants is biased- seeing only
behaving naturally. from participants view.
Non-participant - Open minded- - Covert- unethical
observation outsider - Less insight- separate
separate from - No sympathy- less - Hawthorne effect
people being likely to be bias (people act diff when
observed. - No ethical issues if being watched).
overt. - Time consuming
- Remain objective - Lack of trust with
- More reliable participants=researcher
- No observer bias has less insight-
participants do not
behave naturally.
operationalised
*Behavioural categories- dividing a target behaviour (e.g., stress or aggression) into a subset of
specific and behaviours.
Reliability –
- The consistency of research
- If results remain same or similar over many attempts- reliable
2 types
Internal- test/measure consistent within itself
External- extent to which a test produces consistent results over several occasions
, issues of reliability;
Issue How to deal
Lack of operationalism of the variable- BC
if you don’t clearly state- replication impossible
Order effects Counter balancing – splitting into 2 groups and
reversed order
Lack of standardised procedures – Use standardised procedure and scientific
inconsistency of measuring tools & measuring equipment- MRI scans
uncontrolled environment. Conducting research in lab
*BC- behavioural categories
*Operationalism- = ensuring variables are in a form that can be easily tested.
*Order effects- the order of the conditions having an effect on the participants behaviour- practice
perfect/ boredom.
Inter-rater reliability= the extent to which there is an agreement between two or more
observers involved in observations of a behaviour.
80%
Validity –
- Legitimacy
- Kelly (1927)- if a test is valid, it measures what it claims to measure.
2 types
Internal- the study measures what it intends to measure
External - whether the study paints a true picture of real-life behaviours (mundane realism)
& Would apply to different places, different times, or different people (population validity).
Issue How to deal
Researcher bias Double blind procedure
- Researcher expectations influence the Involves neither researcher nor participant
research Reduces validity knowing true aims of study
Demand characteristics - double blind procedure reduces
demand characteristics
single blind procedure
participants being unaware for the researcher
