Topic 8 - Participant Observation and MIC Flashcards
1
Q
Observations
A
- Participant is used more often than non participant
- Liked by interpretivists
- Positivists sometimes use non participant observation: uses a structured observation schedule -a predetermined list of the types of behaviour the sociologist is interested in. Produces quantitative data
- Most sociologists use unstructured participant observation - offers insight into a groups way of life and so is mainly used by interpretivists
2
Q
Issues with getting in
A
- Some groups are easier to get into than others
- Researcher has to overcome suspicion and gain trust
- Reserachers age, gender or ethnicity may prove an obstacle if different from the group
- Should not disrput groups normal behaviour - not always possible to take a role that is non disruptive and a good vantage point
3
Q
Issues with staying in
A
- Reseracher must be involved in the group but must also be detatched to remain objective and unbiased
- Too detatched = risk of not understanding
- Too involved = risk going native
- Need to strike a balance
- The longer a researcher spends with the group the less strange its ways will appear; observer becomes less observant
- William F Whyte - ‘I started as a non participating observer and foud myself becoming a non observing participant’
4
Q
Issues with getting out
A
- Usually less of a problem
- Leaving the group you have become close to can be difficult as well as re-entering the ‘normal world’
- Loyalty to the group may prevent researchers from fully disclosing everything they have learnt
5
Q
Practical issues
A
- Gives us insight to other peoples lives
- PO allows us sociologists to gain verstehen through first hand experience
- PO produces large amounts of rich, detailed, qualitative data
- May be only suitable method for accessing and studying certain groups
- PO is flexible in comparison to survey methods
- Whyte: ‘I learned answers to questions that I would not have the nerve to ask if i had been using interviews’
- Polsky: ‘Initially keep your eyes and ears open but keep your mouth shut’
6
Q
Overt
A
- Researcher can behave normally
- Dont need special knowledge or personal characteristics to join
- Group may refuse to let outsider in/prevent them witnessing certain activities
- Can ask naiive questions
- Can take notes openly
- Can use interviews or other methods to check insights
- Can opt out of any dangerous/illegal activities
- Risks creating Hawthorne Effect
7
Q
Covert
A
- Must keep up an act
- May need detailed knowledge of groups way of life
- Might be only way of getting info
- Can’t ask naiive questions - could blow cover
- Has to rely on memory
- Cant combine observation with any other methods
- Might have to engage in dangerous/illegal activities to maintain cover
- Less risk of altering groups behaviour
8
Q
Ethical issues
A
- Unethical to deceive people
- Unethical to lie about why they are leaving the group
- May have to participate in immoral/illegal activities
- Overt PO avoids these problems
- PO leads to personal attatchment to the group so researcher risks going native
- NPO avoids these problems but NPO involved ‘spying’ on people without their knowledge and consent
9
Q
Theoretical Issues: Interpretivism
A
- Produces qualitative data - liked by interpretivists
- Produces detailed and authentic pics of actors’ meanings: valid, flexible and grounded theory
- Sociologists have high level of involvement in PO - enables deep, subjective understanding of their meanings thus producing valid, insightful and qualitative data
- Flexibility produces valid data
- Glaser and Strauss - being able to enter reseacrh without fixed hypothesis researcher can develop ideas during research to produce grounded theory
10
Q
Theoretical issues and positivism
A
- Reject the use of PO due to its unscientific method
- Lack of representativeness
- \lack of reliability
- Bias and lack of objectivity
- Lack of validity
11
Q
Representativeness
A
- Groups are usually very small
- Sample is often collected haphazardly therefore group studied turns out to be unrepresentative = doesnt allow generalisations to be made
- Downes and Rock - although PO may provide valid insights, it’s doubtful for how the ‘internally valid’ insights are ‘externall valid’
12
Q
Reliability
A
- Not a standardised, scientific measuring instrument
- Success depends on personal skills/characteristics
- Therefore impossible for other researchers to check original study by replicating it
- Comparisons are difficuly due to qualitative nature of data produced
13
Q
Bias and lack of objectivity
A
- Researchers close involvement = lack of objectivity
- Involvement risks going native
- PO appeals to sociologists who sympathise with the underdog - may be biased in favour of subjects viewpoint
14
Q
Lack of validity
A
- Positivists reject that PO produces valid data
- Finsings are biased subjective impressions of observer
- Observer selects facts they think are worth recording - these are likely to fit in with their own values and prejudices
15
Q
The Hawthorne Effect
A
- Undermines the validity of PO - observers presence may make subjects act diffrently
- Bigger issue for overt observation
- Interpretivists - over time group generally gets used to observers presence and behave normally
- Reseacher can try to adopt a less obtrusive role to minimise the threat to validity
16
Q
Structure vs Action perspectives
A
- PO normall associated with ‘action’ perspectives like interactionism
- Action perspectives see society as being constructed
17
Q
Methods in Context
A
Stuctured observations
- Flanders (1970)
- Sara Dalamont
Untructured observations
- Wright (1992)
- Ball (1993)
- Hammersky
- Lacey
- J.W.B Douglas
- Leon Feinstein (2003)
18
Q
Structured observations
A
- Positivists favour structured observations as they enable them to identify and make quantitative measurements of behaviour
- Usually NPO
19
Q
Practical issues of structured observations
A
Flanders system of interaction analysis categorised (FIAC) - was used to measure pupil-pupil and pupil-teacher interaction quantitatively.
- Observer uses standard chart to recrod interacton at 3 second interviews, placing eac observation in on of 10 pre-defined behaviour categories
- Flanders found in typical American classrooms, 68% of the time is taken up by teachers talking, 20% pupils and 12% silence
- Quicker, cheaper and require less training than unstructured methods
20
Q
Reliability
A
- Techniques such as FIAC are likely to be easily replicated - only uses 10 categories of classroom interaction, easy to apply in standardised way
- Generates quantitative data - easily comparable
21
Q
Validity
A
- Interpretivists criticise it for lack of validity
- Sara Dalamont - simply counting classroom behaviour and classifying it into limited number of pre defined categories ignores meanings that pupils and teachers attatch to it
22
Q
Unstructured observation
A
- Interpretivists favour it because it’s qualitative and flexible
- Access to meanings teachers and pupils give to situations
- Doesnt make assumptions in advance about what the ey research issues will be
- Used more often
23
Q
Practical issues of unstructured
A
- Schools are complex place and more time consuming to observe, however, may be easier to gain permission to observe lessons than to do interviews with teachers and pupils
- Wright: carried out her research - few black teachers, found that African Carribean ethnicity produced antagonistic reactions
- Hammersky: notes down staffroom conversations had to be done covertly. He noted that he may have made mistakes as he relied on on his own interpretation of what was said
24
Q
Ethical issues
A
- Pupils greater vulnerability and limited ability to give informed consent means it has to be done covertly
- Can be argued researcher is obliged to report wrongdoing, however doing so may break trust between pupil and researcher
- Dalamont found given the harm that can be done to pupils, teachers and schools additional care should be taken to protect their identity
25
Validity
- Interpretivists - main strength of PO is validity, however, difference between young people and adults is a barrier to uncovering real attitudes and behaviour - may result in false image = undermines validity
- Teachers may be skilled at disguising feelings and altering behaviour; results in sociologists data lacking validity
- Language from pupil may differnt from the researcher - difficult for researchers to be certain that they understand pupils meanings
26
The Hawthorne effect
- Few cover roles researcher can adopt = difficult to carry out observation; means it has to be covert = difficult to avoid the hawthorne effect
- Difficult to reduce effect of researchers presence
- Ball - asked what did the children make of 'the tall man hiding in the Wendy house?' - childrens awareness of King's presence may have changed their normal behaviour = undermines validity
27
Representativeness
- Scale of educational system is vast, by contrast, most observational studies focus on a small number in single schools
- Small scale study results form the fact that it takes time to become familiar with setting, gain trust and carry out observation
- Small scale means obtaining schools interaction is unlikely to produce representative data
- Hammersky - data collected in staff room were more open to sample bias because many treated him with suspicion = less representative
28
Reliability
- PO studies lack reliability because data recording is unsystematic and hard to replicate
- Personal characteristics of observers are different and so affects outcomes
- Wright - found that as a black female she was met with hostility by some white teachers, but readily accepted by black pupils
29
Longitudinal studies
| Lacey, J.W.B Douglas, Leon Feinstein
- Lacey - 4 year old PO study of Hightown Grammar school
- J.W.B Douglas - study following 5632 children all born in the first week of March in 1946 through their schooling
- Feinstein - used data from NCD's and the British Cohort study which follows a group born in 1970 to study educational differences in educational achievement
30
Advantages and disadvantages of longitudinal studies
- Trace developments over time
- Make comparisons over time
- Sample attition - people dropping out of study; NCD's lost 1/3 of its original sample of 17,400
- Large amounts of data can be difficult to analyse
- Expensive
