The Experimental Method
Involves manipulating an:
- Independent variable (a cause)
- Observing the dependent variable (the effect)
- Along with controlling extraneous variables (other factors)
P Natural scientists like experiments!
- Laboratory experiments - more control, demand characteristics, reliable.
- Field experiements - more natural, less control. Ecogologically valid.
Experiments are RARE in sociology.
Supporting Scientific Methods
Positivists - Durkheim and Comte favour use of scientific methods! :)
^ This was due to their success in producing and testing knowledge in the natural sciences.
- Sociology as the study of social facts - ways in which society influences the behaviour of individuals. (According to Durkheim and Comte)
Durkheim argued that the comparitative study was an appropriate subsitute for the experiment for the experiment because the thought processes behind both were similar.
Scientific methods can produce sociological explanations which are generalizable and testable!
Against Scientific Methods
- Interactionalists argue sociology should study social action not facts.
- Social action involves the social meanings given to behaviour, not just the objective behaviour which cannot be observed.
- Need to discover the reasons FOR behaviour rather than the CAUSES of behaviour.
- Hawthorne Effect - act of studying people may change their behaviour.
- 'Self fulfilling prophecy' - expectations of researcher may influence the behaviour of the participants.
P- Favoured by sociologists who wish to take a scientific approach, e.g. positivists!
- Collection and analysis of data, usually collected by asking questions.
- Data is quantified and anaylsed in a systematic way.
- Can be used to test hypotheses by looking for correlations between variables, e.g. social class and the rate of crime.
Types of Comparative Surveys
- Longitudinal: Follow development of subject over a long period.
- Historical: Compare behaviour in different periods of time.
- Cross sectional: Study different subjects at one moment of time, e.g. before & after.
- Cross-cultural: Study different societies & useful in demonstrating behaviours (e.g. gender roles) are learned through socialisation and aren't natural.
Quantitative Methods - Secondary Data
Sociologists can analyse data collected by somebody else, e.g. Durkheim's study on Suicide.
- Provide an insight into what people think & do.
- Unobtrusive - reactive and investigator effects less likely.
- Archived research widely available on a varitey of topics.
- Often reliable and valid.
- May be incomplete.
- May be representative of one perspective.
- Access to some content may be limited.
- May not provide an insight into participants personal thinking for physical data.
Collect both objective and subjective data. Exam questions tend to focus on participant observation or unstructured interviews.
- Watching and listening to people during their everyday lives - interpretating behaviour by discovering the social meanings that they and others give to the behaviour.
- Involvement of the researcher is minimal.
- Researcher's participation is deliberate.
- Involves joining in with participants in order to understand and describe their everyday lives.
Participant Observation - Advantages
- Participants studied in natural environments to enable detailed accounts of everyday life.
- Researcher can see what's important to the participants instead of imposing their views through questions.
- Researchers can share their experiences of the participants enabling them to discover more information. - This is why sociologists favour participant observation.
- COVERT - minimalizes the possibility of the participants changing their behaviour because they are being observed. (Demand characteristics)
- COVERT - allows access to groups who might exclude outsiders because their behaviour is shameful or secret, e.g. research into gangs.
J. Patrick covertly studied a Glasgow gang using a false name.
Participant Observation - Disadvantages
Tend to be by people who prefer the scientific methods, e.g. positivists.
- More scientifc to be detached from pariticipants & study objective behaviour from the outside.
- Researcher may change the participants behaviour. (Demand characteristics).
- Sharing the participants p.o.v. may encourage bias.
- Cannot be replicated and findings verified - therefore making it unreliable & unscientific.
- Groups studied - unlikely to represent population.
(Practical problems and ethical issues on other cards)
Participant Observation - Disadvantages
- Time consuming.
- Difficult to access some groups - often studied lower class groups compared to upper class groups.
- COVERT - may be limited to groups where the researcher can pass as a participant - class, age, gender, ethnicity may affect this.
- Need to learn the language of the participants to fit in & understand them.
- COVERT - Difficult to ask q's and record behaviour.
- Researchers may witness or be expected to participate in criminal behaviour.
- Deception of participants - no informed consent.
- Professional codes of conduct for medical, psychological & sociological research normally require informed consent from participants.
- Information of private lives may be published without consent.
Information collected by others, e.g. Durkheim's study on Suicide. Evlauations of usefulness based upon reliability, completeness, accuracy and validity.
Types of Secondary Data
- Organisational records - schools, churches, employers etc. keep records.
- Mass media output - TV, radio, recorded music etc. (Sociologists particually interested in representation of gender and ethnicity)
- Diaries, autobiographies and other personal documents - discover personal views & opinions, personal lives and look at the relationships between race, class etc. in the past.
- Published sociological research - used to compare the present with the past or to study a large number of subject over a large area when time and/or funds are limited.
- Produced by a varitey of bodies.
- Includes the government - official stats (briths, deaths, crime, class etc.)
- Non-government - some of the above, and also on church attendence, racial disadvantage & job opportunities for women.
- Positvists favour scientific methods. Assume stats can describe aspects of social life such as the extent of crime, suicide rate & marriage breakdowns.
- Durkheim's Suicide study - the model for research using official stats.
- Stats can be used in comparative surveys to test a hypothesis, e.g. Durkheim.
- Researcher to detach self from participants, avoids changing behaviour.
- Data is reliable!
- May be the only source of data avaliable (e.g. historical studies of a family).
- Cheap, speedy source of large amounts of data - analysed by computers.
Social Statistics - Disadvantages
- Positvists don't accept the use of official stats uncritically. - Not conerned with technical problems of gathering and interpretating data which may affect the accuracy of the research.
- Interactionalists question validity rather than reliability of stats. - Claim that stats are not neutral descriptions of social life but the result of negotiations which lead to the definitions of certain acts, e.g. crime & suicide.
- Stats tell us about the work of coroners rather than suicides & health stats about decisions made by doctors rather than patterns of disease.
Problems with CRIME statistics
- Tell us more about the police and courts than about criminals.
- May be incomplete as not all crimes are reported because: victims afraid or ashamed, victims may see reporting crimes as a waste of time, victims unaware or unconerned, crime is 'victimless' & requires police detection.
- Interactionalists who argue crime figures are the result of police actions and do not reflect real rates of crime, e.g. number of sex offences increased since police officers went undercover in public loos to catch homosexuals.