Research Methods Flashcards Preview

Sociology AS > Research Methods > Flashcards

Flashcards in Research Methods Deck (17):

Quantitative: Positivism

- Compte: Sociology can be scientific
- They like truths, fact
- Stats, trends and patterns
- Value free
- Correlational
- Human behaviour shaped by external stimuli, not internal (in the mind)
- Reductionist?


Quantitative: Experiments

- Hypothesis, IV, DV
- Field experiments: natural setting
- e.g. Rosenthal and Jacobson
- But difficult to control variables, Hawthorne effect, if covert, ethical issues
- Comparative method: looks at past events
- Social groups, times of places are systematically compared to try to isolate variables
- E.g: Durkheim on suicide, Weber on the protestant ethic, Marx on social change, more difficult to isolate variables


Qualitative Methodology

- Meaning and motive, empathy
- People do not simply react to external stimuli but interpret the meaning of stimuli before reacting
- Obersvation, interviews etc
- Cannot be reduced to statistical data
- Weber: Study of social action, requires verstehen, e.g. Calvinism
- Symbolic interactionism, individuals possess a self-concept
- Blumer: Sociologists need to understand the viewpoint of the people whose behaviour they are trying to understand
- Labelling theory
- Phenomenology: reject that casual relationships are possible, believe the social world needs to be classified
- Feminism: Oakley: Feminist, interviewing, rapport


The Research Process

- Areas studied affected by: values and beliefs of researcher
- Developments in the subject e.g. globalization
- Developments in the social world e.g emergence of fundamentalism)
- Gov. Policies
- Funding
- Practicality: time money, access etc



- Confidentiality
- Consent
- Care
- Deceit
- Debrief
- Man an Ghail
- Zimbardo
- Milgram
- Laud Humphreys
- Going native
- Rosenthal and Jacabson
- Black like me - Griffin
- Kinsey: Sexuality (the scale)



- Bryman: 4 types
1. Measurement validity: if something really measures what it intends to (e.g. IQ tests)
2. Internal
3. External: can it be generalized?
4. Ecological



- Opportunity
- Volunteer
- Random
- Stratified random
- Quota sampling: quotas established which determine how many people with a particular characteristics are studies, often used with opinion polls
- Multi- stage: sample of a sample, e.g. sample of voters in a sample of constituencies
- Snowballing
- Non-representative sampling: picked because or despite being untypical or to study specific characteristics, used to falsify general theory, to find key informants, or in a convenience sample


Case Studies

- To develop comprehensive understanding by studying it in depth e.g. Willis' lads
- To falsify a theory (e.g. Gough's: Nayar Tribe and Murdock)
- To develop a typology (e.g. Douglas and suicide)
- To generate a hypothesis
- May lack internal validity
- Not generalisable


Pilot Studies

- Small-scale, preliminary studies carried out before a bigger study to improve, help design or test the feasibility of proposed research
- Used to:
- Test in interview questions are too ambiguous or aren't useful
- To develop ways of getting cooperation with participants
- To develop research skills
- To decide whether or not to proceed with research



- Can be postal or by telephone
- Open or closed
- Advantages:
- Large amounts of data can be collected quick
- Relatively easier
- Easier to analyse data
- Trends and Patterns
- Disadvantages:
- No motive or meaning
- People may lie
- Not many people complete them
- People mat not understand the questions



- Can be overt or covert, participant or non-participant
- Advantages:
- Covert: participants will act more naturally - validity
- Meaning and motive
- Hard to lie
- Overt: may not act naturally: The Hawthorne effect
- Covert: may be unethical
- Can't see trends and patterns
- Can be time-consuming and expensive
- Smaller sample size



- Gov stats cover a wide range, e.g. demography, e.g. employment or ed. achievement, conducts statistical surveys such as the general household survey and a census every decade. These are invaluable because they are easily accessible and ore thorough than a sociologist an produce
- Some positivists (include para about them) such as Durkheim see official stats as valid and reliable, however, this is not always the case e.g. not all crimes are reported or recorded by police
- Some believe it's possible to produce valid and reliable stats through research, e.g. reliable crime stats through methods like self-report studies and victim studies, however, this is not always reliable due to low response or the limitations of sampling
- Some argue that results are invalid, e.g. Box: self-report: people may lie, hide, exaggerate or simply forget crimes. The total number of crimes recorded would depend on the willingness and ability of respondents to be honest and interpretation of the researcher as to whether the crime took place, the researcher also chooses research to include
- Phenomenological view of stats: think all crime stats are invalid, simply a product of categorization procedures used to produce them. Atkinson: sees suicide stats as the product of coroner's assumptions about the sort of people who commit suicide
Circourel: all stats are based on subjective classifications
- Conflict theories: see stats as the product of inequalities in power, Miles and Irvine: gov. stats are not lies but collection procedures ad definitions are manipulated by govs, e.g. redefinitions of unemployment which reduce official unemployment figures, and the manipulation of data ion NHS hospital waiting lists by removing people who have missed appointments for an opporation, poverty stats also rely on gov. definitions,
- some question categories used. Theo Nichols: argues that the official definition of social class ignores the existence of wealth inequalities


Historical Sources

- Vital for studying long-term social changes
- Laslett: used Parish records and Anderson used census data
- Diaries, letters, autobiographies, etc
- Reliability and validity
- Meaning and motive
- Trends and Patterns


Life Documents

- Surviving docs may not be representative
- Open to different interpretations
- Highly subjective (same events would be viewed differently by different people
- content may be influenced by the identity of the person or intended readers


Mass Media and Content Analysis

- Formal content analysis: classifying and counting content
- Thematic analysis: examines a topic and looks for message that lies behind the coverage
- Textual analysis: detailed analysis of small pieces of text
- Audience analysis: how text is interpreted by them


Assessing Secondary Sources

- Authenticity
- Credibility
- Representativeness
- Meaning



- Bryman: argues most sociologists use quants and quals, sociologists as far back as Weber have combined methods
- Other examples: Eileen Barker: study of the Unification church; 'The moonies' used observations, questionnaires and interviews
- the use of stat computer programs to analyse data from ethnography studies
- Winlow's study of bouncers, used observation, interviews and secondary sources
- Hammersley: 3 ways of combining methods:
- Triangulation: findings are cross-checked using different methods
- Facilitation: methods used to assist or develop another
- Complementary: different methods are combined to dovetail different aspects of an investigation
- Bryman: multi-stage research is used in different ways for different purposed
- Multi stage research: fills in gaps
- Some methods study static features of social life and others processual
- different methods gain different perspectives
- Different methods to help generalise
- Quals for meaning and motive, quants for trends and patterns
-Studies diff aspects
- Useful as limitations of each method could be partially overcome