Research Methods

1
Research Methods

• Definition of science
• Data collection and analysis

2
Sociology as a Science

• Debate over the scientific status of sociology
• Sociology is a sciencea view often supported
  by quantitative sociology
• Sociology is an art or humanitya view often
  supported by qualitative sociology
• Caution the distinction is not absolute

3

• Characteristics of science
• Knowledge is based on facts
• Facts are real and observable
• Objectivity in observation
• Use of scientific methods of observation and
  analysis
• The goal of science is truth
• A superior understanding of reality

4

• Should social sciences model themselves on
  natural sciences?
• The natural-science model
• May be irrelevant, because they deal with
  different subject matter
• Is unrealistic

5

• Production of scientific knowledge relies on
  negotiations and rituals, not only on objective
  observations
• Complete objectivity is impossible, because
  researchers cannot fully separate themselves from
  prior knowledge

6

• Does lack of complete objectivity make science
  impossible?
• Yes
• Relativism science is not distinct from other
  ways of acquiring knowledge. There is no ultimate
  truth
• No
• Science is possible, so long as it develops
  conventions that differentiate defensible claims
  to knowledge
• Rules about research methods distinguish good
  sociology from bad sociology

7
Operationalization

• Research tests theories or explores their
  applicability to different situations
• Operationalization is the process of translating
  theories into observable ideas, so that they can
  be tested
• Hypotheses are observable (testable) statements
  consistent with a theory

8

• Variables are observable equivalents of concepts
• They have a range of different values
• Independent variables causes
• Dependent variables effects

9
Validity and Reliability

• Operational definitions of concepts specify
• What will be observed
• How it will be observed
• How observations will be differentiated from
  non-observations
• They should ensure validity and reliability of
  observations

10

• Validity is the degree to which an empirical
  indicator reflects the concept
• Reliability is the degree of consistency of a
  measure
• A reliable measure produces the same measurement
  for the same phenomenon

11
Validity

• Validity of an empirical indicator
• Face validity the indicator fits with our mental
  image of the concept
• Construct validity the indicator performs as
  expected against indicators of related concepts
• Content validity the indicator covers most or
  all meanings of the concept, without overlapping
  onto other concepts

12

• Validity of research findings
• External validity generalizability of results
  beyond the sample
• It depends on sampling and research techniques
• Internal validity the degree to which
  conclusions are actually supported by the data

13
Another Approach to Validity and Reliability

• Goal of qualitative research is to gain
  understanding of the topic or group studied
• Validity and reliability of measurement processes
  are not an issue.
• Validity and reliability of findings are
  established if a convincing argument can be made
  that the topic or group is properly understood

14
Bias and Error

• Error an unintentional, unpredictable random
  error in every research
• Sources sampling error, respondents mistakes,
  errors in coding and analysis
• Bias systematic inaccuracies in data or
  analysis. They distort results in systematic ways
• Multiple sources

15

• Respondent biases
• Acquiescence bias
• Social desirability bias

16
Experiments

• Experiments are traditionally the standard of
  research
• Controlled environment where factors can be
  manipulated to determine effects on the outcome
• Experimenting is not often feasible in social
  sciences
• Variables cannot be manipulated for practical or
  ethical reasons
• External validity of findings is problematic

17

• Surveys are normally done with a
  quasi-experimental design
• Much information is collected about naturally
  occurring phenomena
• Effects of confounding variables can be removed
  during the analysis

18
Surveys and Pseudo-surveys

• Good for study of large populations
• Goals
• Description of a population
• Theory testing
• Generalization

19

• Pseudo-surveys have a low generalizability
  because of
• Poor sampling (based on convenience and luck, not
  at random)
• Low response rate (people who do not answer the
  questionnaire may be systematically different
  from those who do)
• Goal of pseudo-surveys may be manipulation,
  rather than data collection

20
Constructing Survey Questions

• Goal to make questions unambiguous for both the
  respondent and the researcher
• Questions should have focus (only one), brevity,
  clarity, and relevance, and should avoid bias

21

• Potential problems
• Double-barreled questions (have two or more
  subjects or referents)
• Respondents inability to answer accurately
• Threatening topics (threat can be reduced by the
  wording)
• Question order (ask threatening questions near
  the end)

22
Random Sampling, Sample Size and Response Rates

• Random sample each element in the population has
  an equal (non-zero) chance of being selected into
  the sample
• It is not a guarantee that the sample will be
  representative
• Simple random sampling is seldom used, because
  complete list of the population is not often
  available

23

• Generalizability of findings does not depend on
  the size of the sample, but on its selection
  process
• Response rate a percentage of people who
  actually complete the survey
• It influences generalizability

24
Field research Ethnographic or Participant
Observation

• Goal collecting rich, nuanced qualitative data
  that may or may not be generalized
• Techniques
• Participant observation (ethnography)
• In-depth interviewing
• Documentary analysis

25

• The researcher participates in daily activities
  of research subjects, usually for an extended
  period of time
• Field notes (or recordings)
• No rigid research design
• Particularly appropriate for some theories (e.g.,
  Goffmans dramaturgical approach)

26
Field research In-depth interviews/Documentation

• In-depth interviews varying degree of structure
• Semi-structured interviews are usual the
  researcher sets basic questions, but allows
  participants to explore other topics
• Documentation usual when studying formal
  organizations

27

• Choosing a site and informants
• Research topic
• Practicality
• Length of stay on site until the researcher is
  no longer gaining much new information
• Flexibility of field research allows for
  correcting mistakes in research design and
  pursuit of unexpected leads

28
Secondary Data Analysis

• Sources official statistics and existing surveys
• Has grown because of increased availability of
  computers and statistical software
• Advantages
• Broad coverage of data
• Use of data whose collection needs more expertise
  and resources than the researcher has

29

• Disadvantages
• Data are often not directly related to the
  researchers ideas
• The researcher needs to learn new techniques of
  analysis

30
Historical Research and Content Analysis

• Historical methods use of documents, records,
  interviews with participants
• Problems
• Bias in the documents
• Loss or destruction of documents

31

• Content analysis records are sampled and
  analyzed to reveal patterns
• Manifest content analysis words, phrases, and
  images are counted
• Latent content analysis focus is on implicit
  themes

32
Selecting a Research Method

• Any method can be used to investigate social
  phenomena from any theoretical perspective
• Some methods are seldom used in some perspectives
• e.g., surveys in symbolic interactionism
• The problem determines the method
• However, some researchers prefer to use familiar
  techniques, and frame their research questions
  accordingly

33
The Context of Social Research

• Purposes of research
• Exploratory (descriptive) research to find out
  more about a topic
• Explanatory research to test competing theories
  or applicability of a theory
• Predictive research prediction is distinct from
  explanation
• Participatory action research to empower the
  studied group

34
Feminism and Research Methods

• Empowerment is among the goals of feminist
  research
• Feminist critique of traditional science
• Subordination of womens perspective by
  insistence on (male) characteristic of
  objectivity
• Unbiased science is impossible, therefore biases
  should be openly acknowledged

35

• Variants of feminist methodology
• Deliberate abandonment of science
• Use of science to eradicate suppression of women
• Use of lived experiences of women as a standpoint
  to establish a less distorted understanding of
  society

36
Research Ethics

• Nuremberg trials highlighted the contrast between
  the acquisition of knowledge and the protection
  of research subjects

37

• Principles of research ethics
• Voluntary participation (people may not be
  coerced or tricked into participating)
• Informed consent voluntary participation is
  based on it
• Deception may only be used as a last resort, and
  when benefits of the research outweigh the harm
  of deception
• Identity of participants must be protected
  through anonymity or confidentiality
• No unnecessary harm to participants