10 marks

• Interpretivist:

  • Micro
  • Focuses on individuals
  • Interactionist approach
  • Qualitative data
  • Subjective
  • Small scale interactions
  • Socialisation
  • Shared meanings
• Interpretivist example


  Humphreys - Tea Room Trade (non-participant observation)
• Positivist:

  • Objective
  • Quantitative data
  • Statistics
  • Social cause and effect
  • Macro
  • Large scale
  • 'new religion' is science
  • General scope
• Methodology


  Link to the question, does it use primary or secondary data? is it sensitive? strengths of the sampling method?
• Practical:

  • Subject matter - suitable? why?
  • Time and money
  • Research opportunity
  • Personal skills and characteristics taken into account
• Ethical:

  • Deception
  • Informed consent
  • Protection from harm
  • Confidentiality
  • Privacy
  • Right to Withdraw
  • Consequences of publication
• Methodological Pluralism


  Combines methods and types of data for a fuller picture of social life
• Triangulation


  Assesses validity and reliability via investigator, theoretical and methodological concerns
• Gatekeepers


  Hughes (2000) - Key individuals who provide access to a research group
• Research Bargain:

  • Generalisable?
  • Reliable?
  • Objective?
  • Valid?
  • Ethical?
  • Representative?
• Questionnaires (self-report)


  List of questions to collect mainly quantitative data that can be operationalised into categories using open and closed questions for comparisons of data. No researcher is present.

  • Adv: easily repeated on large numbers, cheap, quick, easy to analyse, anonymous (reduces social desirability), standardised, objective, very ethical and less embarrassing than interviews
  • Dis: social desirability bias - not truthful, limited in answers due to multiple choice, might not understand the questions and cannot explain, poor response rate, leading questions and no depth
• Structured interview (self-report)

  Set in advance with standardised questions in a script. The answers are predetermined by the researcher as they are mostly closed questions for quantitative data. This is formal and if probing questions are used, they are scripted.

  • Adv: Highly reliable, less interviewer bias, easy to analyse and collect, objective, easy to answer, can be done on a large sample
  • Dis: less valid as no rich qualitative data, no depth, cannot ask them to go into detail, inflexible, cannot explain questions, limited information, still an interviewer present so interviewer bias
• Semi-structured interview (self-report)


  Few questions are set in advance and can be standardised but still flexible to gain more qualitative data. The interviewer has specific themes to discuss and can give probing questions

  • Adv: Valid, somewhat reliable, rich, in-depth data, flexible, sensitive and comparable
  • Dis: less reliable than structured but more valid, less valid than unstructured but more reliable
• Unstructured interview (self-report)


  Questions are not set in advance, they are given themes to discuss for qualitative data, participants can elaborate on their experiences. The interviewer can freely probe and ask follow-up questions.

  • Adv: High validity, in-depth, rich data, flexible, sensitive, opinionated, not limited, use of own words, explorative
  • Dis: Not reliable, cannot be standardised, not generalisable, not representative, social desirability bias, small-scale, time-consuming, interviewer must be skilled, subjective, not straightforward or easy to analyse
• Participant observation (non-experimental)


  Observes social reality by the researcher participating in what the participants say and do. This is typically small-scale and collects qualitative data
• Participant observation (non-experimental)


  Covert: (participants don't know they're being observed)

  • Adv: Valid, no social desirability bias, insight, depth, rich data, discover meanings
  • Dis: dangerous, risk 'going native', cannot take notes, based on memory, low reliability, not standardised, threat of subjectivity, bias of relationships, not representative at small-scale
• Participant observation (non-experimental)


  Overt: (participants know they're being observed)

  • Adv: High reliability, more able to standardise, less subjective, easier to collect, comprehensive understanding
  • Dis: Less valid, threat of social desirability bias, demand characteristics, time-consuming, dangerous
• Non-participant observation (non-experimental)


  Observer is like a 'birdwatcher', observes without joining in, done in a natural setting for qualitative data.
• Non-participant observation (non-experimental)


  Covert: (don't know they're being observed)

  • Adv: Less likely to influence behaviour, less bias, less demand characteristics
  • Dis: deception, fewer opportunities to discover meanings and subjective
• Non-participant observation (non-experimental)


  Overt: (know they're being observed)

  • Adv: can be recorded easily, can record findngs immediately, consensual, can ask openly, not illegal beaviour, quantitative data too
  • Dis: Hawthorne effect (social desirability), alternate their behaviour, demand characteristics, may not understand why they're being observed
• Controlled observation (experimental?)


  Examining spontaneous behaviour under lab conditions

  • Adv: High levels of control, improves internal validity, less likely to affect DV
  • Dis: Lacks ecological validity/mundane realism - difficult to generalise as it is not representative
• Correlation 


  To discover how 2+ covariables vary with each other in terms of a relationship

  • Adv: Provides precise information and gives a numerical value
  • Dis: Fails to show causation (might be a 3rd factor), decreases the internal validity
• Focus groups (non-experimental)


  An observation where individuals are interviewed as a group to solicit opinions using qualitative data.

  • Adv: some will be more comfortable in a group so open up more, rich, valid, depth, reflective data, follow-up research, observe group dynamics and norms
  • Dis: some may dominate the discussion, must keep the group focused (skilled interviewer), peer group pressure, social desirability bias, difficult to analyse
• Official statistics (secondary data method)


  Numerical, quantitative data produced by national and local government bodies that cover a wide range of behaviour including births, deaths, divorces, crime etc.

  • Adv: representative, generalisable, large-scale, reliable, standardised, objective, easy to anaylse and collect, cheap, quick, available to everyone and comparable
  • Dis: dark figure of crime, not valid, closed questions, limited answers, social desirability bias, political bias and must know construction of data to assess quality
• Historical documents (secondary data method)


  Government reports, treatises, diaries and novels from a particular period, may add qualitative insight or photographs to back up data.

  • Adv: Not exact details (generalisable), easy access, little cost, comprehensive, quick, valid, in-depth and rich data
  • Dis: subjective, not specific to aim of study, authentic?, not reliable, may not represent larger population, not standardised, difficult to analyse or interpret
• Case studies 


  An in-depth study of one person, group, or event. Nearly every aspect of the subject's life and history is analysed to seek patterns and causes of behavior.

  • Adv: rich, in-depth data, longitudinal, valid, more detailed to individual cases
  • Dis: hard to generalise as not representative of general population, unique and specific, lacking population validity
• Content Analysis 


  Indirect observation that is used to analyse qualitative data and transform it into quantitative data.

  • Adv: Easier to analyse and look for patterns
  • Dis: observer bias may affect validity of the findings, cannot draw cause/effect relationships, authors are usually unknown, decreases reliability
• Online research


  Access participants via the internet or social networking, often involving a questionnaire but could be experimental.

  • Adv: Access to large groups, quick, cheap
  • Dis: Response bias
• Qualitative data (written form)


  Adv: depth, rich, greater understanding, valid and opinionated

  Dis: difficult to analyse, lacks objectivity, may have different interpretations, time-consuming
• Quantative data (numerical form)


  Adv: Easy to analyse, easy to collect, reliable, standardised, quick and cheap

  Dis: Large amounts of data, lacks validity and no deep understanding
• Primary data (first-hand from researcher)


  Adv: can be controlled and designed to fit the aims and hypothesis of specific study

  Dis: very lengthy, time-consuming and expensive
• Secondary data (collected by different researchers and analysed to fit a specific study)


  Adv: Simple, cheap, less time-consuming, less equipment needed, little BPS guidelines to abide by

  Dis: may not fit the exact needs of the study, researcher bias - may choose studies that are biased towards their hypothesis, lacks reliability
• Opportunity sampling (convenience/availability)- non-probability


  Involves selecting participants simply because it is convenient as they are readily available to the researcher.

  • Adv: Cheap, high participation rate, easy to do, unlikely to drop out and quick
  • Dis: Biased, unrepresentative, not generalisable
• Random sampling - probability


  Participants randomly chosen from a sampling frame with equal chances.

  • Adv: Unbiased, representative, generalisable, easy to calculate
  • Dis: Time-consuming, hard to be completely random, need minorities so can still be unrepresentative
• Volunteer sampling - non-probability


  Self-selected sampling when asked or responded to an advertisement.

  • Adv: Quick, easy, less cost, unlikely to drop out
  • Dis: Biased, demand characteristics, unrepresentative and not generalisable
• Systematic sampling - probability


  Sample is selected by choosing every nth member of a sampling frame, using a predetermined system.

  • Adv: Easy to select, more applicable, guaranteed to be an even sample, representative, generalisable, valid
  • Dis: May react to hidden periodic trait and could no longer be random
• Stratified sampling - probability


  Sampling frame is divided and subdivided into smaller lists and random samples are taken from each list.

  Adv: representative, generalisable, reduces bias, sub-groups are all represented

  Dis: time-consuming, complicated, costly, large sample sizes, sometimes difficult to identify the strata
• Snowball sampling - non-probability


  Used when the sample is difficult to obtain. One person invited to the study will introduce another and so on.

  • Adv: efficient, good with difficult to obtain groups
  • Dis: biased, not representative, not generalisable, small sample
• Quota sampling - non-probability


  Dividing population according to characteristics e.g. age.

  • Adv: unbiased to random selection but not random in group characteristics
  • Dis: subjective, not everyone has the chance to be selected so biased, not representative, not generalisable
• Multi-stage cluster sampling - probability


  A sample area in a country is chosen randomly (1st) and randomly sampled from each cluster/area (2nd).

  • Adv: less cost, reduces travel, representative, generalisable
  • Dis: sampling error higher than random sampling of the same size, not efficient