Lecture #2 Research Design

What is a Research Design?

  • A methodological framework that facilitates the gathering and analysis of data

  • Research designs may aim to: uncover cause-effect relationships; understand; generalize findings; give voice

    • Quantitative research tries to find cause-effect relationship

    • Generalize finding because sociology works with samples instead of the whole population

  • Research objectives → research design → research methods

    • The research process obj → design → methods

Research Design: Decisions and Considerations

  1. What research orientation do we wish to employ?

    1. Quantitative or qualitative\

  2. Will we conduct a cross-sectional or longitudinal study?

  3. How will our study be structured?

  4. Whom or what do we wish to study?

  5. Are there any ethical considerations?

  6. What are the key variables and how will we measure them?

  7. How will we select our sample?

  8. What type of data do we want to collect and how will we collect it?

  9. How will we go about analyzing our data?

Quantitative Research Designs

  • Goal: to examine relationships between variables

  • Ideally, researcher want to uncover cause-effect relationships between variables

  • Independent Variable (X)

    • Variable as it shifts the dependent variable shifts

    • The perceive cause

  • Dependent Variable (Y)

    • The perceive effect

  • Variable is an entity that takes on different attributes

    • Example; races include white, black, Asian etc.

Independent and Dependent Variable

  • Sex & income

    • Sex = independent variable, income = dependent

  • political ideology & sexual orientation

    • Sexual orientation = independent Variable, political ideology = dependent

  • Number of cancer-related deaths per 1000 population & number of medical doctors per 1000 population

    • #of doctors = independent, #of cancer-related = variable

    • Could be reversed, high cancer rate could bring more doctors

  • Hand washing behaviour & sex

    • Sex = independent, hand wash = dependent

  • occupation & suicide rates

    • occupation = independent, suicide rates = dependent

Quantitative Research Designs: Hypothesis Testing

  • What is a hypothesis?

  • A statement that predicts a relationship between two variables (X→Y)

Conceptual Hypothesis

  • Focuses on the relationship between two variables (an independent variable and a dependent variable)

  • Does not predict the specific direction or nature of the relationship

    • Example: There is a relationship between gender and income in the Canadian Labour Market

Working Hypothesis

  • More specific than the conceptual hypothesis; predicts the direction or nature of the relationship between X and Y

    • Example: On average, women earn less than men in the Canadian Market

Quantitative Research Designs: Theory to Hypothesis

  • Theory: Suicide rates related to degree of social integration within a collectivity (Durkheim, 1897)

  • Conceptual Hypothesis: There is a relationship between religious affiliation and suicide rates

  • Working Hypothesis: Catholics are less likely to commit suicide than protestants.

Quantitative Research Designs: Criteria for causality

  1. Variables must be statistically correlated

    1. They must be statistical evidence that the dependent variable changes are changing the independent variable

  2. Time/order criterion: X must precede Y in time

  3. Relationship must be non-spurious

    1. Spurious relationship is where u think X is causing Y

Qualitative Research Designs: Objective

  • to understand the subjective experiences of participants

    • I.e., the meanings t hat they attach to what they do or their social situation; how they make sense of their world

    • To this end, qualitative research is typically design to:

      • help the researcher become a facilitator and, often a participant

      • help researcher to develop a rapport with participants — give voice to participants

Example of a Qualtative Research Topic

  • Exploring the understandings and perceptions of people who participate in alt-right groups

  • Learn about the white supremacy ideology

Research Questions

  • Must be specific and clearly stated

  • Should inform and direct each stage of the research project

  • Should be the result of a logical refining and honing process: broad study topic → research questions

  • Grounded in existing theory and research

  • Usually formulated based on some void or unresolved question in existing research literature

  • Framed in an open, speculative and ‘objective’ way (not predicting a specific result/finding as with hypotheses)

Research Designs

  • Seven commonly used research designs

    • Experimental (Quantitative)

    • Survey (Quantitative)

    • Secondary analysis (Quantitative)

    • In-depth interview studies (Qualitative)

    • Focus group studies (Qualitative)

    • Ethnography (Could be both, usually qualitative)

    • Content analysis (Could be both)

Cross-sectional Research Design

  • Collection of data from more than one case at a single point in time

  • E.g., Canadian Consensus

Longitudinal Research Designs

  • Study of the same population on more than one occasion

    • Time series study or trend

    • Cohort study

    • Panel study

Units of analysis

Whom or what do we wish to study?

  • individuals

  • Groups

  • Organizations

  • Social categories

  • Social institutions

  • Social artifacts

  • Society

  • CAUTION: ECOLOGICAL FALLACY

    • Drawing conclusions about individuals based on group level data

      • E.g., protestants are less likely to commit suicide compared to catholic, but don’t conclude every catholic is suicidal