The methodological phase

Part 1: Research design

Plan

  • Research design: key concepts
  • Qualitative designs
  • Quantitative designs
  • Mixed methods designs

The methodological phase

  • Ethical considerations (TCPS 2: CORE 2022).

  • Research designs (this week).

  • Sampling (next week).

  • Concept definition and operationalization (next week).

  • Qualitative data collection (in 2 weeks).

  • Quantitative data collection (in 2 weeks).

Research design

Elements to consider

  • Comparisons
  • Interventions
  • Research milieu (location/context).
  • Data collection methods and instruments.
  • Temporality (when are you collecting data)
  • Communications with participants
  • Biases
  • Control of exogenous and endogenous factors

Types of research and their associated designs

Type Goal Research designs
descriptive/exploratory (qualtiative) Describe phenomena Phenomenology; Ethnography; Grounded theory; Case study; Qualitative descriptive study
descriptive (quantitative) Describe population Simple descriptive study; Comparative study; Longitudinal study; Case study.
Correlational descriptive, predictive, confirmative Explore or verify relations correlational descriptive or predictive study; cohort study; case-control study.
Experimental Verify causal relations Before-and-after design with control group; after design with control group; factorial design.
Quasi-experimental Verify causal relations Same as experimental, but with a non-equivalent control group.

Qualitative research designs

Phenomenology

  • Aims to understand the lived experience of participants

  • Focus on a particular phenomenon

  • Data usually collected from interviews

Ethnography

  • Aims to understand behaviours and group cultures.

  • Rooted in anthropology.

  • Focus on a natural setting in which individuals share a common culture.

  • Data collected from participant observation, interviews, and documents

Grounded theory

  • Aims to generate theory from data collected in a natural setting.

  • Rooted in sociology and symbolic interactionism.

  • Focus on a process involving human actions and interactions and how they influence one another.

  • Data collected through interviews or other relevant data sources.

Case study

  • Aims to provide a deep understanding of a person, small group, or situation.

  • Rooted in anthropology and sociology

  • Focus on one or several case(s) in a natural setting.

  • Data collected through observation, interviews, documents, etc.

Qualitative descriptive

  • Aims to provide a simple description of a phenomenon, situation or event.

  • Focus on the who, what, and where of a phenomenon, situation or event in a natural setting.

  • Data collected through any relevant method.

Quantitative research designs

Simple or comparative descriptive

  • Aims to describe the characteristics of a population or a phenomenon.

  • Can describe a single group (simple) or compare the characteristics of different groups (comparative).

  • Uses one or multiple variables.

  • No hypotheses or causal claims.

  • Can be cross-sectional (data collected at one point in time) or longitudinal (data collected at multiple points in time).

Correlational descriptive, predictive, or confirmative

  • Aims to explore, predict, or confirm relations between variables

  • Continuum from descriptive when no or limited pre-existing knowledge of relation to confirmative when there is a theory being tested.

  • Uses one or multiple variables.

  • Hypotheses are used in predictive or confirmative designs.

  • No causal claims.

cohort and case-control design

  • These are common in epidemiology but can also be applied to other fields.

  • Prospective cohort studies follow a group exposed to a risk factor (comes from epidemiology) and another group that is not exposed.

  • Retrospective cohort studies start from a phenomenon and then seek explanations in the past.

  • Case-control designs aim to link a current phenomenon to a past phenomenon by comparing a group in which the phenomenon is currently present to a group in which it is not.

Experimental and quasi-experimental

  • Before-and-after design with a control group.

    • Random assignment to groups → Measures → treatment (experimental group) → Measures

  • after design with a control group.

    • Random assignment to groups → Treatment (experimental group) → Measures

  • factorial design

    • (Quasi-)experimental design in which two or more variables are manipulated.

  • Experimental design requires randomization.

  • In quasi-experimental research, there is limited control of external factors and of the setting, and there is no randomization.

Mixed methods research designs

Exploratory sequential design

QUAL → quan

  • Priority on qualitative data.

  • Aims to refine a theory or explore relations between phenomenons observed during the qualitative phase.

  • Integration of quantitative and qualitative data during the interpretation phase.

Explanatory sequential design

QUAN → qual

  • Priority on quantitative data.

  • Integration of quantitative and qualitative data during the interpretation phase.

  • Qualitative findings aim to clarify or explain quantitative findings.

Parallel design

QUAN+QUAL or QUAL+QUAN

  • The quantitative and qualitative data are equally important (in theory, but generally not in practice).

  • Triangulation is the main purpose.

  • Data are analyzed separately and results are compared.

Embedded design

QUAN/qual or QUAL/quan

  • Qualitative and quantitative data are collected concomitantly.

  • One method dominates and the other is used to complement.

  • Data are integrated during the analysis phase.