Research methodology

A map of common approaches

As a researcher you will encounter many different potential approaches to your research design, many of which overlap, vary in their use of terminology, and use labels in different ways.

Some Epistemological positions

  • Positivism: Emphasises observable evidence, systematic testing and replicable methods.
  • Post-positivism: Retains a commitment to evidence while treating observation and knowledge claims as fallible and provisional.
  • Interpretivism: Seeks to understand subjective and socially situated meanings.
  • Postmodern approaches: Question the possibility of a single stable account of truth and attend to fragmentation, discourse and power.
  • Empiricism and rationalism: Broad philosophical traditions concerning sensory experience and reason respectively as sources of knowledge.
  • Indigenous and decolonial approaches: May foreground relationality, community accountability, reciprocity and the validity of knowledge systems marginalised by colonial research traditions. These approaches should be engaged through appropriate Indigenous scholarship and local protocols.

Broad methodological approaches

  • Quantitative research primarily uses numerical data to measure, compare, model, test or predict.
  • Qualitative research primarily uses language, images, observations, documents or other non-numerical material to explore meaning, process, experience or context.
  • Mixed-methods research deliberately integrates quantitative and qualitative components to address a research problem.

These broad labels do not, by themselves, reveal the researcher’s ontology or epistemology. More specific research designs are listed below:

Often used with realist-leaning questions:

  • Experimental research. Controlled experiments that manipulate variables to test hypotheses and can establish cause-and-effect.
  • Quasi-experimental designs. Studies cause-and-effect between independent and dependent variables based on non-random criteria.
  • Randomised controlled trial (RCT). A form of experiment in which participants are randomly assigned to groups to test an intervention.
  • Longitudinal study. Repeated observations of the same subjects over a long period to study change.
  • Systematic review. Systematically collecting, appraising and synthesising existing studies on a topic. Also used in qualitative and mixed methods approaches.
  • Computational modelling. Using mathematical and computer-based simulation to study complex systems and predict their behaviour.

Often used with constructivist-leaning questions:

  • Ethnography. In-depth study of cultures and social interaction, typically through participant observation and interviews.
  • Case study. In-depth exploration of a specific instance, event, organisation or person in its real-world context.
  • Grounded theory. Develops theory from systematic data gathering and analysis, letting the theory emerge from the data.
  • Narrative inquiry. Uses storytelling to explore how individuals derive meaning and insights from life experiences.
  • Discourse analysis. Analyses language, texts and communication to see how meaning is constructed in social contexts.
  • Phenomenology. Explores individuals’ lived experience, and how people perceive and make sense of their world.

Some research designs have several philosophical variants. For example, case study and grounded theory are used in different ways across paradigms. Do not assume that a particular label settles the philosophical position.

Methods

  • Collecting data, often quantitative: controlled experiments, structured surveys, systematic observation, physiological measurement.
  • Collecting data, often qualitative: semi-structured interviews, focus groups, participant observation, reflective journals.
  • Analysing data: statistical analysis (for example regression) for numerical data; coding and thematic analysis for qualitative data.

Reading across the layers

Certain combinations are common, such as positivist, quantitative and experimental approaches, or interpretivist, qualitative and interview-based approaches, but they are patterns, not strict rules. Critical realists often use mixed methods. What matters is that you can explain why your layers fit together.