3.5 Purposes of mixing methods

Beyond choosing a design, it helps to be clear about why you are mixing methods. Researchers mix methods for different purposes, so articulating your purpose strengthens your rationale.
Greene, Caracelli, and Graham (1989) identified five purposes for mixing methods that remain widely used today:
- Triangulation seeks convergence and correspondence of results from different methods. You use multiple methods to study the same phenomenon, checking whether findings align. This increases validity by counteracting the biases of any single method.
- Example: Comparing survey responses about teamwork quality with observations of team meetings and interview accounts of collaboration.
- Complementarity seeks elaboration, enhancement and clarification. Results from one method help explain or extend results from another. Each method contributes different insights that together create a richer picture.
- Example: Using interviews with staff to understand what factors might drive the patterns and could explain and contextualise the quantitative results.
- Development uses results from one method to inform the design or implementation of another. One strand helps develop the other.
- Example: Using focus group discussions to develop survey items, ensuring the questionnaire reflects participant language and concerns.
- Initiation seeks paradoxes, contradictions and fresh perspectives. When methods yield different results this is not a problem but an opportunity. Contradictions can prompt new insights and deeper investigation.
- Example: Discovering that survey data suggests high satisfaction while interviews reveal significant frustration, prompting an investigation into what explains this discrepancy.
- Expansion extends the breadth and range of inquiry by using different methods for different components. Rather than studying the same thing with multiple methods you use each method to address a different aspect of the problem.
- Example: Using quantitative methods to measure outcomes and qualitative methods to study implementation processes.
When designing a mixed methods study, ask yourself: Why am I mixing methods? What purpose does it serve? At what stage will the strands be connected? What would be lost if one were removed? The degree of benefit you get from mixing depends on how deliberately you employ the two strands; a tightly integrated design will return more than two parallel sub-studies stapled together.
Unit 4 walks through a fully published mixed methods study (Herodotou et al., 2019) end-to-end, showing how each design choice shows up in practice.
Activity 3.4: Matching purposes to examples
Match each research scenario to the primary purpose of mixing methods it represents.
Two lists follow, match one item from the first with one item from the second. Each item can only be matched once. There are 5 items in each list.
Development
Initiation
Expansion
Triangulation
Complementarity
Match each of the previous list items with an item from the following list:
a.A researcher conducts both a survey and interviews about workplace culture, comparing findings to check whether they converge.
b.A researcher uses interview data to explain why some schools show better outcomes than others in the quantitative dataset.
c.A researcher finds that quantitative data suggests a programme is effective, but qualitative data reveals participant dissatisfaction. They investigate this contradiction further.
d.A researcher uses interview findings to develop a culturally appropriate questionnaire for use with a specific community.
e.A researcher measures student achievement quantitatively while using classroom observations to study teaching practices, combining both to understand the learning environment.
- 1 = d,
- 2 = c,
- 3 = e,
- 4 = a,
- 5 = b
Answer
The following table further explains the correct answers.
| Scenario | Purpose | Reason |
|---|---|---|
| A researcher uses interview findings to develop a culturally appropriate questionnaire for use with a specific community. | Development | The qualitative findings are used to develop the quantitative instrument. |
| A researcher finds that quantitative data suggests a programme is effective, but qualitative data reveals participant dissatisfaction. They investigate this contradiction further. | Initiation | The contradiction between quantitative and qualitative findings prompts further investigation and new insights. |
| A researcher measures student achievement quantitatively while using classroom observations to study teaching practices, combining both to understand the learning environment. | Expansion | Different methods address different components of the research (achievement vs teaching practices), extending the breadth of inquiry. |
| A researcher conducts both a survey and interviews about workplace culture, comparing findings to check whether they converge. | Triangulation | Multiple methods study the same phenomenon (workplace culture) to check for convergence. |
| A researcher uses interview data to explain why some schools show better outcomes than others in the quantitative dataset. | Complementarity | The qualitative findings elaborate and explain the quantitative patterns, adding depth and meaning. |
3.4 Choosing a design: a decision framework


