ARCH9090 Chap.5 Choosing and Combining Methods
Choosing and Combining Methods
Method choice shapes who participates, what can be learned and how confidently conclusions can be drawn. This chapter uses purpose, participants, depth, resources and risk to compare qualitative and quantitative approaches. It explains breadth and depth, validity and reliability, then shows how mixed methods add value only when their roles complement each other and integration is planned before collection.
What this chapter covers
- 01
The decision and information need as the starting point for method choice
- 02
Purpose, participants, depth, resources and risk as selection criteria
- 03
Quantitative techniques for scale, frequency, location and relationships
- 04
Qualitative techniques for experience, meaning, explanation and unexpected issues
- 05
Validity, reliability, credibility and transparency as different quality tests
- 06
Sequential mixed methods that connect explanation, comparison and integration
Worked example · free
Combine depth and breadth for an evening service review
- 1Use service and observation data to identify when and where use falls. This mark allocation belongs to an AskSia-authored practice task and is not a published University marking scheme.
- 1Recruit contrasting users and non-users for interviews that explore transport, safety, hours and services.
- 1Build a later questionnaire from surfaced reasons to compare patterns within the reached sample.
- 1Add targeted recruitment because current-user channels miss people who already avoid the branch.
- 1Integrate observed patterns, reported explanations and sample limits in the finding.
Key terms
- Method fit
- The alignment between a technique, the decision, the information need, participants, resources and risk.
- Breadth
- Coverage across more people, places or categories through consistent questions or measures.
- Depth
- Detailed contextual understanding of experience, meaning and reasoning from flexible inquiry.
- Credibility
- The degree to which an interpretation is visibly supported by relevant evidence.
Choosing and Combining Methods FAQ
Are quantitative methods more reliable than qualitative methods?
Not automatically. Consistency can support reliability, but a standardised method may consistently measure the wrong concept. Qualitative work can be systematic and credible when recruitment, questioning, interpretation and evidence are transparent. Quality must be judged against the method’s purpose.
When should mixed methods be used?
Use them when different evidence types answer connected parts of the decision. Interviews may surface reasons that shape a questionnaire, or a questionnaire may identify contrasting cases for interviews. Collection alone is insufficient; the analysis must compare and integrate the evidence.
What is a common method mismatch?
Using a closed survey alone to understand motivation is weak because its options constrain explanation. One focus group is also weak for estimating prevalence. Start from the information need, then choose a route that gives the required depth or breadth with an honest inference boundary.
How should resource limits change a method plan?
Narrow the scope, stage the work or qualify the intended claim rather than keeping an ambitious design and cutting consent, recruitment or analysis quality. If depth matters most, focus on carefully recruited conversations and avoid prevalence claims. If comparison matters, use consistent questions and be honest about the reached sample.
Explain the trade-off to the client and record which evidence remains exploratory, so a smaller design does not silently inherit the claims of a larger one. Document the stopping rule so any later method request must identify the distinct uncertainty it will resolve and the participant burden it creates.
Assessment move
Create a method comparison matrix with columns for information produced, likely participants, strengths, exclusions, resource demand, ethical risk and inference limit. Populate it with interviews, questionnaires, observation, mapping and workshops from the course. Then practise with decision prompts: explain why one method is weak, select a better direction and justify whether another method adds complementary evidence.
Learn the quality terms by applying them, not reciting definitions. Ask what could make a questionnaire invalid, an interview interpretation incredible or a process opaque. For mixed designs, draw the sequence and write exactly how the first stage changes the second.
In assessment writing, use method language carefully: compare, explore, estimate within the sample and interpret are often more accurate than prove, represent or demonstrate causation. Create three versions of one design under different resource ceilings. Preserve the information need while narrowing reach, depth or sequence transparently; do not pretend the smallest version supports the same claim.
For every proposed combination, complete the sentence: evidence from the first method will change the second by. If nothing follows, the methods may be parallel rather than integrated. Practise locating bias before collection by asking who feels safe in the setting, who encounters the recruitment route and which answers the format makes easy.
After drafting an analysis plan, work backward to the instrument and confirm that the data required can actually be produced. Compare validity with reliability using concrete failures: consistent but leading questions may be reliable in administration yet invalid for the intended concept.
End each method recommendation with a boundary statement describing the people, context and decision to which the finding can reasonably apply.