Qualitative and Mixed-Methods Research

Counts show a pattern—not its meaning

Suppose above-no-wake passages cluster on crowded weekend afternoons.

Is that confusion about markers, deliberate resistance, safety during crossing, habit, or something else?

Qualitative inquiry asks how people experience and interpret the situation.

What you should be able to do

• Match a qualitative question to an appropriate method and cases.
• Apply four standards of qualitative rigor.
• Explain why positionality and reflexivity belong in design.
• Name how quantitative and qualitative strands integrate.

Qualitative data preserve expression and context

Qualitative research collects and interprets language, actions, interactions, and social artifacts.

It develops rich, contextual understanding of lived experience, meaning, and social process—not a numerical estimate for a population.

Use it when depth is the need

Qualitative approaches are well suited when:

• little is known;
• perspectives are complex or varied;
• context and nuance matter;
• cases are small, distinctive, or hard to reduce to numbers; or
• the question asks how or why people make meaning.

Depth is not statistical generalization

A purposive interview sample can reveal an unanticipated access problem and explain how it operates.

It cannot establish what percentage of all bay users share that experience.

The design earns credibility through relevant variation and transparent logic, not by pretending cases were randomly selected.

Four standards define trustworthiness

Dependability makes the process visible

Document recruitment, interview guides, changes, field decisions, coding, and analysis so others can understand and critique what was done.

Consistency does not require freezing an iterative design; it requires a clear record of what changed and why.

Credibility tests the interpretation

Strategies include prolonged engagement, persistent observation, triangulation, peer debriefing, and member checking.

Member checking asks contributors whether an interpretation fairly represents their views; it does not turn every participant into a veto holder.

Confirmability traces claims to evidence

Maintain transcripts or field notes, a codebook, analytic memos, decision logs, and examples connecting themes to data.

The goal is not a claim of value-free neutrality. It is transparent reasoning that allows others to see how the interpretation was produced.

Transferability gives the user context

Describe the people, place, time, relationships, and conditions richly enough that another person can judge whether insights apply elsewhere.

The research user makes that judgment. Transferability is not statistical generalizability from a representative sample.

Positionality locates the researcher

Positionality includes identities, experiences, roles, commitments, and relationships that shape what a researcher notices, asks, and accepts as true.

MDWP’s lead owns a shoreline home and supports restrictions. That fact neither automatically invalidates nor disappears from the study.

Reflexivity is an ongoing practice

Researchers critically examine how their position, assumptions, power, and relationships shape recruitment, interaction, analysis, and reporting.

Tools include reflexive journaling, peer debriefing, positionality statements, and explicit records of analytic decisions.

Qualitative design still begins with a question

A strong design connects:

purpose → question → methodology → cases → method → analysis → quality standard → claim

“Do interviews” is no more complete than “send a survey.”

Select information-rich cases

Non-probability sampling is strategic, not careless.

Select people or artifacts with relevant knowledge, experiences, positions, or variation. Document inclusion logic and recruitment pathways.

One organization-appointed speaker cannot stand for every member of a public.

Interviews explore individual experience

Semi-structured interviews use a guide but allow follow-up and unexpected meaning to emerge.

Use an interview funnel: begin broadly, move toward specific experiences, and listen actively. Memo immediately afterward and explain follow-up.

Focus groups reveal interaction

Facilitated discussion can show agreement, disagreement, memory, norms, and power dynamics as participants respond to one another.

They are poorly suited when confidentiality or hierarchy prevents candid talk. Separating groups can reduce—but not erase—those risks.

Observation records practice in context

Qualitative observation asks what happens, how, where, with whom, and what it appears to mean in a natural setting.

Record concrete actions and context before interpretation. “Hostile” is a conclusion; raised voices, interruptions, and gestures are evidence.

Archives preserve—and omit—social traces

Meeting minutes, complaint narratives, signs, news, photographs, and public comments can reveal language and institutional history.

They were produced for other purposes, may omit marginalized experiences, and usually cannot answer follow-up questions.

Analysis is iterative interpretation

Researchers organize data, code segments, compare cases, develop categories, test themes against contrary evidence, memo decisions, and refine claims.

Software can store and retrieve text. It does not take responsibility for the interpretation.

Mixed methods assigns distinct jobs

Mixed methods integrates quantitative and qualitative approaches in one study or program.

The justification is not “more data is better.” Each strand should answer a different part of the problem, and the design must say where they meet.

Four integration logics

• Convergent: collect separately; compare or combine.
• Explanatory sequential: QUAN first; QUAL explains patterns.
• Exploratory sequential: QUAL first; QUAN tests or extends what emerged.
• Embedded: a smaller strand sits within a dominant design.

Choose a Willow Bay sequence

Observation shows when above-no-wake passages cluster. Interviews then ask people using those periods how they interpret markers, congestion, access, and fairness.

Which mixed-methods logic is this, and where does integration occur?

Check the sequence

It is explanatory sequential if the quantitative observation results guide the later interviews and the qualitative findings explain those patterns.

Integration occurs when observation results shape sampling or questions and again when both strands are interpreted together.

Triangulation is not a vote

Different sources may converge, complement one another, or contradict.

A contradiction can reveal different definitions, contexts, or weaknesses. Do not count methods and let “two out of three” win without examining why evidence differs.

Prepare for Friday

Your design must align:

question → design → method and cases → quality protection → integration → bounded claim

Read the case definitions carefully. Support, compliance, participation, registered owner, and bay user name different things.