Lecture 20 ARM

Question 1

Mixed methods

Answer 1

Combines quantitative and qualitative approaches within one study

Used to get a more holistic understanding than either method alone

Common in social sciences

Question 2

Why using mix methods?

Answer 2

1) Broader answers - both BREADTH and DEPTH (statistical breadth - narrative depth)
2) Generalisability - adding quantitative data brings larger “N”for external validity
3) Context and depth: Quantitative trends get details from meat on the bones by qualitative evidence
4) Triangulation (credibility): if different methods agree - findings are more credible. The cross-verification is called triangulation
5) Explain or explore - each methods can help explain surprises from the other and help explore what to focus on next.

Question 3

Quant vs Qual vs Mixed

Answer 3

Quant: numbers, statistics, testing hypotheses, large samples, generalizable results, deductive reasoning

Qual: textual, observations, depth, detail, meaning, smaller samples, inductive reasoning

Mixed: integrates both, leverages each other strengths. statistical breadth + narrative depth = comprehensive understanding.

Question 4

Deductive reasoning

Answer 4

Start with theory or hypothesis and test it with data.
Top down approach
Genreal idea –> specific observation.
Prediction, hypothesis, falsification logic
Quantitative

Question 5

Inductive reasoning

Answer 5

Start with specific observations and build to broader patterns or theories.
Verification logic - discovering, developing theory
Qualitative

Mixed uses both - inductive to generate a hypothesis, then deductive testing or vice versa

Question 6

Explanatory sequential design

Answer 6

1. Quantitative data first - analyse first - then qualitative second
2. The qual phase is designed to explain or clarify the quan results - especially useful if numbers are puzzling, in need of context
3. Clear follow up - qual sample or questions are chosen based on the initial quant outcomes

Example: Start with a survey, then conduct interviews or focus groups with some of the survey participants to understand the reasons behind the trend

Question 7

Exploratory sequential design

Answer 7

1. Qualitative first, quantitative second
2. The quant phase is designed to test or generalise insights from the initial qualitative findings
3. Often used to develop a new survey or instrument based on qual data, to see how widespread the discovered pattern is

Example: Start with ethnographic fieldwork or interviews to identify key factors or create a theory: then design a survey or experiment to verify those factors in a larger sample

Question 8

Convergent parallel design

Answer 8

Collecting quantitative and qualitative data at the same time - then you compare and contrast them

Concurrent approach, during the time timeframe

Analyse datasets separately first, then merge or compare results to see agree/differ

Purpose: To triangulate findings for a comprehensive view - convergence can validate results, divergence can reveal new insights

Example: Survey community attitudes while doing ethnographic observation - comparing stats and field notes

Question 9

Challenges to mixing methods

Answer 9

1) Time and workload
2) Expertise - must master both statistical and qualitative analysis - interdisciplinary teams
3) Data integration - conflicting results, makes interpretation harder
4) Paradigm differences-Quant and qual approaches rely on different assumptions (objectivism vs constructivism, positivism vs interpretivism, deductive vs induction)
5) Resource intensive: More data = more costs and managing write up is complex with more data