Design the study
Introduction to Social Research Methodology
Aim
This exercise puts today’s session to work. You will design a complete two-phase mixed-methods study for a real organisational problem, justify it in writing to a sceptical audience, and then critique another pair’s design against the same standards. The skills being tested are exactly the ones from the lecture: choosing a design that fits the question, marking a genuine integration point, and writing a justification that answers the five questions — fit, sequence, priority, integration, feasibility.
The brief
The moment has arrived. After two years of watching staff turnover climb from 14% to 27%, Meridian’s board has approved a budget for a proper study of the turnover problem. You are the junior research team, and the board’s instruction is short: “Find out how bad it is, why it is happening, and what would make people stay. You have six months and one external research budget. Don’t bring us two reports — bring us one answer.”
What you already know:
- Meridian employs ~450 people across 32 stores and a growing e-commerce arm.
- The HR dashboard reports turnover monthly, by store and role, and can report it by tenure band.
- Exit interviews exist but are unstructured, patchy, and conducted only with employees who volunteer for them.
- Store managers vary widely in style; several stores have far worse turnover than others, but nobody has checked whether the difference is statistically meaningful.
- The board meets quarterly and will expect an interim report at three months.
Task 1: design a two-phase mixed-methods study (20 minutes, pairs)
With your partner, design a two-phase mixed-methods study of Meridian’s turnover problem. Choose one of the three core designs from today’s session — convergent (QUAN + QUAL), explanatory sequential (QUAN → qual), or exploratory sequential (QUAL → quan) — and commit to it. Write your design on one page, covering all of the following:
- The design, named, with its notation — including which strand has priority (capitalisation is a decision, not decoration).
- Phase 1: the method (be specific — if it is a survey, name the mode: PAPI, CAWI, or CATI), the participants (who, how many, how selected), and rough timing (which weeks of the six months).
- Phase 2: the same three elements — method, participants, timing.
- The integration point, clearly marked. State exactly where and how one phase informs, builds, or is merged with the other: at which step, by what procedure, producing what output (for example: “phase-1 results select the phase-2 sample and set the interview guide” or “a joint display compares survey scales with interview themes”). An arrow on a diagram is not enough — describe the mechanism.
Constraints to respect: the study must fit the six-month window including analysis and reporting, must be feasible for a 450-person company (you cannot interview everyone, and store staff need manager approval for release time), and must produce something presentable at the three-month interim board meeting.
Task 2: the justification (10 minutes, pairs)
Write a five-sentence justification of your design, addressed to the board. One sentence per question, in this order:
- Fit — why this question needs both a quantitative and a qualitative strand (name the two components of the board’s question).
- Sequence — why your phases run in this order, or why they run in parallel (what does the arrow — or the plus sign — depend on?).
- Priority — which strand leads, and why that follows from what the board needs to decide.
- Integration — where the strands meet, and what the integration produces.
- Feasibility — why this fits six months, one budget, and a 450-person company.
Write for the board, not for a methods examiner: plain language, no jargon left unexplained. But every sentence must do real justificatory work — “this will give a richer picture” earns zero.
Task 3: swap and critique (15 minutes, pairs → fours)
Swap your design and justification with another pair. Read theirs carefully, then assess it against the checklist below. For each item, mark pass, weak, or fail, and write one line of evidence for your verdict.
Critique checklist
| # | Check | The question to ask |
|---|---|---|
| 1 | Sequence justified? | Could phase 2 have been fully specified before phase 1 reported? If yes, the sequence is a schedule, not a logic. (For convergent designs: is there a stated reason why neither phase needs the other first?) |
| 2 | Real integration or a staple-job? | Can you point to the exact place where one strand changes what the other does or means? Would either phase change in any way if the other were deleted? |
| 3 | Feasible for Meridian? | Does it fit six months including analysis? Is the sample realistic for ~450 employees across 32 stores? Does it respect release-time and re-contact constraints? Is there something to show the board at three months? |
| 4 | Priority coherent? | Does the declared priority match the design — does the leading strand actually get the larger share of time, sample, and budget? |
| 5 | Justification specific? | Do all five sentences make checkable claims, or do any of them retreat into “richer picture” territory? |
When both pairs have finished, join up as a four and give each other your verdicts — spend roughly three minutes per design. Where you marked weak or fail, propose the smallest change that would fix it.
Plenary comparison (10 minutes)
Each four nominates one design from its pairing to present to the class in ninety seconds: the design’s name and notation, the integration point, and the one critique it received that most improved it. As we compare designs, keep two questions in view:
- Different pairs will have chosen different designs for the same problem — and several may be defensible. What does each choice assume about what the board most needs to know first?
- Which failed checklist items came up most often across the class? The most common failure in real mixed-methods research is item 2 — was it here too?
Before you leave
In one sentence each, write down: (a) the design your pair chose and the single strongest sentence of your justification; (b) one thing your critics found that you had not seen. Keep both — when we reach research reporting later in the course, you will see that the justification you wrote today is a miniature of the methods section you will one day have to write.