From question to design
Introduction to Social Research Methodology
The scenario
You are the junior research team advising the board of Meridian, a mid-sized Polish retail-and-services company: around 450 employees, 32 stores, and a growing e-commerce arm. Annual staff turnover has risen from 14% to 27% in two years, customer satisfaction scores are falling, and the board wants evidence, not hunches.
In this exercise you will practise the full move covered in today’s session: from research question to hypothesis, from hypothesis to variables, and from problem to design — ending with the skeleton of a research proposal. Keep your work from Task 3: you will build on it in later sessions.
Task 1: From questions to hypotheses (20 minutes)
Work in pairs. Below are three research questions that your team has already agreed with Meridian’s board. For each question:
- Write one testable hypothesis. It may be directional or non-directional, but it must satisfy the properties discussed in the session: testable and falsifiable, clear and specific, a statement of relationship between named variables, grounded in a plausible rationale, and value-neutral.
- Identify the unit of analysis, the independent variable, the dependent variable, and at least one control variable. For the control variable, add one sentence explaining why it needs to be controlled — what rival explanation does it represent?
The research questions:
- Q1. Does the quality of first-line management affect staff turnover across Meridian’s 32 stores?
- Q2. Is customer satisfaction lower in stores with higher staff turnover?
- Q3. Do employees who complete Meridian’s onboarding programme stay with the company longer than those who do not?
Use this table format for each question:
| Component | Your answer |
|---|---|
| Hypothesis | |
| Unit of analysis | |
| Independent variable | |
| Dependent variable | |
| Control variable(s) + why |
Check before moving on: could your hypothesis be proven wrong by data? If you cannot describe the result that would falsify it, revise it.
Task 2: Matching designs to managerial problems (20 minutes)
Work in the same pairs. Below are four problems raised at the last Meridian board meeting. For each one, choose the most appropriate research design from the session’s typology — cross-sectional, longitudinal, case study, comparative, or experimental/quasi-experimental — and justify your choice in 2–3 sentences. Your justification should say what the design does well for this problem and acknowledge one limitation you would flag to the board.
- Scenario A. The regional director for the south says: “Our two Kraków stores opened in the same year, are the same size, and pay exactly the same rates. One loses a third of its staff every year; the other hardly anyone. I want to understand what separates them.”
- Scenario B. The finance director says: “Pay talks with the unions open in three weeks. They claim that unpaid overtime is now routine on our shop floors. Before I sit down at that table, I need to know how common it really is — store by store and role by role.”
- Scenario C. The operations director says: “Five stores in the north switched to the new shift-swap app in March, because their managers asked for it; the other 27 are still on paper rotas. Absences in the five have dropped since. Before we pay to roll the app out everywhere, I need to know whether it was the app that did it.”
- Scenario D. The HR director says: “We took on 60 new store staff this September, and we know most of our leavers go within their first year. I want to see how their commitment develops over the coming months, and to catch the point where it starts to slip.”
Be ready to defend your choices: after 15 minutes we will compare answers in plenary, and pairs who chose different designs for the same scenario will argue it out. Note that some scenarios can be defensibly answered with more than one design — what matters is the quality of the justification.
Task 3: A one-page proposal skeleton (15 minutes)
Still in pairs. Choose one of the three research questions from Task 1 and sketch a one-page research proposal skeleton for it, using the five headings below. One or two sentences per heading is enough — this is a skeleton, not a full proposal.
| Section | What to write |
|---|---|
| Problem | What is wrong at Meridian, and why is it worth the board’s money to study it? |
| Research question | The question from Task 1, refined if needed |
| Hypothesis | Your hypothesis from Task 1, refined if needed |
| Design | Which design will test it, and in one sentence, why it fits |
| Data needed | What data must be collected or obtained — include both any new data (e.g. a CAWI survey) and existing company records you would use |
When specifying survey data, state the mode (PAPI, CAWI, or CATI) and briefly say why it suits Meridian’s workforce.
Keep this page. You will return to this skeleton in later sessions, when we add conceptualisation and measurement (session 8), sampling (session 10), and data-collection instruments (sessions 11–12) to it.
Wrap-up (5 minutes)
Before we finish, be ready to answer one question as a pair: of the three moves you practised today — writing the hypothesis, assigning the variables, or choosing the design — which was hardest, and why?