The revision circuit
Introduction to Social Research Methodology
How the circuit works
This exercise revises the whole course through four task stations. Each station is a different kind of work: correcting misleading charts, rewriting a summary, quick-fire methods questions, and exam-style planning. All the material you need for every station is printed on this sheet.
Work in groups of 3–4. Each group starts at a different station and spends 10 minutes there before rotating to the next (A → B → C → D → A). Appoint a scribe at each station; you will need your notes for the wrap-up. If the room does not allow physical rotation, simply work through the stations in order at your table — the timings are the same.
Everything at every station draws on Meridian — the mid-sized Polish retail-and-services company (about 450 employees, 32 stores, a growing e-commerce arm) whose rising turnover (14% to 27% in two years) and falling customer satisfaction your team has been diagnosing all course.
Total time: 5 minutes set-up, 4 × 10 minutes at stations, 10 minutes wrap-up.
Station A: Error correction — the misleading charts (10 minutes)
A colleague has drafted two figures for the board pack. Both are described precisely below; the underlying data are given in the tables.
Figure 1 as drafted
A vertical bar chart titled “Stores dramatically outperform Logistics on engagement”. The y-axis is labelled “Mean engagement score” and runs from 3.0 to 4.0. There are four bars. Nothing on the chart mentions sample sizes, the margin of error, or the fact that the scale runs from 1 to 5.
The data behind it (from the staff engagement survey, 312 respondents):
| Division | Mean engagement (1–5 scale) | n |
|---|---|---|
| Stores | 3.9 | 118 |
| E-commerce | 3.8 | 54 |
| Head office | 3.7 | 61 |
| Logistics | 3.6 | 79 |
Because the axis starts at 3.0, the Stores bar (0.9 units above the axis floor) is drawn 1.5 times the height of the Logistics bar (0.6 units above the floor).
Figure 2 as drafted
A line chart titled “Customer satisfaction in free fall”. The y-axis runs from 40 to 80 on a 0–100 satisfaction index. The solid line connects three data points: June = 74, July = 71, August = 69. From August, a dashed line continues at the same downward slope for twelve more months, ending at 44 the following August, labelled “projected”.
Your tasks
- List every distortion you can find in the two figures — aim for at least four in Figure 1 and four in Figure 2. For each, say how it misleads the board (what false impression a busy reader would take away). (6 minutes)
- Describe the honest versions. For each figure, state in two or three sentences what a corrected chart would look like — axis, form, title, labels — and what impression the board would then (correctly) take away. (4 minutes)
Message in the title, honest axes, no clutter, labels where the eyes are — and remember what a bar’s length is supposed to encode, and how many data points make a trend.
Station B: Writing — the executive summary (10 minutes)
Below is the summary your colleague has drafted for the front page of the Meridian turnover report — about 120 words, methodology first, jargon throughout.
In accordance with the diagnostic mandate conferred by the Board, the research team operationalised a convergent mixed-methods design. A stratified random sample (n = 312; response rate 69%) of Meridian personnel completed a CAWI-administered instrument comprising a nine-item engagement index (Cronbach’s alpha = 0.88) and closed items on remuneration, scheduling, and supervisory support. Concurrently, semi-structured interviews (n = 12) were conducted with recent leavers, transcribed verbatim, and subjected to thematic analysis. Quantitative data were analysed via descriptive statistics and multiple regression. Analyses indicate that scheduling unpredictability and perceived supervisory support are significantly associated with turnover intention among first-year store personnel, whereas remuneration effects were non-significant. Interview data corroborated these patterns. It is recommended that consideration be given to piloting fixed scheduling arrangements.
Your tasks
- Diagnose it first. In one minute, agree on the two or three biggest faults of this text as an executive summary for the board. (1 minute)
- Rewrite it as an executive summary of at most 60 words: findings first, plain language, the numbers that matter, and any recommendation clearly labelled as a recommendation. Count your words. (7 minutes)
- Apply the test. Read your version aloud: could a board member who reads nothing else make a defensible decision from it? Adjust if not. (2 minutes)
Station C: Quick-fire methods (10 minutes)
Eight short scenarios spanning the course. Answer as a group, against the clock — about one minute each. For every question, the scribe records the answer and a one-line reason; an answer without a reason scores nothing in the wrap-up.
- Choose the design. The board wants to know whether a 10% discount voucher actually increases repeat purchases among e-commerce customers. What design gives the most defensible causal answer, and what would it look like?
- Name the sampling strategy. HR wants findings that generalise to all 450 Meridian employees and has a complete, up-to-date staff list broken down by division. What sampling strategy should the team use?
- Spot the ethical problem. A regional manager asks the research team for the named, individual responses of “his” store’s employees from the engagement survey, which was announced to staff as confidential. What is the problem, and what should the team do?
- Qual or quant? The board wants to understand what resigning employees actually experienced in their final months — what the decision to leave felt like from the inside. Which approach, and which method?
- Choose the mode. The team must survey around 2,000 customers quickly and cheaply; Meridian holds a valid e-mail address for every loyalty-programme member. Which survey mode (PAPI, CAWI, or CATI) fits, and what is its main coverage risk?
- Name the sampling strategy. The team wants to interview former employees who left angry and never answered the exit survey. One agrees to talk and offers to introduce two ex-colleagues, who each know others. What is this strategy called, and what is its main limitation?
- Spot the measurement problem. A manager proposes measuring engagement with a single yes/no question: “Are you engaged at work?” Name the problem in the language of the course, and propose something better.
- Correlation or causation? The survey shows that stores with lower engagement have higher turnover. A director drafts a slide saying “low engagement is driving staff out”. May the team approve this wording? If not, what wording — and what further design — would be defensible?
Station D: Exam practice — open questions (10 minutes)
Two open questions in the style of the examination. Answer each as a bullet-point plan — the form you would use to structure a full written answer in the exam. Five minutes each.
Question D1 — plan a small study (5 minutes)
Meridian’s customer satisfaction is falling in stores but is stable online. Plan a small study to diagnose why. Your plan must state: a research question; a design and approach (quantitative, qualitative, or mixed — with a reason); a sampling strategy; the data-collection method(s), including survey mode where relevant; one ethical safeguard; and one limitation of your study.
Question D2 — explain and illustrate a core distinction (5 minutes)
Explain the difference between validity and reliability, illustrating each with an organisational measurement example. Then explain why a measure can be perfectly reliable and still not valid — and why this means both must be established.
Marks follow reasons. “Stratified random sampling” earns little; “stratified random sampling, because the staff list makes probability sampling feasible and stratifying by division guarantees each is represented” earns the marks.
Wrap-up discussion (10 minutes)
Each group reports briefly: the one distortion from Station A other groups may have missed; your 60-word summary from Station B (the class picks the best); any Station C question your group disagreed on; and where your Station D plans diverged. Finish by noting privately which station felt hardest — that is where your revision should begin.