Transcripts
Part 3: You will hear two students discussing ideas for a statistics assignment and deciding what data to use.
Nora: James, have you decided what to do for the statistics assignment? I keep reading the brief and it still feels a bit open-ended.
James: Not really. I’ve got a shortlist, but every option has a drawback. They keep stressing appropriate methods and limitations, so I want something we can defend without advanced modelling.
Nora: Same. I don’t want a topic where the numbers look good but the definitions are fuzzy, because then we end up writing a glossary instead of an analysis.
James: Exactly. One main dataset, a clear question, and then we justify the data, or explain its weaknesses properly. And we need to show we understand bias, not just calculate averages.
Nora: Right. First idea: sports participation, gym attendance, team sports, that sort of thing.
James: Manageable. We wouldn’t need complicated modelling, just comparisons and trends, descriptive stats and a simple chart. But we’d have to say that membership isn’t the same as attendance.
Nora: Fine. Second: local transport usage, bus numbers or cycle counts.
James: I like it, but the figures may not be consistent. Councils count differently, and sometimes they move counting points or change sensors, so the trend can be an artefact rather than real change.
Nora: True, and then we spend ages explaining gaps. Third: air quality, particulate readings.
James: Mm, that does sound interesting, but it can get complicated.
Nora: Interesting, but we haven’t been given much guidance. The handout was mostly general comments, not clear variable choices, and it didn’t explain how to handle seasonal patterns.
James: Yes, more discussion notes than instructions. I’d worry we pick the wrong measure and realise too late.
Nora: Fourth: housing rents. I found a public dataset with monthly medians by area.
James: That’s promising. Supporting information is easy to find, government dashboards, local reports, even newspapers, and the variables are usually defined clearly. It’s also easy to explain to the marker what the numbers represent.
Nora: Fifth: diet and nutrition surveys, like food diaries.
James: That worries me. People forget what they ate, estimate portions badly, or give socially better answers. Different surveys also ask differently, so the figures might not be dependable.
Nora: So housing rents seems safest.
James: Agreed, but we need a specific question, say, how rents vary by distance to transport, comparing two neighbouring districts. Then we can add one transport indicator as context, but only if the data is clean.
Nora: And we need limitations beyond just small sample. In my district, the biggest day-to-day complaint is parking, especially near the station and the main shopping street.
James: Parking? I thought you’d say crime.
Nora: Crime comes up sometimes, but parking is constant. We can use it as context, not a main variable, and we can say we’re using resident feedback to interpret the figures.
James: OK. We should add one positive change too, so it doesn’t sound like a list of problems.
Nora: They improved the café area by the station, actually, that’s just refurbishment. The real new facility is a small cinema next to the station, only two screens, but it’s brought more evening footfall.
James: Right, mention it briefly as a possible influence, but be careful not to claim it caused rent rises.
Nora: Redevelopment is another issue. An old warehouse site is being converted into flats, and during the works the area gets covered in litter, packaging, debris, bits of insulation.
James: Great limitation point. The dataset might show rent changes but not short-term disruption like mess and reduced street cleanliness. We can also mention that advertised from prices can distort early figures.
James: For an external case, Hammarby Sjöstad in Stockholm is often cited for its waste system and recycling model. We can reference it briefly, as long as we cite one reliable overview source.
Nora: Fine. Keep it short and factual.
James: And we need a local proposed development: Eastbrook Quays.
James: Was that the one where people complained the buildings were too tall?
Nora: No, not height. I’m pretty sure the controversy was about flood risk, the floodplain mapping and river levels, but I’ll check council documents so we don’t guess.
James: Perfect. I’ll handle the main dataset. You research Eastbrook Quays and summarise the objections. Then we can meet tomorrow, lock the methods, and start writing before we drown in charts.
