Week 2 — Sociology-in-Action Workshop · "Read the Data"

Course: Introduction to Sociology (SOC 1) · Silver Oak University (fictional sample) · Prof. Adeyemi
Objective: Objective 2 — read social data critically (what it measures, over what population, what it shows and doesn't) and distinguish correlation from causation · SLO B (reason from and evaluate evidence)
Worth 50 points · Sociology Workshops group = 15% of the grade · Workshop 2
Mode this week: data interpretation (this is the course's signature read-the-data workshop; other weeks alternate with observation/reflection). No special tools — just a browser and an approved chatbot.

This is the course's signature weekly component. Every instructional week has one Sociology Workshop. Some weeks you'll read real social data (a chart or table from the Census, Pew, BLS, the World Bank, or Our World in Data); other weeks you'll observe and reflect on your own social world. Either way you'll end by catching an AI's mistakes. All external resources are links — nothing to buy or download.

Part 1 — The Big Picture

This week you learned how sociologists gather evidence — and the discipline that keeps them honest: a number describes a pattern, and two things moving together is a clue, not a verdict. This workshop puts that to work on one real statistic. You'll read it the way a sociologist does — slowly, asking exactly what it measures and what it can't tell us — and then run a correlation-vs-causation drill so the most expensive mistake in social science never sneaks past you.

The guiding question: When a source reports a number, what does that number actually show about society — and what does it NOT show?

Part 2 — The Figure (read it at the source)

We'll read one clean, current statistic about American life. It has been verified at the source for this workshop — but the whole point is that you confirm it yourself.

THE FIGURE: About nine-in-ten — 91% — of U.S. adults own a smartphone.
Source: Pew Research Center, "Mobile Fact Sheet."
What/where/when: % of U.S. adults who say they own a smartphone; based on a Pew survey of 5,022 U.S. adults conducted Feb. 5–June 18, 2025 (fact-sheet page dated Nov. 20, 2025).
🔗 https://www.pewresearch.org/internet/fact-sheet/mobile/
A second figure from the same source (for the drill in Part 4): 16% of U.S. adults are "smartphone dependent" — they own a smartphone but do not subscribe to home broadband — and that share is far higher among lower-income adults (about 34% of those in households earning under \$30,000, versus 4% of those earning \$100,000 or more). (Same Pew survey, 2025.)

Do this first: open the link, find the 91% yourself, and note the year and what's measured. Never repeat a number you haven't seen on the source's own page — that's the whole discipline of this course.

Part 3 — The Read-the-Data Scaffold (fill this in)

Work the four questions for the 91% figure. This is the read-the-data move you'll use all term.

A model is provided in the answer key below — try it yourself first, then check.

Part 4 — Correlation-vs-Causation Drill

Use the second figure: smartphone-dependent adults are 16% overall, but 34% of the lowest-income group versus 4% of the highest-income group.

A careless headline might say: "Being poor makes people glued to their phones," or "Smartphone dependence keeps people in poverty."

1. Both headlines commit the same error. Name it. (One sentence.)
2. Name a plausible third (confounding) variable that could drive both lower income and smartphone dependence. (Hint: what costs money?)
3. Could the arrow run in reverse for either headline? Explain in one sentence why direction matters here.
4. What would it take to actually establish that one of these causes the other? (One sentence — think back to the method built for causation.)

Part 5 — Analysis Questions

Answer in a sentence or two each:
1. In your own words, what is the difference between a statistic that describes a pattern and one that proves a cause? Use the 91% figure.
2. The 91% comes from a sample of about 5,000 adults, not from asking every adult. Why can a well-chosen sample of 5,000 reasonably represent the whole country — and what would make a sample of even 50,000 untrustworthy?
3. The figure is higher for younger and higher-income adults. State one thing this gap does show and one thing it does not show.
4. Pick one way this Pew number could be misused in an argument (e.g., a marketer or a politician citing it) — and how a sociologist would push back.

Part 6 — AI-Critique Moment (required — this is the BYOAI step)

Now bring in your approved chatbot (Gemini, Claude, or ChatGPT) and be the sociologist who checks its work.

1. Ask it: "What percentage of U.S. adults own a smartphone, and is there a study showing that smartphone use causes lower well-being? Give me the statistic and the citation."
2. Check everything it says against a real source:
   - Did it give a smartphone-ownership number? Search for it at Pew (the Mobile Fact Sheet). Does it match what you saw (≈91%, 2025), or did it state a different/outdated/made-up figure?
   - Did it produce a "study" and a citation? Try to find that study. Chatbots fabricate plausible-sounding studies and citations constantly — if you can't locate it, treat it as fabricated and say so.
   - Did it slide from correlation to causation — reporting that smartphone use causes lower well-being, when the evidence (if any) is a correlation (and reverse direction is possible)?
   - Did it overgeneralize — treat a group average as true of every member, or stereotype "people who use phones a lot"?
3. Write 3–4 sentences reporting what the AI got right and at least one thing you had to correct, verify, or flag — a wrong/unverifiable number, a fabricated study or citation, or a correlation reported as a cause. (If it happened to get everything right, say how you verified each claim at the source — that's the skill.)

The habit all term: the tool drafts, you verify. A chatbot will confidently invent a statistic, fake a citation, or call a correlation a cause — catching it is the point, and this is the week it slips the most.

Part 7 — What to Submit

Submit a single document (or text entry) with: your completed read-the-data scaffold (Part 3), your correlation-vs-causation drill answers (Part 4), your Part 5 answers, and your Part 6 AI-critique paragraph. Note in one line that you confirmed the 91% figure on Pew's page. Due Sunday, Sep 13, 11:59 p.m. (50 points).

Instructor answer key & model responses — REMOVE BEFORE PUBLISHING TO STUDENTS

Figure verified live at the source for this build: Pew Research Center, Mobile Fact Sheet (page dated Nov. 20, 2025; data from a survey of 5,022 U.S. adults, Feb. 5–June 18, 2025) — "About nine-in-ten (91%) own a smartphone, up from just 35% in the Center's first survey of smartphone ownership conducted in 2011." The smartphone-dependency figures (16% overall; 34% of households under \$30K vs. 4% of \$100K+) are from the same Pew source (2025), 🔗 https://www.pewresearch.org/internet/fact-sheet/mobile/ and the companion digital-divide short read. All figures were seen on Pew's own pages before shipping.

Model — Part 3 (the four questions on 91%):
- 1. What is measured? The percentage of U.S. adults who self-report owning a smartphone — a share of a population, from a survey (not a count of devices, not "people," but adults).
- 2. Over what population and period? U.S. adults, as measured in a 2025 Pew survey of ~5,022 people, with an overall margin of error of about ±1.9 points. It does not cover children, and it's self-reported.
- 3. Shows / does NOT show: It shows that smartphone ownership is now near-universal among U.S. adults and far above the 2011 level (35%). It does NOT show why people own them, what they do with them, which phones, or how ownership differs within small subgroups beyond what Pew breaks out.
- 4. Correlation or causation? The higher ownership among younger/higher-income adults is a descriptive pattern (a correlation) — it does not establish that income or age causes ownership; many factors could be involved.

Model — Part 4 (drill):
1. Both headlines confuse correlation with causation (treating an association between income and phone dependence as proof of cause).
2. A plausible third variable: household income / the cost of home broadband — lower income makes a \$60/month home-internet plan unaffordable, so a phone becomes the only connection, driving both the low income context and the dependence. (Accept: cost of living, housing, regional broadband availability.)
3. Reverse direction: "smartphone dependence keeps people poor" assumes phones cause poverty, but it's far more likely that being lower-income leads to relying on a phone (can't afford broadband) — the arrow runs the other way. Direction matters because the policy implication flips entirely.
4. To establish a cause, you'd need a controlled study — ideally an experiment (random assignment), or at minimum longitudinal data with third variables held constant — not a single cross-sectional correlation.

Expected answers (Part 5): Q1 a descriptive statistic reports a pattern/prevalence; a causal claim says one thing produces another, which requires ruling out third variables and direction (an experiment). Q2 a representative random sample of ~5,000 mirrors the population within a margin of error, so it generalizes; a self-selected 50,000 (e.g., a click-in poll) can't, because respondents opted in — representativeness beats size. Q3 Shows: a real, documented ownership gap by age/income. Does not show: the cause of the gap, or anything about within-group variation or behavior. Q4 any reasonable misuse (e.g., "everyone has a smartphone, so a phone-only service reaches all customers" — ignores the ~9% without one and the access gaps) with a sociologist's correction.

Part 6 (AI-critique): full credit for a specific catch — most commonly the AI fabricating a "study" and citation about smartphones and well-being, stating a wrong or unverifiable ownership number, reporting a correlation as a cause, or overgeneralizing. Full credit also if the student verified each AI claim against Pew/a real source and reported how.

Grading rubric — 50 points

Quality gate (self-checked): the one published figure used (Pew 91% U.S. adult smartphone ownership, 2025) was verified live at the Pew Research Center Mobile Fact Sheet (pewresearch.org) before shipping, as were the smartphone-dependency figures (16% overall; 34% under \$30K vs. 4% \$100K+, 2025, same source); each is pre-stated with source + year, and students are required to re-confirm the figure at the source. Every part keeps correlation ≠ causation front and center: the scaffold's question 4, the entire Part 4 drill, and the AI-critique all explicitly treat the demographic gaps as descriptive patterns, not causes, and target fabricated statistics/studies, overgeneralization, and correlation-as-causation — the discipline's load-bearing AI risks. No correlation is presented as causation anywhere in this workshop.