BUS 210 Week 5 Plan

Three pillars, one arc, six steps

Put these on the board in the first five minutes and leave them there. Every case below is the same walk through them.

NarrativeThe arc: tension, conflict, climax, resolution. Without it, the audience loses interest and forgets (slide 5).

DataA few data points that show what is at stake, so the claim is more than an opinion.

VisualsA chart matched to the question and to the audience's data-graphics literacy (slides 24–25).

QuestionThe conflict or problem

AudienceWho is affected, what is at stake

3 data pointsThe evidence

ConnectionCause, effect, suggestion and alternatives

One-pagerQuestion, 3 data points, suggestion

RefineIterate: question, analysis, narrative

Say the definitions out loud once: a data story is a structured narrative that explains how elements connect and leads to a resolution. Data storytelling translates complex analysis into terms that inform a decision or action. A data-driven story blends numbers, visuals and narrative to communicate insights effectively.

Run sheet

0–3	Hook	Ask: "You want something expensive from your parents. Which works better: I train a lot, or one chart?" Take two answers. That's storytelling with data.
3–8	The recipe	Board: three pillars, the arc (slide 6) and the six steps. Use Hawking (slide 8) in one sentence per arc stage as the non-data example.
8–25	Case 1: the race bike	Walk the six steps (details below). At minute 18, give pairs 3 minutes to rewrite the vague question so it's specific and answerable, then take two answers.
25–42	Case 2: the pre-math pilot	Same six steps, new audience. At minute 33, pairs take 3 minutes to name one alternative to "make it compulsory" with one pro and one con.
42–45	Exit question	"Same recipe, two audiences: what changed in the charts we chose, and why?" Expected answer: the audience sets the level of technical detail and visual sophistication.

Case 1 · Session 1, 8–25 min

"Why I need a new race bike"

A student who races bicycles wants to persuade his parents to buy him a new bike. Low technical audience, high personal stakes.

TensionHe trains six days a week and money is tight at home.

ConflictHis times improved for months, but now they're flat on climbs while his training stays the same.

ClimaxHis competitors with lighter bikes gain about the same time on every hill.

ResolutionA mid-range bike, on a deal: he keeps his training hours, or the bike gets sold.

Vague question: "Shouldn't I get a better bike?"
Refined question: "Will a bike 1.5 kg lighter cut my 8 km club-climb time enough to move me into the top 5 of my age group this season?" It's specific, it interests the audience, and the reasoning is plausible (less weight means less energy uphill).
Audience: His parents. They care about cost, commitment and safety. They read simple charts, so show one idea per chart, with no raw 180-day logs.
Descriptive Q: "How many hours did I train per week over the last six months?"
Exploratory Q: "Do club riders who train more finish faster? Where does my gap come from: fitness or the climbs?"
Answerability: He has no data on himself riding the new bike yet. Use direct data where he can, such as a test ride on the same climb, and be honest when he relies on indirect data like manufacturer claims (slide 22).
Alternatives: A new bike, new wheels only (cheaper, gets most of the weight saving), a second-hand bike, or waiting a season. Give one pro and one con for each. He recommends one.

Data point	Chart	Why this chart
Weekly training hours, last 26 weeks	Line chart	One variable over time shows that he trains regularly.
Monthly time on the club climb	Line chart with the plateau annotated	Shows progress, then the flat stretch. The annotation tells parents where to look (slide 25).
His race times vs. 8 competitors	Bar chart, sorted, his bar highlighted	Compares a categorical variable (the rider). It's easy to read.
Training hours vs. race time for club riders	Scatter plot	Two numerical variables changing together. Keep this one for himself, because it may be too technical for his parents.

Ask the class: Which of the four charts would you drop for this audience, and why? (Usually the scatter plot, since it's too technical. That shows how the audience shapes the visuals.)

Quiz link: specificity, answerability, plausible framework, descriptive vs. exploratory questions, the audience setting the level of technical detail.

Case 2 · Session 1, 25–42 min

"Make pre-math part of the curriculum"

Math teachers ran a pilot pre-math course in the summer. Pre-course and post-course surveys and later grades suggest it worked. They want the dean and the department heads to adopt it. Expert audience, institutional stakes.

TensionMany Business freshmen arrive afraid of math and struggle in their first math course.

ConflictThe curriculum is full, a course costs money, and skeptics will say only motivated students volunteered.

ClimaxConfidence jumps from pre- to post-survey, and participants' first-term grades are higher.

ResolutionAdopt it as a curriculum course, with a plan to keep measuring.

Vague question: "Was the summer course good?"
Refined question: "Did the 3-week summer pre-math course raise students' math confidence and their first-semester math grades compared with students who didn't take it?"
Audience: The dean and the department heads. They're data literate and short on time. They care about cost, workload and evidence. They can handle more technical detail, so a comparison of distributions is fine here.
Descriptive Q: "What share of students rated their confidence 4 or 5 out of 5, before and after?"
Exploratory Q: "Do students with bigger confidence gains also earn higher grades?"
Answerability: Grades of a comparison group are needed, not just enrollment counts. Slide 22 makes the same point about online learning: use performance data, not enrollment figures.
The skeptic's point: Volunteers may differ from everyone else. This is sampling bias from Quiz 1. A strong story raises this objection itself and answers it, for example: "Next summer, offer it to everyone below the placement score and compare."
Alternatives: A compulsory credit course, an elective, a course only for students below a placement score, an online self-study module, or no change. Give a pro and a con for each, then the teachers' recommendation.

Data point	Chart	Why this chart
Confidence (1–5 Likert), pre vs. post	Side-by-side or stacked bar chart	Distribution of a categorical (ordinal) variable at two points in time.
First-semester math grades, participants vs. non-participants	Two histograms or box plots	Distribution of a numerical variable. This audience can read box plots.
Confidence gain vs. grade	Scatter plot	Two numerical variables changing together. Answers the exploratory question.

Ask the class: Compare with Case 1. Why is a box plot fine here when the scatter plot was too much for the parents? (The audience is different: it sets the level of technical detail and visual sophistication.)

Quiz link: audience and technical detail, answerability, actionable insights (slide 23), chart for a numerical vs. a categorical distribution.

Run sheet

0–11	Case 3: carbon emissions	This is the M2-1 example as a full one-pager. Students check it against the six steps, then critique one chart (details below).
11–24	From story to proposal	Map the recipe onto the proposal sections. Walk through the example PDF (5 min). Pairs spend 4 minutes drafting their own question and audience on paper, then you cold-call two pairs.
24–32	Rapid review	Quick-fire questions using slide 27 and the review list below. Hands up, no slides.
32–33	Quiz logistics	Phones away, one tab only, 12 minutes.
33–45	Quiz	Start an on-screen timer.

Case 3 · Session 2, 0–11 min

"What can we do about carbon emissions?"

The worked example from teaching note M2-1. It's the model one-pager students should be able to copy.

TensionGlobal average temperature is rising.

ConflictEmissions come from many sectors, so where should action start?

ClimaxThe last 50 years show the fastest rise, and a few sectors dominate emissions.

ResolutionTarget the high-emission sectors with specific actions.

Question: "What actions can we take to reduce carbon emissions and slow global warming?"
Audience: Policy makers, community leaders and the general public. Mixed literacy, so keep the charts familiar.
3 data points: Emissions by sector, temperature and CO₂ over the past century, and the estimated impact of actions such as renewables and less fossil fuel.
Visuals: A line chart of temperature over time with the last 50 years highlighted, and a pie chart of emissions by sector annotated with the suggested action for each sector.
Conclusion: Targeting high-emission sectors with concrete actions can slow further warming.

Critique (2 min): With 7+ sectors, would a sorted bar chart beat the pie? And to show that CO₂ and temperature move together, which chart from slide 24 fits? (A scatter plot.) Tie this to slide 25: not all visualizations are good. Keep the data minimal, the guides clear and the chart matched to the audience.

Quiz link: the full story structure, chart-type matching (histogram, bar/pie, line, scatter), the three pillars.

From story to proposal

The proposal is the first draft of a data story. It summarizes the chosen dataset (source and key aspects) and the objectives, and goes to you for approval. It's due in 12 days: Sunday 15 November 2026.

Proposal section	Story element	In the example (sleep and stress)
Title as a question	The question or conflict	"How do occupations influence stress levels and sleep patterns?"
Introduction	Tension: why it matters	Work pressure disrupts sleep, and sleep affects health.
Data overview	Data pillar	374 observations, 13 variables, with variables named. Ask students to state each variable's type (Quiz 1: nominal or numerical).
Data source	Credibility	A Kaggle link
Objective	Audience and resolution	Recommendations for professionals, researchers and health enthusiasts
Data points (3–4)	Sub-questions, each with a chart	Occupation → stress (bar chart and box plot), stress → sleep quality (scatter plot and correlation), stress groups → duration (box plot), activity → stress (regression line)
Conclusion	Expected resolution	Targeted support for high-stress jobs

What the example does well

The title is a question, and every data point is a sub-question.
It names a chart for each sub-question and says why.
It names the audience explicitly.
It states the size of the dataset (observations × variables).

What to push students to do better

Survey data shows association, not cause. Prefer "is associated with" to "affects".
Make the question specific: which occupations, and which sleep measure.
Check answerability before choosing a dataset: do the variables exist?
Name alternatives or recommendations, not just "patterns".

Approval checklist (read it aloud): a public dataset with a working link · at least 100 observations and a mix of categorical and numerical variables · a specific, answerable question · a named audience · 3–4 data points, each with a chart type · about 2 pages. Suggested sources: Kaggle, TÜİK, World Bank, Our World in Data.

What students should be able to answer

Three pillars of data storytelling?Narrative, data and visuals.

Main consequence of poor storytelling?It fails to engage or enlighten, so the audience loses interest and forgets.

The arc, in order?Tension, conflict, climax, resolution.

Iterative refinement?Repeatedly refining the question, the analysis and the narrative until the question is clear and answerable.

Descriptive vs. exploratory?Descriptive summarizes ("What is the average income?"). Exploratory looks for patterns or relationships ("Does education relate to income?").

Answerability?The data needed (market data, indicators) is available and sufficient to answer the question.

Specificity?The question is precise and leads to clear, actionable insights.

The five traits of a good question?Interest to the audience, not yet answered, a plausible framework, answerability, specificity.

Audience and visual design?The audience decides the level of technical detail and visual sophistication.

Which chart?Numerical distribution: histogram. Categorical: bar or pie chart. Change over time: line chart. Two variables together: scatter plot.