Scroll Stats Project Guide
Bivariate Data Project
Scroll Stats: Age vs. Screen Time
Name: ___________________________
Date: ______________ Period: ______
Project Mission
Does age influence how much time people spend scrolling on social media? In this project, you will analyze sample survey data from 14 individuals to model the relationship between age (years) and average daily social media screen time (hours).
Deliverables
- 1. Variable setup & hypothesis
- 2. Labeled scatter plot & trend line
- 3. Linear equation \(y = mx + b\)
- 4. Interpretation & predictions
Sample Dataset: Social Pulse Survey (\(n = 14\))
Use all 14 ordered pairs for your analysis
| ID |
Age (yrs) |
Screen Time (hrs/day) |
| #1 |
13 |
4.8 |
| #2 |
15 |
4.5 |
| #3 |
17 |
5.1 |
| #4 |
19 |
3.9 |
| #5 |
22 |
4.2 |
| #6 |
25 |
3.4 |
| #7 |
28 |
3.1 |
| ID |
Age (yrs) |
Screen Time (hrs/day) |
| #8 |
32 |
2.8 |
| #9 |
35 |
2.3 |
| #10 |
39 |
2.5 |
| #11 |
43 |
1.9 |
| #12 |
47 |
1.5 |
| #13 |
51 |
1.7 |
| #14 |
55 |
1.1 |
1 Part 1: Variable Setup & Hypothesis
Independent Variable (\(x\)-axis) & Unit:
Dependent Variable (\(y\)-axis) & Unit:
Explain why you chose which variable is independent and which is dependent:
Hypothesis: Before plotting, predict the direction (positive, negative, or none) and explain why:
2 Part 2: Scatter Plot & Line of Best Fit
Use a ruler for your trend line
Directions: Label both axes (with units) and give your graph a title. Plot all 14 ordered pairs from Page 1. Then draw a straight line of best fit that balances the points above and below the line.
Graph Title:
0 1 2 3 4 5 6 0 5 10 15 20 25 30 35 40 45 50 55 60
3 Part 3: Equation of the Line of Best Fit
Choose two points on your trend line (they do not need to be original data points) to calculate the slope \(m\) and \(y\)-intercept \(b\).
1. Two Points on Your Trend Line:
2. Slope Calculation \(m = \frac{y_2 - y_1}{x_2 - x_1}\):
3. Solve for \(y\)-intercept (\(b\)) using \(y = mx + b\):
4. Final Linear Equation:
4 Part 4: Statistical Interpretation & Predictions
Use complete sentences and units
1. Association Analysis: Describe the direction, form, and strength of the relationship. Identify any potential outliers or clusters.
2. Meaning in Context: Interpret the slope (\(m\)) and \(y\)-intercept (\(b\)) in terms of age and daily social media hours. Does the \(y\)-intercept make sense in real life? Explain.
3. Predictions & Extrapolation: Use your equation to predict daily screen time for a 30-year-old and a 75-year-old. Show your work, then explain which prediction is more reliable.
Calculations (\(x = 30\) and \(x = 75\)):
Reliability Comparison (Interpolation vs. Extrapolation):
4. Critical Thinking: Does this data prove that getting older causes people to use social media less? Suggest one confounding variable that might explain this trend.
Project Grading Rubric (20 Points Total)
Score: _____ / 20
| Criterion |
Exemplary (5 pts) |
Proficient (3–4 pts) |
Developing (1–2 pts) |
| Scatter Plot & Labels |
All 14 points accurate; title, axes, and units clearly labeled. |
1–2 plotting errors or missing title/unit label. |
Multiple plotting errors or missing axis labels. |
| Trend Line & Equation |
Balanced trend line; slope & \(y\)-intercept calculated accurately. |
Reasonable trend line; minor arithmetic error in \(m\) or \(b\). |
Trend line does not fit data or major equation errors. |
| Contextual Meaning |
Accurately interprets association, \(m\), and \(b\) with real-world units. |
Partially interprets \(m\) or \(b\); misses units or context. |
Vague or incorrect interpretation of model parameters. |
| Predictions & Reasoning |
Correct predictions; clearly distinguishes interpolation vs. causation. |
Predictions mostly correct; brief reasoning on reliability. |
Incomplete calculations or confuses correlation/causation. |
Scroll Stats Answer Key
Teacher Resource • Answer Key
Scroll Stats: Age vs. Screen Time
Note: Student trend lines may vary slightly
1 Part 1: Variable Setup & Hypothesis
Independent Variable (\(x\))
Age (in years)
Dependent Variable (\(y\))
Daily Social Media Screen Time (hours/day)
Reasoning: Age is the independent variable because a person's age may help explain or predict how much time they spend on social media, whereas screen time cannot change a person's age.
Sample Hypothesis: Negative association—younger people (teens and young adults) typically spend more hours per day on social media than older adults.
2 Part 2: Completed Scatter Plot & Line of Best Fit
Age vs. Daily Social Media Screen Time (\(n = 14\))
Daily Screen Time (hours) 0 1 2 3 4 5 6 0 5 10 15 20 25 30 35 40 45 50 55 60 Age (years)
3 Part 3: Equation of the Line of Best Fit (Sample Solution)
1. Two Points on Trend Line:
\((15, 4.6)\) and \((55, 1.0)\)
2. Slope Calculation (\(m\)):
\(m = \frac{1.0 - 4.6}{55 - 15} = \frac{-3.6}{40} = -0.09\)
3. Solve for \(y\)-intercept (\(b\)):
\(4.6 = -0.09(15) + b\)
\(4.6 = -1.35 + b \implies b = 5.95\)
4. Final Linear Equation
\(y = -0.09x + 5.95\)
Teacher Reference (Exact Least-Squares Regression): \(\hat{y} \approx -0.091x + 5.91\), with correlation coefficient \(r \approx -0.97\). Accept student equations with \(m \in [-0.11, -0.07]\) and \(b \in [5.5, 6.4]\).
4 Part 4: Statistical Interpretation & Predictions
1. Association Analysis:
There is a strong, negative, linear association between age and daily social media screen time. As age increases, daily screen time decreases steadily. There are no major outliers, though point #3 \((17, 5.1)\) sits slightly above the trend.
2. Meaning in Context:
Slope (\(m = -0.09\)): For each additional year of age, daily social media screen time decreases by about \(0.09\) hours (roughly \(5.4\) minutes) per day.
\(y\)-Intercept (\(b = 5.95\)): The model predicts a newborn (age \(0\)) spends \(5.95\) hours/day on social media. This does not make sense in real life because infants do not use social media and \(x = 0\) is outside the data range (\(13\text{–}55\)).
3. Predictions & Extrapolation:
Age 30: \(y = -0.09(30) + 5.95 = \mathbf{3.25\text{ hrs}}\)
Age 75: \(y = -0.09(75) + 5.95 = \mathbf{-0.80\text{ hrs}}\)
The prediction for a 30-year-old is much more reliable because \(30\) lies inside the observed domain of \(13\text{–}55\) (interpolation). Age \(75\) is outside the data range (extrapolation) and yields an impossible negative screen time.
4. Critical Thinking (Correlation vs. Causation):
No—observational data shows correlation, not causation. Possible confounding variables include free time / work & family responsibilities, generational comfort with technology, or lifestyle differences.