Graphs as Models
Read a graph as a working model of how a system behaves, then use a line of best fit to predict inside the data (interpolation) and, more carefully, beyond it (extrapolation).
A graph is more than a picture of numbers: it is a model, a simplified stand-in for how a real system behaves. Once you fit a line or curve to your data, that model lets you do something the raw table cannot: predict values you never measured. Predicting inside the range you measured is called interpolation and is usually safe. Predicting beyond your data is extrapolation, and it is riskier because the pattern may change where you never looked. Epidemiologists model case counts to forecast how far an outbreak might spread, pharmacologists model dose-versus-response curves to choose a safe dose, financial analysts model trends to plan budgets, and climate scientists model temperature data to project future change. Every one of them also has to know the limits of the model, because a line that fits today can mislead tomorrow. Learning to treat a graph as a model, and to respect where the model stops being trustworthy, is one of the most useful science skills you will carry into any career.
- Common Core · HSS-ID.B.6Represent data on two quantitative variables on a scatter plot, and describe how the variables are related; fit a function (such as a line) to the data.
- Common Core · HSS-ID.C.7Interpret the slope (rate of change) and the intercept of a linear model in the context of the data.
- Ohio · Ohio HS S.ID.6Fit a linear model to a scatter plot and use it to reason about the relationship between two measured variables.
- NGSS · SEP-4Analyzing and Interpreting Data: use graphical displays and lines of best fit to identify patterns and make predictions from data.
- AP · AP Bio SP 4 (Data Analysis)Analyze data, including using a trend line or model to predict values and to state the limits of what the model supports.
- Read points on an x-y coordinate graph: You must be able to locate a point by its x and y values before you can read a trend or a prediction off a graph.
- Plot ordered pairs from a data table: A graph model starts as plotted data, so students must turn (x, y) pairs into points.
- Understand slope and the equation y = mx + b: A line of best fit is written as y = mx + b, so predicting values means substituting into that equation.
Prerequisites are inferred: pending teacher review.
Re-learn the skill with worked practice and clear examples.
Treat the graph as a model: fit a line to the data, read its equation as y = mx + b, and use it to predict. Predicting a value inside the data range is interpolation (usually safe); the surviving trend line is what does the predicting.
Using the best-fit model y = 2x + 3, predict the count at x = 4 (which is inside the data). Is this interpolation or extrapolation?
Reviewed- A.y = 8, extrapolation
- B.y = 11, interpolation
- C.y = 11, extrapolation
- D.y = 14, interpolation
Show the worked solution ▾
Answer: B. y = 11, interpolation
- Step 1: Substitute x = 4: y = 2(4) + 3 = 8 + 3 = 11.
- Step 2: Check the range: The data runs from x = 1 to x = 6, and 4 is inside that range, so predicting here is interpolation.
Why it's right: y = 2(4) + 3 = 11, and because x = 4 is inside the measured range of 1 to 6, this is interpolation.
- A: y = 2(4) + 3 = 11, not 8, and x = 4 is inside the data.
- C: The value 11 is right but x = 4 is inside the data, so this is interpolation, not extrapolation.
- D: y = 2(4) + 3 = 11, not 14.
Aligned to Common Core HSS-ID.C.7: predict from a linear model · reading level ~grade 9
The table shows a data point at each hour and the model's prediction from y = 2x + 3. Which hour's measured count is exactly ON the best-fit line?
Reviewed| Hour (x) | Measured count | Model y = 2x + 3 |
|---|---|---|
| 1 | 5 | 5 |
| 2 | 8 | 7 |
| 3 | 8 | 9 |
| 4 | 12 | 11 |
| 5 | 12 | 13 |
| 6 | 16 | 15 |
- A.Hour 2
- B.Hour 1
- C.Hour 4
- D.Hour 6
Show the worked solution ▾
Answer: B. Hour 1
- Step 1: Compare measured to model: A point is on the line when the measured count equals the model value in the same row.
- Step 2: Scan the rows: At hour 1 the measured count is 5 and the model gives 5, so they match. Every other row has a small gap.
Why it's right: At hour 1 the measured count (5) equals the model value (5), so that point lands exactly on the best-fit line while the others scatter above or below it.
- A: At hour 2 the measured count is 8 but the model gives 7, so it is above the line.
- C: At hour 4 the measured count is 12 but the model gives 11, so it is above the line.
- D: At hour 6 the measured count is 16 but the model gives 15, so it is above the line.
Aligned to NGSS SEP-4: compare data to a model · reading level ~grade 9
A best-fit line for a plant's height is h = 1.5d + 4, where d is the day and h is the height in centimeters. Predict the height on day 6.
Reviewed- A.10 cm
- B.11.5 cm
- C.13 cm
- D.15 cm
Show the worked solution ▾
Answer: C. 13 cm
- Step 1: Substitute d = 6: h = 1.5(6) + 4.
- Step 2: Compute: 1.5 x 6 = 9, and 9 + 4 = 13 cm.
Why it's right: h = 1.5(6) + 4 = 9 + 4 = 13 cm.
- A: This uses 1.5 x 4 + 4 (wrong day) or forgets to add 4 correctly.
- B: This adds only part of the intercept or misreads the slope.
- D: This uses h = 1.5(6) + 4 with an extra 2 added, giving 15 instead of 13.
Aligned to Common Core HSS-ID.C.7: apply a linear model · reading level ~grade 9
- A student fits y = 2x + 3 to lab data and predicts the hour-3.5 count by interpolation before the next reading is taken.
- A greenhouse worker uses a best-fit line of height versus day to estimate how tall seedlings will be midway through the week.
- A lab tech compares each measured point to the model line to spot which reading was an outlier.
Fill these in as you work through the lesson.
- Model (a simplified stand-in for a real system):
- Line of best fit (one line closest to all the points):
- Interpolation (predicting inside the data):
- Extrapolation (predicting beyond the data, riskier):
A graph is a of a system; predicting a value inside the measured range is called , and predicting beyond it is called , which is riskier because the pattern may .
- Explain in your own words why interpolation is usually safer than extrapolation.
- Given y = 2x + 3, predict y at x = 4 and state whether that is interpolation or extrapolation.
- Describe one situation where a straight line is the wrong model and a growth curve fits better.
A best-fit model is y = 2x + 3 and the data was measured from x = 1 to x = 6. To predict at x = 5, substitute: y = 2(5) + 3 = ____. Because x = 5 is inside the data range, this prediction is ____ (interpolation or extrapolation).
The vocabulary of this topic, shown in the way you will meet it.