What the Slope Says
In “Reading the Equation of a Regression Line,” you learned that \(b\) is the slope in the sample regression equation \(\hat{y}=a+bx\). Identifying \(b\) is only the first step. To interpret it, explain what happens to the predicted response when the predictor increases by one unit, using the variables’ names and measurement units.
The slope describes the direction and rate of change in the values predicted by the regression line. A positive slope means the predicted response increases as the predictor increases. A negative slope means the predicted response decreases. The size of the slope gives the amount of predicted change for a one-unit increase in \(x\).
The phrase “one-unit increase” must match how \(x\) is measured. If \(x\) is measured in days, one unit means one day. If \(x\) is measured in hundreds of kilometers, one unit means 100 kilometers. The context—not the number of decimal places in the equation—tells you what a unit represents.
The units of the slope follow from comparing a change in response with a change in predictor: response units divided by predictor units. For example, if the response is plant height in centimeters and the predictor is time in days, the slope is measured in centimeters per day. Those units help you avoid mixing up which variable is changing and which variable is being predicted.
This rate can also be seen directly from the regression equation. If \(x\) increases by one, the predicted response changes from \(a+bx\) to \(a+b(x+1)\). Subtracting the first prediction from the second leaves \(b\). That is why the coefficient of \(x\) gives the change in the line’s prediction for a one-unit increase in \(x\).
This interpretation is about predictions from the fitted line. It does not say that every individual case will change by exactly \(b\) response units, or that changing \(x\) causes the response to change. The slope summarizes how the line’s predicted response varies with the predictor. As discussed in “Why Correlation Does Not Imply Causation,” an association by itself does not establish a cause-and-effect relationship.
A Sentence Pattern for Interpreting Slope
A reliable sentence names both variables, states what an increase in the predictor means, and reports the predicted change in the response with its units. Keep the slope’s sign in mind: a negative slope means the predicted response goes down, not up.
For a negative slope, use the positive size of the decrease in the sentence: “decreases by 1.7 degrees Celsius,” rather than “decreases by \(-1.7\) degrees Celsius.” The negative sign already tells you the direction. You may also say “changes by \(-1.7\) degrees Celsius,” but “decreases by 1.7 degrees Celsius” is often clearer.
An interpretation should not omit the word “predicted.” For instance, “plant height increases by 0.8 centimeters per day” could sound like a statement about every plant’s actual growth. A more precise sentence says “the predicted plant height increases by 0.8 centimeters for each additional day.” This wording keeps the claim attached to the regression line.
Translate a one-unit increase in \(x\) into the actual measurement, such as one day or one kilogram.
A positive slope means the predicted response increases; a negative slope means it decreases.
Use the magnitude of the slope and name the unit of \(y\). Make clear that the response is predicted.
Worked Examples: Slope in Context
Worked Example: Plant Height and Time
A fictional community garden models the height of seedlings, \(y\), in centimeters, using the number of days since planting, \(x\). The fitted line is \(\hat{y}=6+0.75x\). Interpret the slope in context.
The slope is the coefficient of \(x\), so \(b=0.75\). Here, one unit of \(x\) is one day, and the response is measured in centimeters. The slope’s units are centimeters per day. Since the slope is positive, the line’s predicted height increases as the number of days increases.
A precise interpretation is: For each additional day since planting, the predicted seedling height increases by 0.75 centimeters. This describes the predictions from the fitted line; it does not guarantee that each seedling actually grows by exactly 0.75 centimeters every day.
Worked Example: Cooling Water
A fictional science club models the temperature of water in a container, \(y\), in degrees Celsius, using the time since the container was filled, \(x\), in hours. The fitted line is \(\hat{y}=78-2.4x\). Interpret the slope.
The slope is \(b=-2.4\). One unit of \(x\) is one hour, and the response unit is degrees Celsius, so the slope is measured in degrees Celsius per hour. The negative sign indicates that the predicted temperature decreases as time increases. The amount of the decrease for one additional hour is 2.4 degrees Celsius.
A precise interpretation is: For each additional hour since the container was filled, the predicted water temperature decreases by 2.4 degrees Celsius. Do not say that the temperature “decreases by \(-2.4\) degrees”; the decrease is 2.4 degrees, and the slope’s negative sign indicates its direction.
Worked Example: Recycling Collected
A fictional neighborhood project uses the number of collection bins, \(x\), to predict the mass of recyclable material collected, \(y\), in kilograms. Its fitted line is \(\hat{y}=5.3+3.6x\). Write an interpretation of the slope.
The coefficient multiplying \(x\) is \(3.6\), so the slope is positive. One unit of \(x\) means one additional bin. Since \(y\) is measured in kilograms, the slope’s units are kilograms per bin. Thus, the fitted line’s predicted mass rises by 3.6 kilograms for each additional bin.
A complete interpretation is: For each additional collection bin, the predicted mass of recyclable material collected increases by 3.6 kilograms. The intercept \(5.3\) is not part of this slope interpretation; the sentence describes how the predicted response changes, not the prediction at a particular number of bins.
Worked Example: Predictor Units That Are Larger Than One
A fictional trail group models the number of maintenance hours needed, \(y\), from the length of a trail section, \(x\), measured in units of 10 kilometers. Its fitted line is \(\hat{y}=4+1.8x\). Interpret the slope in the original context.
The slope is \(1.8\), but one unit of \(x\) represents 10 kilometers—not one kilometer. The response is measured in hours, so the slope’s units are hours per 10 kilometers. The positive slope means the predicted maintenance time increases as the trail section gets longer.
The contextual interpretation is: For each additional 10 kilometers of trail, the predicted maintenance time increases by 1.8 hours. Saying “for each additional kilometer” would misstate the predictor’s unit. If expressing the same rate per kilometer, divide \(1.8\) hours by 10: \(1.8/10=0.18\) hours per kilometer. The regression equation, however, uses \(x\) in 10-kilometer units, so the direct interpretation should make that clear.
Common Mistakes and AP Exam Tips
- Leaving out the predicted response. “Temperature decreases by 2.4 degrees per hour” can sound like a claim about every observation. A full interpretation says “the predicted water temperature decreases.”
- Giving the wrong direction for a negative slope. If \(b=-2.4\), the prediction decreases as \(x\) increases. Do not report an increase just because the magnitude is 2.4.
- Keeping a negative amount after saying “decreases.” Say “decreases by 2.4 degrees,” not “decreases by \(-2.4\) degrees.” The direction and amount should agree.
- Mixing up the units. The slope is in response units per predictor unit. If \(y\) is kilograms and \(x\) is bins, use kilograms per bin—not bins per kilogram.
- Ignoring how the predictor is recorded. If \(x\) is in 10-kilometer units, one unit means 10 kilometers. State that actual change, rather than calling it one kilometer.
- Turning an association into a causal claim. A regression slope describes how predictions vary with \(x\). Unless the study design supports a causal conclusion, do not say that increasing \(x\) will cause the response to change.
For full-credit communication, use a sentence that includes the predictor’s one-unit increase, the predicted response, the direction, the amount, and the response units. Check the units before you write: the numerator comes from \(y\), and the denominator comes from \(x\). Then confirm that the direction matches the sign of the slope.
Check Your Understanding
For each fitted line, identify the slope’s direction and write a contextual interpretation with the correct units.
- A fictional orchard predicts the mass of apples harvested, \(y\), in kilograms, from the number of trees, \(x\): \(\hat{y}=12+8.5x\). What does the slope mean?
- A fictional bakery predicts the minutes needed to prepare an order, \(y\), from the number of items, \(x\): \(\hat{y}=14+1.2x\). State the slope’s units and interpret it.
- A fictional lake-monitoring team models water clarity, \(y\), in centimeters, using the number of days after a storm, \(x\): \(\hat{y}=21-0.6x\). Interpret the slope without writing a negative amount after “decreases.”
- A fictional delivery model predicts fuel used, \(y\), in liters, from distance \(x\), measured in 100-kilometer units: \(\hat{y}=3+7x\). What actual distance does one unit of \(x\) represent, and how should the slope be interpreted?
- In a regression of response \(y\) in dollars on predictor \(x\) in hours, the slope is \(-4\). What are the slope’s units, and what direction does it describe for the predicted response?