Least-Squares Regression Line: Formula and Interpretation
Connect the least-squares equation to correlation, standard deviations, prediction, residuals, and contextual interpretation.
Direct answer
least squares regression line: The least-squares regression line ŷ=a+bx predicts a response from an explanatory variable by minimizing squared vertical residuals. Interpret slope in context, check residual behavior, and avoid extrapolation beyond the observed x-range.
Quick reference: Least-Squares Regression Line: Formula and Interpretation
| Slope | b1=r(sy/sx) |
|---|---|
| Intercept | b0=ȳ−b1x̄ |
| Prediction | ŷ=b0+b1x |
Least-Squares Regression Line: Formula and Interpretation: complete lesson
The least-squares line predicts y from x
The least squares regression line has the form ŷ=a+bx, where b is the slope and a is the y-intercept. For each observed x, the line produces a predicted response ŷ. A residual is observed minus predicted: e=y−ŷ. Least squares chooses a and b to minimize the sum of squared residuals, Σe².
Because vertical residuals are minimized, the line is asymmetric: the regression of y on x is generally not the algebraic inverse of the regression of x on y. The explanatory and response roles must therefore be chosen from context before fitting or interpreting the model.
Slope is a contextual rate of predicted change
The slope b tells how much the predicted y changes for a one-unit increase in x. If ŷ=42+3.6x predicts exam score from study hours, then each additional study hour is associated with an increase of 3.6 points in the predicted score, on average within the observed range. The wording “predicted” or “associated” is safer than “causes” unless the study design supports causation.
Units are y-units per x-unit. A slope of −2.4 minutes per mile means predicted time decreases by 2.4 minutes for each additional mile only if the variables and sign make contextual sense. Stating units often exposes a reversed-variable interpretation.
The intercept is meaningful only when x=0 is relevant
The intercept a is the predicted response when x=0. Algebraically it is always part of the line, but context determines whether it deserves interpretation. If the data cover houses between 1,000 and 4,000 square feet, the predicted price at 0 square feet is outside the meaningful range and should not be presented as a real-world estimate.
Centering x around a meaningful reference can make an intercept more useful without changing the fitted predictions. For example, use x−18 for age so the intercept represents predicted outcome at age 18 rather than at age 0.
Slope can be calculated from correlation and standard deviations
For the least-squares regression of y on x, b=r(sy/sx). The sign of b therefore matches the sign of r. If x and y are standardized to z-scores, both standard deviations become 1, so the slope of the standardized regression line is exactly r.
The intercept then follows from a=ȳ−bx̄, which means the least-squares line always passes through the point (x̄,ȳ). This provides a useful arithmetic check when summary statistics are given instead of raw data.
Residuals measure prediction errors vertically
A positive residual means the observed response lies above the regression line; the model underpredicted that case. A negative residual means the observation lies below the line; the model overpredicted. Residual magnitude is measured in the units of y.
Residual plots help evaluate linearity and constant spread. A random cloud around 0 supports the linear form. Curvature indicates a systematic pattern left unexplained by the line. A funnel shape suggests changing variability. A single extreme residual can identify an unusual response even when x is ordinary.
r-squared describes variation explained by the linear model
For simple linear regression with an intercept, R²=r². If R²=0.64, about 64% of the variation in observed y-values is accounted for by their linear relationship with x in the fitted model. This is not “64% of y is caused by x,” and it is not the percent of points that lie exactly on the line.
A high R² does not guarantee the line is appropriate. A strongly curved pattern can have a high R² over a restricted range, and an influential point can inflate R². Residual diagnostics and context remain necessary.
Extrapolation is prediction beyond the observed x-range
A regression line summarizes the relationship in the region where data were observed. Predicting far outside that range assumes the same pattern continues, which may be implausible. A linear model for fuel use between 30°F and 80°F should not automatically be projected to −40°F.
Interpolation—prediction inside the observed range—is usually less risky, but it still inherits model error and individual variability. A point prediction ŷ is not a guarantee for one future observation.
Influence combines leverage and residual behavior
A high-leverage point has an unusual x-value and can pull the fitted line toward itself. If it also has a large residual relative to the rest of the pattern, it can substantially alter slope, intercept, correlation, and R². Comparing the fitted model with and without the point is a diagnostic, not an automatic deletion rule.
An observation near x̄ can have a large residual yet limited influence on slope because it lacks leverage. Distinguishing outlying y behavior from x leverage is central to regression diagnostics.
Worked least-squares regression examples
Regression example 1: Study hours and score
Suppose the fitted line is ŷ=52+4.1x. At x=6, the predicted response is ŷ=52+4.1(6)=76.600. The slope is 4.1 response-units per x-unit; in context, a one-unit increase in x is associated with a 4.1-unit increase in the predicted response.
If the observed response is y=81, the residual is y−ŷ=81−76.600=4.400. Because the residual is positive, the observation lies above the fitted line and the model underpredicted this case.
The calculation should be treated as interpolation only if x=6 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 2: Temperature and heating use
Suppose the fitted line is ŷ=120-1.3x. At x=55, the predicted response is ŷ=120-1.3(55)=48.500. The slope is -1.3 response-units per x-unit; in context, a one-unit increase in x is associated with a 1.3-unit decrease in the predicted response.
If the observed response is y=44, the residual is y−ŷ=44−48.500=-4.500. Because the residual is negative, the observation lies below the fitted line and the model overpredicted this case.
The calculation should be treated as interpolation only if x=55 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 3: House size and price
Suppose the fitted line is ŷ=35+0.22x. At x=1800, the predicted response is ŷ=35+0.22(1800)=431.000. The slope is 0.22 response-units per x-unit; in context, a one-unit increase in x is associated with a 0.22-unit increase in the predicted response.
If the observed response is y=450, the residual is y−ŷ=450−431.000=19.000. Because the residual is positive, the observation lies above the fitted line and the model underpredicted this case.
The calculation should be treated as interpolation only if x=1800 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 4: Delivery distance and time
Suppose the fitted line is ŷ=12+1.8x. At x=15, the predicted response is ŷ=12+1.8(15)=39.000. The slope is 1.8 response-units per x-unit; in context, a one-unit increase in x is associated with a 1.8-unit increase in the predicted response.
If the observed response is y=42, the residual is y−ŷ=42−39.000=3.000. Because the residual is positive, the observation lies above the fitted line and the model underpredicted this case.
The calculation should be treated as interpolation only if x=15 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 5: Practice sessions and errors
Suppose the fitted line is ŷ=18-0.9x. At x=10, the predicted response is ŷ=18-0.9(10)=9.000. The slope is -0.9 response-units per x-unit; in context, a one-unit increase in x is associated with a 0.9-unit decrease in the predicted response.
If the observed response is y=7, the residual is y−ŷ=7−9.000=-2.000. Because the residual is negative, the observation lies below the fitted line and the model overpredicted this case.
The calculation should be treated as interpolation only if x=10 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 6: Engine size and mpg
Suppose the fitted line is ŷ=42-5.2x. At x=3, the predicted response is ŷ=42-5.2(3)=26.400. The slope is -5.2 response-units per x-unit; in context, a one-unit increase in x is associated with a 5.2-unit decrease in the predicted response.
If the observed response is y=25, the residual is y−ŷ=25−26.400=-1.400. Because the residual is negative, the observation lies below the fitted line and the model overpredicted this case.
The calculation should be treated as interpolation only if x=3 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 7: Advertising and sales
Suppose the fitted line is ŷ=80+6.5x. At x=12, the predicted response is ŷ=80+6.5(12)=158.000. The slope is 6.5 response-units per x-unit; in context, a one-unit increase in x is associated with a 6.5-unit increase in the predicted response.
If the observed response is y=170, the residual is y−ŷ=170−158.000=12.000. Because the residual is positive, the observation lies above the fitted line and the model underpredicted this case.
The calculation should be treated as interpolation only if x=12 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 8: Elevation and pressure
Suppose the fitted line is ŷ=101.3-0.01x. At x=1200, the predicted response is ŷ=101.3-0.01(1200)=89.300. The slope is -0.01 response-units per x-unit; in context, a one-unit increase in x is associated with a 0.01-unit decrease in the predicted response.
If the observed response is y=88.7, the residual is y−ŷ=88.7−89.300=-0.600. Because the residual is negative, the observation lies below the fitted line and the model overpredicted this case.
The calculation should be treated as interpolation only if x=1200 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 9: Age and reaction time
Suppose the fitted line is ŷ=180+3.2x. At x=20, the predicted response is ŷ=180+3.2(20)=244.000. The slope is 3.2 response-units per x-unit; in context, a one-unit increase in x is associated with a 3.2-unit increase in the predicted response.
If the observed response is y=252, the residual is y−ŷ=252−244.000=8.000. Because the residual is positive, the observation lies above the fitted line and the model underpredicted this case.
The calculation should be treated as interpolation only if x=20 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 10: Rainfall and yield
Suppose the fitted line is ŷ=18+1.1x. At x=25, the predicted response is ŷ=18+1.1(25)=45.500. The slope is 1.1 response-units per x-unit; in context, a one-unit increase in x is associated with a 1.1-unit increase in the predicted response.
If the observed response is y=49, the residual is y−ŷ=49−45.500=3.500. Because the residual is positive, the observation lies above the fitted line and the model underpredicted this case.
The calculation should be treated as interpolation only if x=25 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 11: Screen time and sleep
Suppose the fitted line is ŷ=9.1-0.42x. At x=5, the predicted response is ŷ=9.1-0.42(5)=7.000. The slope is -0.42 response-units per x-unit; in context, a one-unit increase in x is associated with a 0.42-unit decrease in the predicted response.
If the observed response is y=6.4, the residual is y−ŷ=6.4−7.000=-0.600. Because the residual is negative, the observation lies below the fitted line and the model overpredicted this case.
The calculation should be treated as interpolation only if x=5 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Regression example 12: Mileage and used-car value
Suppose the fitted line is ŷ=28-0.00022x. At x=60000, the predicted response is ŷ=28-0.00022(60000)=14.800. The slope is -0.00022 response-units per x-unit; in context, a one-unit increase in x is associated with a 0.00022-unit decrease in the predicted response.
If the observed response is y=13.9, the residual is y−ŷ=13.9−14.800=-0.900. Because the residual is negative, the observation lies below the fitted line and the model overpredicted this case.
The calculation should be treated as interpolation only if x=60000 lies within the data’s observed x-range. The fitted line summarizes average linear behavior; an individual case can differ substantially from ŷ even when the model is useful overall.
Prediction-diagnostics laboratory
Prediction-diagnostics laboratory 1: school survey
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this school survey, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 1: school survey,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 1: school survey,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 2: public-health study
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this public-health study, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 2: public-health study,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 2: public-health study,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 3: manufacturing process
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this manufacturing process, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 3: manufacturing process,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 3: manufacturing process,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 4: transportation system
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this transportation system, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 4: transportation system,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 4: transportation system,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 5: consumer study
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this consumer study, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 5: consumer study,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 5: consumer study,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 6: environmental monitoring
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this environmental monitoring, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 6: environmental monitoring,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 6: environmental monitoring,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 7: sports analysis
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this sports analysis, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 7: sports analysis,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 7: sports analysis,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 8: education program
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this education program, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 8: education program,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 8: education program,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 9: service operation
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this service operation, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 9: service operation,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 9: service operation,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 10: technology experiment
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this technology experiment, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 10: technology experiment,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 10: technology experiment,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Prediction-diagnostics laboratory 11: community poll
For each fitted line, state slope units, assess whether the intercept is meaningful, calculate at least one residual, and inspect whether the requested x-value lies inside the observed range before trusting the prediction. In this community poll, write the target in words before using notation so the calculation remains tied to the variable, population, or model actually being studied.
A high-quality solution should also include a self-check tailored to this topic: compare the sign, units, boundary, probability range, or design logic with what the scenario makes plausible. If the numerical output contradicts that check, revisit the setup before changing the conclusion. In “Prediction-diagnostics laboratory 11: community poll,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Finally, distinguish what the statistical method establishes from what the study design does not establish. The strongest response gives the numerical or graphical evidence and then limits the claim to the population, process, association, or legacy-enrichment scope justified by the data-generating mechanism. In “Prediction-diagnostics laboratory 11: community poll,” apply this check to the named variables and numerical direction rather than treating it as a reusable sentence from another exercise. Check slope units, residual sign, and the observed x-range before interpreting a fitted prediction or using the line outside its data-supported region.
Least-Squares Regression Line: Formula and Interpretation: 22 multiple-choice questions
These items practice slope/intercept interpretation, prediction, residuals, R-squared, influence, and extrapolation limits.
Question 1. Least-Squares Regression Line
For lesson completion at a digital learning platform in Prairie District during a spring 2027 pilot, x̄=33.89, ȳ=67.754, sx=7.649, sy=10.797, and r=0.812. Find the least-squares line and predict y at x=27.82.
Answer: C
b1=r(sy/sx)=0.812(10.7977.649)=1.146. b0=ȳ−b1x̄=67.754−1.146(33.89)=28.91. Thus ŷ=28.91+1.146x, and at x=27.82, ŷ=60.797. The slope predicts a 1.146-unit change in y for each one-unit increase in x, within the observed range.
Question 2. Least-Squares Regression Line
For appointment completion at a regional hospital in Pine Ridge during a semester-long cohort study, x̄=52.528, ȳ=89.572, sx=14.034, sy=12.304, and r=0.835. Find the least-squares line and predict y at x=52.371.
Answer: B
b1=r(sy/sx)=0.835(12.30414.034)=0.732. b0=ȳ−b1x̄=89.572−0.732(52.528)=51.118. Thus ŷ=51.118+0.732x, and at x=52.371, ŷ=89.457. The slope predicts a 0.732-unit change in y for each one-unit increase in x, within the observed range.
Question 3. Least-Squares Regression Line
For appointment wait time at a university advising center in Coastal Plains during a monthly quality review, x̄=48.996, ȳ=40.462, sx=7.826, sy=16.288, and r=0.559. Find the least-squares line and predict y at x=53.234.
Answer: D
b1=r(sy/sx)=0.559(16.2887.826)=1.163. b0=ȳ−b1x̄=40.462−1.163(48.996)=-16.541. Thus ŷ=-16.541+1.163x, and at x=53.234, ŷ=45.393. The slope predicts a 1.163-unit change in y for each one-unit increase in x, within the observed range.
Question 4. Least-Squares Regression Line
For security wait time at a regional airport authority in South Harbor during a fall 2026 audit, x̄=24.05, ȳ=64.884, sx=7.334, sy=14.469, and r=0.774. Find the least-squares line and predict y at x=20.715.
Answer: D
b1=r(sy/sx)=0.774(14.4697.334)=1.527. b0=ȳ−b1x̄=64.884−1.527(24.05)=28.16. Thus ŷ=28.16+1.527x, and at x=20.715, ŷ=59.791. The slope predicts a 1.527-unit change in y for each one-unit increase in x, within the observed range.
Question 5. Least-Squares Regression Line
For algebra benchmark completion at a public high school in Desert County during a two-month observation window, x̄=34.136, ȳ=56.516, sx=14.679, sy=15.307, and r=0.598. Find the least-squares line and predict y at x=35.861.
Answer: B
b1=r(sy/sx)=0.598(15.30714.679)=0.624. b0=ȳ−b1x̄=56.516−0.624(34.136)=35.229. Thus ŷ=35.229+0.624x, and at x=35.861, ŷ=57.592. The slope predicts a 0.624-unit change in y for each one-unit increase in x, within the observed range.
Question 6. Least-Squares Regression Line
For program satisfaction at a city recreation department in Desert County during a pre-exam training cycle, x̄=32.777, ȳ=92.175, sx=14.948, sy=9.355, and r=0.611. Find the least-squares line and predict y at x=27.53.
Answer: B
b1=r(sy/sx)=0.611(9.35514.948)=0.382. b0=ȳ−b1x̄=92.175−0.382(32.777)=79.642. Thus ŷ=79.642+0.382x, and at x=27.53, ŷ=90.169. The slope predicts a 0.382-unit change in y for each one-unit increase in x, within the observed range.
Question 7. Least-Squares Regression Line
For sample concentration at a food safety laboratory in Great Lakes during a six-week field trial, x̄=46.73, ȳ=82.415, sx=13.028, sy=9.872, and r=0.657. Find the least-squares line and predict y at x=49.547.
Answer: A
b1=r(sy/sx)=0.657(9.87213.028)=0.498. b0=ȳ−b1x̄=82.415−0.498(46.73)=59.151. Thus ŷ=59.151+0.498x, and at x=49.547, ŷ=83.817. The slope predicts a 0.498-unit change in y for each one-unit increase in x, within the observed range.
Question 8. Least-Squares Regression Line
For vaccination appointment completion at a public health department in Riverbend during a community outreach cycle, x̄=35.589, ȳ=72.025, sx=10.709, sy=12.908, and r=0.559. Find the least-squares line and predict y at x=31.495.
Answer: B
b1=r(sy/sx)=0.559(12.90810.709)=0.674. b0=ȳ−b1x̄=72.025−0.674(35.589)=48.046. Thus ŷ=48.046+0.674x, and at x=31.495, ŷ=69.267. The slope predicts a 0.674-unit change in y for each one-unit increase in x, within the observed range.
Question 9. Least-Squares Regression Line
For lesson completion at a digital learning platform in Cedar Grove during a summer implementation review, x̄=49.9, ȳ=87.037, sx=6.128, sy=16.18, and r=0.839. Find the least-squares line and predict y at x=45.444.
Answer: A
b1=r(sy/sx)=0.839(16.186.128)=2.215. b0=ȳ−b1x̄=87.037−2.215(49.9)=-23.504. Thus ŷ=-23.504+2.215x, and at x=45.444, ŷ=77.166. The slope predicts a 2.215-unit change in y for each one-unit increase in x, within the observed range.
Question 10. Least-Squares Regression Line
For recovery time at a wildlife clinic in North Valley during a spring 2027 pilot, x̄=29.413, ȳ=69.279, sx=5.471, sy=21.494, and r=0.523. Find the least-squares line and predict y at x=29.263.
Answer: B
b1=r(sy/sx)=0.523(21.4945.471)=2.055. b0=ȳ−b1x̄=69.279−2.055(29.413)=8.844. Thus ŷ=8.844+2.055x, and at x=29.263, ŷ=68.971. The slope predicts a 2.055-unit change in y for each one-unit increase in x, within the observed range.
Question 11. Least-Squares Regression Line
For lunch-program participation at a school district in Cedar Grove during a community outreach cycle, x̄=31.35, ȳ=57.671, sx=9.788, sy=19.899, and r=0.676. Find the least-squares line and predict y at x=31.335.
Answer: A
b1=r(sy/sx)=0.676(19.8999.788)=1.374. b0=ȳ−b1x̄=57.671−1.374(31.35)=14.586. Thus ŷ=14.586+1.374x, and at x=31.335, ŷ=57.65. The slope predicts a 1.374-unit change in y for each one-unit increase in x, within the observed range.
Question 12. Least-Squares Regression Line
For daily energy output at a solar installer in Capital Region during a pre-exam training cycle, x̄=45.403, ȳ=96.947, sx=12.997, sy=19.21, and r=0.485. Find the least-squares line and predict y at x=55.038.
Answer: D
b1=r(sy/sx)=0.485(19.2112.997)=0.717. b0=ȳ−b1x̄=96.947−0.717(45.403)=64.4. Thus ŷ=64.4+0.717x, and at x=55.038, ŷ=103.9. The slope predicts a 0.717-unit change in y for each one-unit increase in x, within the observed range.
Question 13. Least-Squares Regression Line
For mobile-deposit adoption at a community bank in Central County during a fall 2026 audit, x̄=50.018, ȳ=76.606, sx=6.009, sy=8.377, and r=0.487. Find the least-squares line and predict y at x=47.396.
Answer: C
b1=r(sy/sx)=0.487(8.3776.009)=0.679. b0=ȳ−b1x̄=76.606−0.679(50.018)=42.648. Thus ŷ=42.648+0.679x, and at x=47.396, ŷ=74.826. The slope predicts a 0.679-unit change in y for each one-unit increase in x, within the observed range.
Question 14. Least-Squares Regression Line
For program satisfaction at a city recreation department in Mountain Region during a spring 2027 pilot, x̄=33.279, ȳ=93.121, sx=7.611, sy=15.862, and r=0.489. Find the least-squares line and predict y at x=37.87.
Answer: B
b1=r(sy/sx)=0.489(15.8627.611)=1.019. b0=ȳ−b1x̄=93.121−1.019(33.279)=59.206. Thus ŷ=59.206+1.019x, and at x=37.87, ŷ=97.8. The slope predicts a 1.019-unit change in y for each one-unit increase in x, within the observed range.
Question 15. Least-Squares Regression Line
For lesson completion at a digital learning platform in Capital Region during a school-year data collection, x̄=49.106, ȳ=94.929, sx=5.512, sy=9.827, and r=0.724. Find the least-squares line and predict y at x=48.574.
Answer: B
b1=r(sy/sx)=0.724(9.8275.512)=1.291. b0=ȳ−b1x̄=94.929−1.291(49.106)=31.544. Thus ŷ=31.544+1.291x, and at x=48.574, ŷ=94.242. The slope predicts a 1.291-unit change in y for each one-unit increase in x, within the observed range.
Question 16. Least-Squares Regression Line
For lesson completion at a digital learning platform in Sunbelt district during a winter readiness review, x̄=28.752, ȳ=54.308, sx=9.988, sy=18.093, and r=0.631. Find the least-squares line and predict y at x=29.531.
Answer: A
b1=r(sy/sx)=0.631(18.0939.988)=1.143. b0=ȳ−b1x̄=54.308−1.143(28.752)=21.443. Thus ŷ=21.443+1.143x, and at x=29.531, ŷ=55.198. The slope predicts a 1.143-unit change in y for each one-unit increase in x, within the observed range.
Question 17. Least-Squares Regression Line
For part diameter at a regional manufacturer in Westview during a pre-exam training cycle, x̄=35.933, ȳ=40.484, sx=12.894, sy=21.126, and r=0.825. Find the least-squares line and predict y at x=42.436.
Answer: B
b1=r(sy/sx)=0.825(21.12612.894)=1.352. b0=ȳ−b1x̄=40.484−1.352(35.933)=-8.087. Thus ŷ=-8.087+1.352x, and at x=42.436, ŷ=49.274. The slope predicts a 1.352-unit change in y for each one-unit increase in x, within the observed range.
Question 18. Least-Squares Regression Line
For trail-use duration at a state park in Riverbend during a multiweek validation study, x̄=53.019, ȳ=61.833, sx=7.902, sy=16.418, and r=0.521. Find the least-squares line and predict y at x=48.95.
Answer: D
b1=r(sy/sx)=0.521(16.4187.902)=1.082. b0=ȳ−b1x̄=61.833−1.082(53.019)=4.441. Thus ŷ=4.441+1.082x, and at x=48.95, ŷ=57.428. The slope predicts a 1.082-unit change in y for each one-unit increase in x, within the observed range.
Question 19. Least-Squares Regression Line
For crop yield at a farm cooperative in New England network during a regional benchmarking study, x̄=45.62, ȳ=88.257, sx=7.98, sy=19.913, and r=0.468. Find the least-squares line and predict y at x=51.074.
Answer: B
b1=r(sy/sx)=0.468(19.9137.98)=1.168. b0=ȳ−b1x̄=88.257−1.168(45.62)=34.981. Thus ŷ=34.981+1.168x, and at x=51.074, ŷ=94.626. The slope predicts a 1.168-unit change in y for each one-unit increase in x, within the observed range.
Question 20. Least-Squares Regression Line
For lunch-program participation at a school district in Prairie District during a semester-long cohort study, x̄=49.171, ȳ=55.704, sx=13.336, sy=19.591, and r=0.762. Find the least-squares line and predict y at x=43.464.
Answer: D
b1=r(sy/sx)=0.762(19.59113.336)=1.119. b0=ȳ−b1x̄=55.704−1.119(49.171)=0.662. Thus ŷ=0.662+1.119x, and at x=43.464, ŷ=49.316. The slope predicts a 1.119-unit change in y for each one-unit increase in x, within the observed range.
Question 21. Least-Squares Regression Line
For appointment wait time at a university advising center in Metro East during a multiweek validation study, x̄=50.222, ȳ=62.56, sx=5.646, sy=17.898, and r=0.751. Find the least-squares line and predict y at x=52.458.
Answer: D
b1=r(sy/sx)=0.751(17.8985.646)=2.381. b0=ȳ−b1x̄=62.56−2.381(50.222)=-57.003. Thus ŷ=-57.003+2.381x, and at x=52.458, ŷ=67.883. The slope predicts a 2.381-unit change in y for each one-unit increase in x, within the observed range.
Question 22. Least-Squares Regression Line
For mobile-deposit adoption at a community bank in New England network during a semester-long cohort study, x̄=35.24, ȳ=58.314, sx=5.909, sy=14.421, and r=0.702. Find the least-squares line and predict y at x=35.407.
Answer: B
b1=r(sy/sx)=0.702(14.4215.909)=1.713. b0=ȳ−b1x̄=58.314−1.713(35.24)=-2.061. Thus ŷ=-2.061+1.713x, and at x=35.407, ŷ=58.6. The slope predicts a 1.713-unit change in y for each one-unit increase in x, within the observed range.
Least-Squares Regression Line: Formula and Interpretation: 6 free-response questions
For each free-response prompt, interpret the fitted coefficients in units, calculate predictions or residuals as requested, inspect model diagnostics, and limit conclusions to the observed range and design.
FRQ set 1: Least-Squares Regression Line
Scenario. For vaccination appointment completion at a public health department in Sunbelt district during a baseline measurement week, x̄=52.904, ȳ=46.958, sx=10.603, sy=16.332, and r=0.683. Find the least-squares line and predict y at x=57.37.
- Define the variable, units, groups, and requested distribution or model feature.
- Show the required calculation or graphical/model reasoning with labeled quantities.
- Interpret the numerical result in the context of the data rather than as an isolated number.
- Identify an unusual feature, limitation, or condition that affects the conclusion.
Model response
b1=r(sy/sx)=0.683(16.33210.603)=1.052. b0=ȳ−b1x̄=46.958−1.052(52.904)=-8.699. Thus ŷ=-8.699+1.052x, and at x=57.37, ŷ=51.656. The slope predicts a 1.052-unit change in y for each one-unit increase in x, within the observed range.
FRQ set 2: Least-Squares Regression Line
Scenario. For security wait time at a regional airport authority in Desert County during a summer implementation review, x̄=57.235, ȳ=44.413, sx=12.334, sy=15.633, and r=0.798. Find the least-squares line and predict y at x=61.087.
- Define the variable, units, groups, and requested distribution or model feature.
- Show the required calculation or graphical/model reasoning with labeled quantities.
- Interpret the numerical result in the context of the data rather than as an isolated number.
- Identify an unusual feature, limitation, or condition that affects the conclusion.
Model response
b1=r(sy/sx)=0.798(15.63312.334)=1.011. b0=ȳ−b1x̄=44.413−1.011(57.235)=-13.477. Thus ŷ=-13.477+1.011x, and at x=61.087, ŷ=48.309. The slope predicts a 1.011-unit change in y for each one-unit increase in x, within the observed range.
FRQ set 3: Least-Squares Regression Line
Scenario. For application processing time at a housing authority in Prairie District during a fall 2026 audit, x̄=48.571, ȳ=68.386, sx=11.359, sy=10.525, and r=0.527. Find the least-squares line and predict y at x=47.282.
- Define the variable, units, groups, and requested distribution or model feature.
- Show the required calculation or graphical/model reasoning with labeled quantities.
- Interpret the numerical result in the context of the data rather than as an isolated number.
- Identify an unusual feature, limitation, or condition that affects the conclusion.
Model response
b1=r(sy/sx)=0.527(10.52511.359)=0.488. b0=ȳ−b1x̄=68.386−0.488(48.571)=44.668. Thus ŷ=44.668+0.488x, and at x=47.282, ŷ=67.757. The slope predicts a 0.488-unit change in y for each one-unit increase in x, within the observed range.
FRQ set 4: Least-Squares Regression Line
Scenario. For appointment completion at a regional hospital in Riverbend during a pre-exam training cycle, x̄=44.845, ȳ=92.071, sx=10.942, sy=9.326, and r=0.627. Find the least-squares line and predict y at x=50.195.
- Define the variable, units, groups, and requested distribution or model feature.
- Show the required calculation or graphical/model reasoning with labeled quantities.
- Interpret the numerical result in the context of the data rather than as an isolated number.
- Identify an unusual feature, limitation, or condition that affects the conclusion.
Model response
b1=r(sy/sx)=0.627(9.32610.942)=0.534. b0=ȳ−b1x̄=92.071−0.534(44.845)=68.106. Thus ŷ=68.106+0.534x, and at x=50.195, ŷ=94.93. The slope predicts a 0.534-unit change in y for each one-unit increase in x, within the observed range.
FRQ set 5: Least-Squares Regression Line
Scenario. For checkout time at a grocery cooperative in South Harbor during a six-week field trial, x̄=31.742, ȳ=40.237, sx=13.523, sy=19.311, and r=0.476. Find the least-squares line and predict y at x=41.672.
- Define the variable, units, groups, and requested distribution or model feature.
- Show the required calculation or graphical/model reasoning with labeled quantities.
- Interpret the numerical result in the context of the data rather than as an isolated number.
- Identify an unusual feature, limitation, or condition that affects the conclusion.
Model response
b1=r(sy/sx)=0.476(19.31113.523)=0.68. b0=ȳ−b1x̄=40.237−0.68(31.742)=18.661. Thus ŷ=18.661+0.68x, and at x=41.672, ŷ=46.987. The slope predicts a 0.68-unit change in y for each one-unit increase in x, within the observed range.
FRQ set 6: Least-Squares Regression Line
Scenario. For weekly material weight at a recycling program in Capital Region during a winter readiness review, x̄=46.789, ȳ=82.15, sx=14.955, sy=8.018, and r=0.61. Find the least-squares line and predict y at x=44.282.
- Define the variable, units, groups, and requested distribution or model feature.
- Show the required calculation or graphical/model reasoning with labeled quantities.
- Interpret the numerical result in the context of the data rather than as an isolated number.
- Identify an unusual feature, limitation, or condition that affects the conclusion.
Model response
b1=r(sy/sx)=0.61(8.01814.955)=0.327. b0=ȳ−b1x̄=82.15−0.327(46.789)=66.848. Thus ŷ=66.848+0.327x, and at x=44.282, ŷ=81.33. The slope predicts a 0.327-unit change in y for each one-unit increase in x, within the observed range.
Continue with the next connected AP Statistics skill
The most useful next step is to connect this topic to a neighboring method rather than repeating the same question type indefinitely.
