Eighth Grade Mathematics: Fitting an Informal Trend Line (Line of Best Fit)
When bivariate data displays a clear linear association, mathematicians and scientists draw a straight line through the scattered points to model the relationship and make quantitative predictions. This modeling line is called a line of best fit, or a trend line. While professional statisticians use advanced calculus and least-squares regression algorithms to find the mathematically optimal line, eighth grade students learn to fit an informal line of best fit using visual and geometric balance. In this chapter, you will master the criteria for drawing a valid trend line, learn to assess visual fit, and recognize poorly placed lines.
The Purpose and Criteria of a Trend Line
A line of best fit is a straight line drawn through the center of a cloud of data points on a scatter plot that best represents the underlying linear trend:
The Three Golden Rules of an Informal Trend Line:
1. Follow the Trend:
The line must capture the general direction (positive or negative slope) and steepness of the points.
2. Equal Balance of Points:
The line should have roughly an EQUAL NUMBER of data points lying above the line
and below the line.
3. Minimize Overall Distances (Residuals):
The line should pass as close as possible to all data points simultaneously,
threading directly through the center of the cloud rather than along an edge.
Visualizing a Well-Fitted Trend Line:
y
10 | * /
8 | * / *
6 | * / *
4 | * / *
2 | / *
0 +-----------------------> x
0 2 4 6 8 10 12
The line threads cleanly through the middle with equal points above and below!
A common misconception is that a trend line must connect the very first data point to the very last data point, or that it must pass through the origin (0, 0). A line of best fit does not need to touch any specific data points; its job is to capture the overall collective trend of the entire swarm.
Evaluating Poorly Fitted Trend Lines
To develop a sharp visual eye for trend lines, you must learn to identify the three classic mistakes made when drawing informal lines:
Classic Trend Line Flaws:
Flaw 1: Tilted Incorrectly (Wrong Slope)
The line is too steep or too shallow, failing to match the angle of the swarm.
Flaw 2: Offset Upward or Downward (Wrong Intercept)
The line follows the slope but sits along the top edge or bottom edge of the swarm,
placing almost all points on one side of the line.
Flaw 3: Skewed by an Outlier
The line bends or shifts unnaturally to chase an extreme isolated outlier,
sacrificing its fit across the majority of the data points.
Flawed Trend Lines:
Too High (Offset): Wrong Slope (Too Steep):
y y
| / | /
| / * * * | / *
| / * * * | / *
| / * * | / *
+-----------------> x +-----------------> x
All points BELOW line! Cuts across the cloud at wrong angle!
A valid trend line balances points above and below across the entire length of the line, from the lowest x-values to the highest x-values.
Using a Clear Ruler and Two Reference Points
The most effective technique for drawing an informal trend line by hand is using a transparent plastic ruler:
The Ruler Balancing Method:
Step 1: Lay a clear ruler across the scatter plot so you can see all data points beneath it.
Step 2: Adjust the angle and position until roughly half the points lie above the ruler
and half lie below, evenly distributed along the line.
Step 3: Draw a single continuous straight line through the entire scatter plot.
Step 4: Identify TWO CLEAR GRID INTERSECTIONS lying directly on your drawn line
(do not pick data points unless the line crosses them exactly!).
Identifying two clean grid intersections on your drawn line is essential because those two coordinates will allow you to write the algebraic equation of your trend line in the next chapter.
Chapter Exercises
Exercise 1: A student looks at a scatter plot with 20 data points showing a positive linear association. The student connects the point with the smallest x-value to the point with the largest x-value using a straightedge. Explain in paragraph form why connecting the first and last data points is often a poor method for constructing a line of best fit.
Exercise 2: A scatter plot contains 16 data points showing the relationship between hours of sunshine and solar panel energy production. A proposed trend line has 14 points lying below the line and only 2 points lying above the line. Evaluate whether this line is a valid line of best fit, explaining what adjustment must be made.
Exercise 3: Explain in paragraph form how an investigator should handle a severe outlier when deciding where to place an informal line of best fit on a scatter plot.
Solutions and Explanations
Solution 1: Connecting the first data point to the last data point is often a poor method for constructing a line of best fit because individual data points are subject to natural variation, measurement noise, and random fluctuations. If either the first point or the last point happens to sit unusually high or unusually low compared to the rest of the dataset, connecting them anchors the line to two potentially unrepresentative extremes. Doing so forces the line to tilt at an inaccurate angle that ignores the collective pattern formed by the remaining 18 points in the center of the cloud. A true line of best fit must reflect the average trajectory of all 20 data points combined, threading through the geometric center of the data cloud rather than being dictated by two arbitrary edge points.
Solution 2: The proposed trend line is not a valid line of best fit because it fails the fundamental criterion of balance. A valid trend line must pass through the middle of the data cloud, having roughly an equal number of points lying above the line and below the line across its entire length. In this proposal, 14 out of the 16 data points lie below the line, meaning that over 87 percent of the data is on one side. This indicates that the line is offset too high (its y-intercept is too large), significantly overestimating the solar energy production for almost every recorded value of sunshine. To correct the line, it must be shifted vertically downward parallel to its current slope until approximately 8 points lie above and 8 points lie below.
Solution 3: When deciding where to place an informal line of best fit in the presence of a severe outlier, an investigator should focus on modeling the general relationship that governs the vast majority of the data points, rather than allowing the single outlier to distort the model. Because an outlier represents an unusual anomaly, a data entry error, or an unrepeatable physical exception, forcing the trend line to move toward the outlier creates a model that fits neither the main group nor the outlier well. Best statistical practice requires fitting the line of best fit to the primary cluster of representative data points, while clearly identifying and noting the outlier separately in the written analysis. This preserves the predictive accuracy of the linear model for normal operations while highlighting the anomaly for further scientific investigation.