What Does Direction Mean?
A scatterplot shows paired quantitative data: each point represents one individual or observational unit, with one variable placed on the horizontal axis and the other on the vertical axis. Once the points are plotted, we can ask whether the vertical values tend to rise, fall, or show no clear overall movement as the horizontal values increase. This overall tendency is the direction of the scatterplot’s pattern.
Direction describes a pattern in the plotted data, not a rule that every point must obey. Real data often include variation: two individuals with similar horizontal values may have different vertical values, and a point may run against the overall tendency. The question is whether the cloud of points, considered as a whole, tends to move upward, downward, or neither from left to right.
The order of the variables matters when describing direction. A positive pattern means “as the horizontal variable increases, the vertical variable tends to increase.” A negative pattern means the vertical variable tends to decrease as the horizontal variable increases. These descriptions refer to the variables as placed on the axes; they are not claims that one variable causes the other to change.
How to Read the Overall Tendency
Imagine scanning the plot from left to right. Do not focus on one point or a small group of points in isolation. Instead, consider where the point cloud tends to sit vertically at smaller horizontal values and compare that with where it tends to sit at larger horizontal values. If the points generally occupy higher vertical positions on the right, the direction is positive. If they generally occupy lower vertical positions, the direction is negative. If there is no consistent overall shift, the pattern is directionless.
A simple way to check your reading is to divide the horizontal range roughly into a left part and a right part, then compare the vertical positions of the points in each part. This is a visual aid, not a formal rule: the split is not a special statistical threshold, and the points need not be evenly divided. It helps keep attention on the overall tendency rather than on whichever point is most striking.
Identify which quantity is horizontal and which is vertical. Direction is always described in terms of what happens vertically as the horizontal value increases.
Compare the general vertical positions of points on the left with those on the right, without expecting every point to follow the same movement.
State whether the vertical variable tends to increase, decrease, or show no clear overall direction as the horizontal variable increases.
Worked Example: A Positive Direction
Suppose a hypothetical instructor records each of eight students’ weekly study hours and quiz scores. Study hours are on the horizontal axis, and quiz score is on the vertical axis. The plotted pairs are (1, 55), (2, 62), (3, 60), (4, 70), (5, 68), (6, 77), (7, 73), and (8, 85).
Read the points from left to right. At the smaller study-hour values, the scores are 55, 62, 60, and 70. At the larger values, the scores are 68, 77, 73, and 85. Some scores dip between nearby points: for example, the score goes from 62 to 60 as study hours increase from 2 to 3, and from 77 to 73 as hours increase from 6 to 7. Those individual decreases do not outweigh the overall tendency for scores to be higher toward the right.
Conclusion: The scatterplot has a positive direction. In context, students with more weekly study hours tend to have higher quiz scores in these hypothetical data. This description concerns an association in the plotted data; it does not establish that studying more caused a higher score.
Worked Example: A Negative Direction
Imagine a hypothetical set of observations on daily outdoor temperature and household heating use. Temperature is on the horizontal axis, and heating use, measured in kilowatt-hours, is on the vertical axis. The plotted pairs are (5, 30), (10, 27), (15, 29), (20, 23), (25, 24), (30, 18), (35, 20), and (40, 14).
As we move right, temperature increases from 5 to 40 degrees. The heating-use values do not decrease at every step: use rises from 27 to 29 kilowatt-hours between 10 and 15 degrees, and from 18 to 20 between 30 and 35 degrees. But the point cloud is generally higher on the left and lower on the right. The larger temperatures are associated with lower heating-use values overall.
Conclusion: The scatterplot has a negative direction. In context, higher outdoor temperatures tend to occur with lower household heating use in these hypothetical observations. The direction alone does not show that temperature is the only factor affecting heating use or prove a causal relationship.
When There Is No Clear Direction
A directionless plot does not have a consistent overall upward or downward tendency from left to right. The vertical values may be mixed across the horizontal range: high and low points appear at both smaller and larger horizontal values, without the cloud generally shifting up or down.
“Directionless” is a visual description of direction, not a claim that the variables are unrelated in every possible way. A pattern might have a more complicated shape that does not produce a simple upward or downward tendency. It is also possible for a real but subtle tendency to be hard to see in a particular plot. For this reason, report what the plotted pattern shows rather than making a broader claim that the variables have no relationship.
Worked Example: A Directionless Pattern
Consider hypothetical data on shoe size and reading score for eight students. Shoe size is on the horizontal axis, and reading score is on the vertical axis. The plotted pairs are (5, 72), (6, 84), (7, 68), (8, 90), (9, 75), (10, 81), (11, 70), and (12, 85).
Scores do not generally rise as shoe size increases: the score is 90 at shoe size 8, then 75 at shoe size 9. Nor do they generally fall: the score rises from 75 at size 9 to 81 at size 10, then falls to 70 at size 11 and rises to 85 at size 12. High and low scores are spread across the horizontal range, with no clear overall upward or downward tendency.
Conclusion: For these hypothetical observations, the scatterplot is directionless. This means there is no clear overall direction in the plotted points; it does not prove that shoe size and reading score are independent in all populations or under all ways of measuring them.
Why Direction Is Useful
Describing direction is an efficient first step in understanding paired quantitative data. A clear sentence about what tends to happen to the vertical variable as the horizontal variable increases helps summarize the broad pattern without pretending that the data follow an exact rule. It also makes comparisons between plots easier: two plots may show opposite directions even when they involve different variables or units.
Direction is only one feature of a scatterplot. It does not tell us how closely the points follow the tendency, and it does not tell us whether the pattern follows a straight or curved shape. Those are separate questions. Keep the description focused: first say whether the overall tendency is upward, downward, or unclear; do not infer other features from direction alone.
The way the variables are assigned to the axes also matters. If the horizontal and vertical variables are exchanged, an upward tendency remains upward and a downward tendency remains downward, but the sentence describing the direction must name the new horizontal variable first. For example, “higher temperature tends to go with lower heating use” describes the same negative association as “higher heating use tends to go with lower temperature.” Neither sentence identifies a causal direction.
Common Mistakes
- Expecting every point to follow the direction. The reasoning “one point goes down, so the plot cannot be positive” gives too much weight to a local change. The correction is to judge the overall point cloud; a positive pattern can include some decreases between neighboring points.
- Reversing the verbal description. Saying “as the horizontal variable increases, the vertical variable decreases” describes a negative direction, not a positive one. Always name which variable increases and what the other tends to do.
- Calling a plot directionless because it is scattered. Points can vary noticeably and still show a clear overall upward or downward tendency. Direction asks about the general movement, not how tightly points cluster.
- Equating directionless with no relationship. A plot with no overall upward or downward tendency may still have a more complicated pattern. State only that no clear overall direction is visible.
- Turning association into causation. “The vertical variable tends to be higher when the horizontal variable is higher” describes the plot. “Increasing the horizontal variable makes the vertical variable higher” makes a causal claim that the plot alone does not establish.
Check Your Understanding
For each question, focus on the overall direction and describe it using the variables in context.
- A hypothetical plot places daily exercise time on the horizontal axis and resting heart rate on the vertical axis. The points tend to be lower on the right than on the left. What is the direction, and how would you describe it in context?
- In a positive scatterplot, one point has a lower vertical value than the point immediately to its left. Does this contradict the positive direction? Explain.
- A plot has high and low vertical values across both small and large horizontal values, with no visible upward or downward tendency. What direction should you report, and what should you avoid claiming?
- Why is “as the horizontal variable increases, the vertical variable tends to decrease” a description of negative direction?
- Does a positive direction by itself show that the horizontal variable causes the vertical variable to increase? Explain.