A Final Self-Check for Scatterplot Descriptions
In “Free-Response Practice Describing a Scatterplot,” you brought direction, unusual features, form, and strength together in a contextual answer. This tutorial puts those skills to work across several different data sets. The goal is not to memorize a particular description, but to make sure each part of your answer matches the pattern the scatterplot actually shows.
A useful synthesis check is to connect every claim to evidence. Before writing, note what happens to the response as the explanatory variable increases, sketch the overall form in a few words, judge how closely the points follow that form, and identify any important departures. Then check that the parts agree: for example, a point far from a linear pattern may affect how strong the overall association appears.
An Evidence-First Review Routine
DUFS, from “Describing a Scatterplot With DUFS,” is still a helpful organizer. For a final review, make the organizer more specific: write a short piece of evidence beside each feature before composing your answer. This helps prevent a confident-sounding description from drifting away from the data.
Identify what each point represents, the explanatory variable on the horizontal axis, and the response variable on the vertical axis. Include units when they are given.
Move from left to right. Describe whether the response tends to increase, decrease, or show no clear direction, then decide whether the overall pattern is roughly linear, curved, or another shape.
Ask how closely the points follow the overall pattern. Look for outliers, clusters, gaps, or other notable features, and consider whether they affect your strength description.
Use contextual language, locate important unusual points with both coordinates when possible, and check that your direction, form, strength, and unusual-feature statements do not contradict one another.
This is a review routine, not a requirement to write four disconnected sentences. A concise paragraph can cover several features at once. The evidence notes simply make it easier to see whether a claim is supported. For group-coded plots, as in “Lurking Patterns: Time and Subgroups,” add one more check: describe whether the pooled pattern hides meaningful differences among groups.
Worked Example: Stream Temperature and Dissolved Oxygen
Worked Example: Stream Temperature and Dissolved Oxygen
A fictional environmental science class records water temperature and dissolved oxygen concentration at nine locations along a stream. Each point represents one location. The table gives the paired measurements that appear in the scatterplot.
| Location | Water temperature, °C | Dissolved oxygen, mg/L |
|---|---|---|
| 1 | 8 | 11.2 |
| 2 | 10 | 10.8 |
| 3 | 12 | 10.1 |
| 4 | 14 | 9.7 |
| 5 | 16 | 9.1 |
| 6 | 18 | 8.6 |
| 7 | 20 | 8.1 |
| 8 | 22 | 7.5 |
| 9 | 24 | 10.4 |
Question: Synthesize the scatterplot’s direction, form, strength, and unusual features.
Evidence check: Across the first eight locations, dissolved oxygen generally decreases as temperature increases. Those points follow a roughly straight, fairly close downward pattern. The point at \(24\)°C and \(10.4\) mg/L is far above that pattern.
Full response: “For these stream locations, there is a negative, roughly linear association between water temperature and dissolved oxygen concentration: locations with warmer water generally have lower dissolved oxygen. The main pattern is fairly close, although one location is an outlier, with a temperature of 24°C and dissolved oxygen concentration of 10.4 mg/L, well above the pattern. The unusual point makes the overall association less consistent than the other locations suggest.”
The answer describes the overall direction and form, then uses the point’s two coordinates to locate the outlier. It also explains why the outlier matters to the strength description rather than calling the entire scatterplot uniformly strong. The final statement stays descriptive: these paired measurements show an association, not proof that temperature alone caused the observed oxygen levels.
When Strength and Form Need Separate Attention
“Recognizing Linear and Nonlinear Form” and “Judging Strength of an Association” distinguish two questions that are easy to blend together. Form is the shape of the overall pattern; strength is how closely points follow that pattern. A close curve can show a strong association even though the association is not linear. Conversely, a roughly straight pattern can be weak if points are widely scattered around it.
Unusual points deserve a deliberate check as well. In “Spotting Outliers in Bivariate Data,” you learned that being far out on one axis is not by itself enough to make a point a bivariate outlier. Judge the point against the overall pattern, and, when useful, describe its approximate \(x\)- and \(y\)-values. The next example checks whether a description accurately represents a curved pattern without inventing an outlier.
Worked Example: Time Spent Charging and Battery Level
Worked Example: Time Spent Charging and Battery Level
For a fictional technology demonstration, students record the battery level of the same model of handheld device at different times during a charging session. The scatterplot uses minutes since charging began as the explanatory variable and battery level as the response variable.
| Minutes charging | Battery level, percent |
|---|---|
| 0 | 5 |
| 10 | 27 |
| 20 | 46 |
| 30 | 62 |
| 40 | 75 |
| 50 | 85 |
| 60 | 92 |
Question: A student writes, “There is a strong positive linear association, with one outlier at 60 minutes.” What should be revised?
Evidence check: Battery level rises at every listed charging time, so direction is positive. The increases become smaller as charging time grows, so the pattern bends and then levels off rather than following a straight line. The point at 60 minutes continues that curve; it is not far from the overall pattern.
Revised response: “For this device, battery level has a strong, positive, curved association with time spent charging. The battery level rises as charging time increases, but it rises less quickly at later times, so the pattern levels off. No point appears to be an outlier from the curve.”
The revision separates direction, form, and strength. It replaces “linear” with “curved,” describes how the curve bends, and removes the unsupported outlier claim. A point can have a large value on an axis—in this case, the largest charging time—without being unusual relative to the pattern. The response describes the plotted data and does not claim that every device will follow precisely the same curve.
Review Group Patterns Without Losing the Overall Picture
A plot with group labels calls for careful synthesis. You can describe the overall association and still note that groups form separate clusters or follow different patterns. Avoid treating every point as if its group membership were irrelevant. At the same time, do not let a single unusual point erase a visible trend within the rest of a group.
Worked Example: Training Program and Sorting Errors
Worked Example: Training Program and Sorting Errors
A fictional warehouse training team compares hours of practice with the number of sorting errors made during a timed task. Points are marked by Program A or Program B. The table includes one unusual result.
| Program | Practice hours | Sorting errors |
|---|---|---|
| A | 1 | 20 |
| A | 2 | 18 |
| A | 3 | 17 |
| A | 4 | 15 |
| A | 5 | 14 |
| A | 6 | 12 |
| B | 1 | 15 |
| B | 2 | 13 |
| B | 3 | 12 |
| B | 4 | 10 |
| B | 5 | 9 |
| B | 6 | 7 |
| B | 6 | 14 |
Question: Describe the association while accounting for the program labels and the unusual observation.
Evidence check: Within each program, sorting errors generally decrease as practice hours increase, in a roughly linear pattern. The points form two groups: for similar practice hours, most Program B participants have fewer errors than Program A participants. The Program B observation at \(6\) hours and \(14\) errors is well above the rest of that program’s pattern.
Full response: “For these trainees, practice hours and sorting errors have a negative, roughly linear association within each program: participants with more practice generally make fewer errors. The points form two clusters by program, with Program B participants generally making fewer errors than Program A participants at similar practice hours. One Program B result is unusual: the participant with 6 hours of practice made 14 errors, well above the other Program B results. The plot shows these associations but does not establish that practice hours or program membership caused the differences.”
The response includes both the within-program direction and the group separation. It does not let the unusual Program B result stand in for the whole group, and it locates that result using both variables. This is more informative than describing only the pooled pattern or saying simply that the plot has “an outlier.” As in “Lurking Patterns: Time and Subgroups,” group membership can add useful context to the description.
Common Mistakes and AP Exam Tips
- Using a direction label without context. “Negative” alone does not say which variables move together. A complete statement names both variables and tells how the response tends to change as the explanatory variable increases.
- Calling a curve linear because it rises or falls. Direction does not determine form. Check whether the pattern is roughly straight or bends, and describe the bend when it is noticeable.
- Calling a point an outlier just because it is extreme on an axis. Judge its distance from the overall pattern. Give approximate values for both coordinates when the plot or table makes them available.
- Ignoring the effect of an outlier on strength. Describe the main pattern, but consider whether a point far from it makes the association less consistent overall. Do not describe the points as uniformly close when one plainly departs from the pattern.
- Leaving out group labels. If the plot distinguishes programs or other categories, look for clusters and compare the groups at similar explanatory-variable values. A pooled summary may hide an important feature.
- Confusing association with causation. Describe what tends to happen together. Do not claim that the explanatory variable caused the response unless the study design supports a causal conclusion.
- Writing a checklist instead of a connected description. A string of terms such as “negative, linear, moderate, one outlier” is less clear than sentences connecting the features to the variables and the setting.
For a final audit, point to the phrase in your answer that describes each feature. Then ask whether the values or visual pattern support that phrase. If you cannot locate evidence for a claim—or if two claims conflict—revise before moving on. “Writing Descriptions in Context” provides the guiding principle: name the individuals, variables, and units, and describe what the graph shows without going beyond it.
Check Your Understanding
Use the evidence-first routine to review each description. Explain what the available values or plot features support.
- In the stream data, which observation is unusual, and how does it affect the description of strength?
- Why is “strong positive linear association” inaccurate for the charging data?
- In the training data, what do the program clusters show that a single pooled direction statement would leave out?
- What two coordinates locate the unusual Program B observation?
- Rewrite this statement more carefully: “The training program causes fewer sorting errors.”