Scatter Plots & Association
Learn to create scatter plots from two-variable data, identify outliers and clusters, and recognize positive, negative, and no association patterns.
What you'll do in this lesson
A voice-first session with the Crimsora tutor on Scatter Plots & Association, then targeted practice and FRQs — with the tutor adapting to where you get stuck.
What this lesson covers
When you collect information about two different measurements—like hours studied and test scores, or temperature and ice cream sales—you get what's called bivariate data. A scatter plot is the most powerful tool for seeing whether those two things actually move together, and if so, how. In this lesson, you'll learn to build scatter plots, spot the patterns that matter, and describe what those patterns tell you about the relationship between the variables.
What Is a Scatter Plot?
A scatter plot is a graph that shows pairs of values from bivariate data. One variable goes on the horizontal axis (the -axis) and the other on the vertical axis (the -axis). Each point represents one observation. For example, if you're looking at the relationship between hours of sleep and test performance, each dot shows one student's hours of sleep paired with their test score.
Scatter plots let you see the overall pattern at a glance. Instead of staring at a table of numbers, you can instantly spot whether the points climb together, fall together, or scatter randomly. This visual summary is why scatter plots are so useful in science, business, and everyday decision-making—they make relationships visible.
When you construct a scatter plot, choose a reasonable scale for each axis so the points spread across the graph and use the full space available. Crowding all the points in one corner makes patterns hard to see. Label both axes clearly with the variable names and units.
Scatter plots let you see the overall pattern at a glance. Instead of staring at a table of numbers, you can instantly spot whether the points climb together, fall together, or scatter randomly. This visual summary is why scatter plots are so useful in science, business, and everyday decision-making—they make relationships visible.
When you construct a scatter plot, choose a reasonable scale for each axis so the points spread across the graph and use the full space available. Crowding all the points in one corner makes patterns hard to see. Label both axes clearly with the variable names and units.
Recognizing Positive and Negative Association
If two variables have a positive association, as one variable increases, the other tends to increase too. Points cluster around an upward-sloping trend from lower left to upper right. Examples: height and weight, practice time and skill level, hours worked and money earned.
If two variables have a negative association, as one variable increases, the other tends to decrease. Points cluster around a downward-sloping trend from upper left to lower right. Examples: price and quantity bought, temperature and heating costs, exercise frequency and resting heart rate.
If there is no association (also called no correlation), knowing one variable tells you nothing predictable about the other. Points scatter randomly with no clear trend. Examples: shoe size and favorite color, birth month and math ability, height and number of siblings.
Importantly, association describes only the pattern you see in the graph. It does not claim that one variable causes the other—that requires deeper investigation. A strong association can exist purely by coincidence or because both variables are influenced by some third factor.
If two variables have a negative association, as one variable increases, the other tends to decrease. Points cluster around a downward-sloping trend from upper left to lower right. Examples: price and quantity bought, temperature and heating costs, exercise frequency and resting heart rate.
If there is no association (also called no correlation), knowing one variable tells you nothing predictable about the other. Points scatter randomly with no clear trend. Examples: shoe size and favorite color, birth month and math ability, height and number of siblings.
Importantly, association describes only the pattern you see in the graph. It does not claim that one variable causes the other—that requires deeper investigation. A strong association can exist purely by coincidence or because both variables are influenced by some third factor.
Clustering, Outliers, and Spread
Clustering occurs when many points bunch together in one region of the plot. This shows consistency: many observations share similar pairs of values. For instance, if you plot the relationship between class size and average class grade across many schools, you might see a cluster around 25 students and a grade of 82, meaning that class size and performance are relatively consistent.
An outlier is a point that lies far away from the main cluster or pattern. Outliers often reveal unusual or noteworthy cases. If a student studied only 2 hours but scored 95 on a test, that point would sit far above the trend line and stand out as exceptional. Outliers deserve investigation: they might be data entry errors, unique circumstances, or genuinely interesting cases that break the normal pattern.
The overall spread of points around a trend—how tightly they pack or how widely they scatter—tells you about the strength of association. A tight cluster shows a strong association; scattered points show a weak association. A student who plots bivariate data should always describe both the direction (positive, negative, or none) and the strength (tight cluster or loose scatter).
An outlier is a point that lies far away from the main cluster or pattern. Outliers often reveal unusual or noteworthy cases. If a student studied only 2 hours but scored 95 on a test, that point would sit far above the trend line and stand out as exceptional. Outliers deserve investigation: they might be data entry errors, unique circumstances, or genuinely interesting cases that break the normal pattern.
The overall spread of points around a trend—how tightly they pack or how widely they scatter—tells you about the strength of association. A tight cluster shows a strong association; scattered points show a weak association. A student who plots bivariate data should always describe both the direction (positive, negative, or none) and the strength (tight cluster or loose scatter).
Linear vs. Nonlinear Patterns
A linear pattern means the points roughly follow a straight line. The rate of change stays roughly constant: as increases by 1, tends to increase (or decrease) by about the same amount each time. Examples: distance traveled at constant speed over time, total cost when buying items at a fixed price per item.
A nonlinear pattern means the relationship curves or changes speed. The points might follow a U-shape, an upside-down U, an exponential curve, or some other bent path. Examples: distance fallen by a dropped object over time (speeds up due to gravity), profit at different production volumes (might increase then decrease after a peak due to overhead costs), height of a bouncing ball over time (zigzag pattern).
When describing a scatter plot, always note whether any trend is linear or nonlinear. If it's linear, you can later use a straight line to model the data. If it's nonlinear, a straight line won't work well, and you'd need a different approach. Recognizing this distinction early saves time and prevents fitting the wrong model.
A nonlinear pattern means the relationship curves or changes speed. The points might follow a U-shape, an upside-down U, an exponential curve, or some other bent path. Examples: distance fallen by a dropped object over time (speeds up due to gravity), profit at different production volumes (might increase then decrease after a peak due to overhead costs), height of a bouncing ball over time (zigzag pattern).
When describing a scatter plot, always note whether any trend is linear or nonlinear. If it's linear, you can later use a straight line to model the data. If it's nonlinear, a straight line won't work well, and you'd need a different approach. Recognizing this distinction early saves time and prevents fitting the wrong model.
Describing a Scatter Plot Completely
A complete description of a scatter plot addresses four elements: the variables (what is measured on each axis), the association (positive, negative, or none), the strength (tight clustering or loose scatter, and whether association is strong or weak), and any noteworthy features (clusters, outliers, or nonlinear patterns).
For example: "The scatter plot shows the relationship between hours of after-school tutoring (horizontal) and quiz scores (vertical) for 20 seventh-grade students. There is a strong positive linear association: as tutoring hours increase, quiz scores tend to increase roughly proportionally. Most points cluster between 1 and 4 hours of tutoring with quiz scores from 75 to 95. One student received 5 hours of tutoring but scored only 70, which is an outlier below the main trend—possibly indicating other factors affected that student's performance."
This type of description is precise, evidence-based, and ready to guide further analysis or decision-making.
For example: "The scatter plot shows the relationship between hours of after-school tutoring (horizontal) and quiz scores (vertical) for 20 seventh-grade students. There is a strong positive linear association: as tutoring hours increase, quiz scores tend to increase roughly proportionally. Most points cluster between 1 and 4 hours of tutoring with quiz scores from 75 to 95. One student received 5 hours of tutoring but scored only 70, which is an outlier below the main trend—possibly indicating other factors affected that student's performance."
This type of description is precise, evidence-based, and ready to guide further analysis or decision-making.
Key terms
- Bivariate data.
- Paired measurements of two different variables for the same objects or individuals, such as height and arm span for each student.
- Scatter plot.
- A graph where each point represents one pair of values, with one variable on the horizontal axis and the other on the vertical axis.
- Positive association.
- A relationship where one variable tends to increase as the other increases, shown by points trending upward from lower left to upper right.
- Negative association.
- A relationship where one variable tends to decrease as the other increases, shown by points trending downward from upper left to lower right.
- No association.
- No consistent relationship between two variables; points scatter randomly with no clear trend.
- Outlier.
- A point that lies far away from the main cluster or pattern and may indicate an unusual case or error.
- Clustering.
- The concentration of multiple data points in one region of a scatter plot, showing consistency among those observations.
- Linear pattern.
- A relationship where points roughly follow a straight line, indicating a constant rate of change.
Worked example
A science class collected data on the number of minutes 12 students spent studying for a biology test and the score each student earned. Create a scatter plot from the data and describe the pattern.
Data: (15, 72), (20, 78), (25, 81), (30, 85), (10, 65), (35, 88), (40, 92), (18, 75), (28, 84), (22, 79), (38, 90), (5, 58)
Data: (15, 72), (20, 78), (25, 81), (30, 85), (10, 65), (35, 88), (40, 92), (18, 75), (28, 84), (22, 79), (38, 90), (5, 58)
Step 1: Set up the axes. The independent variable (time spent studying) goes on the horizontal axis, and the dependent variable (test score) goes on the vertical axis. Label them clearly: "Minutes Studying" and "Test Score." Choose scales that fit all the data comfortably—minutes from 0 to 45, scores from 50 to 95.
Step 2: Plot each point. For (15, 72), find 15 on the horizontal axis and 72 on the vertical axis, then mark the intersection. Repeat for all 12 pairs: (20, 78), (25, 81), (30, 85), (10, 65), (35, 88), (40, 92), (18, 75), (28, 84), (22, 79), (38, 90), (5, 58).
Step 3: Look for patterns. Notice that as minutes increase, test scores generally increase too. The points form a cloud that trends upward from lower left to upper right—no point sits far away from this trend, so there are no obvious outliers. The points are fairly tightly clustered around an imaginary upward line.
Step 4: Describe what you see. "The scatter plot shows a strong positive linear association between study time and test score. As study minutes increase from 5 to 40, test scores increase from 58 to 92. The points cluster consistently along an upward trend with no dramatic outliers. This pattern suggests that studying longer is associated with higher test performance." This description identifies the variables, the type and direction of association, the pattern (linear), the strength (strong, tight cluster), and the practical meaning.
Step 2: Plot each point. For (15, 72), find 15 on the horizontal axis and 72 on the vertical axis, then mark the intersection. Repeat for all 12 pairs: (20, 78), (25, 81), (30, 85), (10, 65), (35, 88), (40, 92), (18, 75), (28, 84), (22, 79), (38, 90), (5, 58).
Step 3: Look for patterns. Notice that as minutes increase, test scores generally increase too. The points form a cloud that trends upward from lower left to upper right—no point sits far away from this trend, so there are no obvious outliers. The points are fairly tightly clustered around an imaginary upward line.
Step 4: Describe what you see. "The scatter plot shows a strong positive linear association between study time and test score. As study minutes increase from 5 to 40, test scores increase from 58 to 92. The points cluster consistently along an upward trend with no dramatic outliers. This pattern suggests that studying longer is associated with higher test performance." This description identifies the variables, the type and direction of association, the pattern (linear), the strength (strong, tight cluster), and the practical meaning.
Practice questions
A scatter plot shows the relationship between temperature (horizontal axis, in degrees Celsius) and the number of people at a city swimming pool (vertical axis). The points trend upward from lower left to upper right, clustering tightly around an imaginary line, with no points far from the trend. Which statement correctly describes the association?
- There is a strong negative linear association between temperature and pool attendance.
- There is a strong positive linear association between temperature and pool attendance.
- There is no association between temperature and pool attendance.
- There is a nonlinear relationship between temperature and pool attendance.
Answer: There is a strong positive linear association between temperature and pool attendance.
The description "upward from lower left to upper right" means as one variable increases, the other increases—that is positive association. The "tight clustering around a line" means strong association and a linear pattern. Negative association would trend downward; no association would scatter randomly; nonlinear patterns follow curves or bent paths, not straight lines.
A student plots age (in years) versus hours of sleep per night for 25 people and observes one point at (85, 10) while the rest cluster between ages 20–40 with 7–9 hours of sleep. What is this isolated point called, and why might the student investigate it further?
Answer: The point (85, 10) is an outlier. The student should investigate it because the person is much older than the others and sleeps much less than the main cluster, suggesting either a unique circumstance (such as a sleep disorder or health condition in elderly people, which differs from the younger group), a data entry error, or a genuinely noteworthy case.
An outlier lies far from the main cluster or trend. This point is isolated both horizontally (age 85 vs. 20–40) and vertically (10 hours vs. 7–9). Outliers deserve investigation because they might reveal errors, special conditions, or real phenomena that differ from the typical pattern. Simply ignoring them misses important information.
On a scatter plot of age versus height for students in grades 1 through 8, the points form a cloud that curves upward steeply for younger ages, then flattens out at older ages. How would you describe the shape of this relationship, and what does it tell you about the rate of change?
Answer: The relationship is nonlinear because the points follow a curve rather than a straight line. The steep part for younger grades shows rapid height increase; the flat part for older grades shows that height change slows down. This tells you that the rate of change is not constant—students grow quickly when young and much more slowly as they get older.
A linear pattern shows constant rate of change (the same increase in for each unit increase in ), which appears as a straight line. A curve indicates the rate changes—in this case, growth rate decreases over time. Recognizing nonlinearity matters because a straight-line model would fit poorly, especially for predicting height at different ages.
FAQ
- Do I need to connect the dots on a scatter plot?
- No. A scatter plot shows individual data points as isolated dots. Do not draw lines connecting the dots in order. If a trend exists, points cluster around an imaginary trend, but connecting them point-to-point creates a meaningless zigzag. Connecting dots is used in different graphs, like line graphs that show how one quantity changes over time.
- If there is a strong association, does that mean one variable causes the other?
- Not necessarily. Association only describes the pattern you see in the graph—that two variables move together. Causation means one variable directly influences the other. You might see a strong association between ice cream sales and drowning deaths (both increase in summer), but neither causes the other; both increase because of warm weather. Determining causation requires deeper investigation beyond the scatter plot.
- How do I know if I should describe an association as strong or weak?
- Look at how tightly the points cluster around the trend. If most points lie very close to an imaginary line or curve running through them, the association is strong. If points scatter widely and loosely, the association is weak. A strong association means the two variables are closely related; a weak one means other factors also play a big role, so the relationship is less reliable for prediction.
- What should I do if my scatter plot looks like a random cloud with no pattern at all?
- That means there is no association between the two variables. Write that explicitly: 'There is no apparent association between X and Y.' Knowing that two variables are unrelated is valuable information—it tells you that one variable does not help predict the other, and you should look elsewhere to explain changes in your variable of interest.
Learn this with a teacher, not a page
The Crimsora tutor teaches Scatter Plots & Association live — explaining on a whiteboard, asking you questions, and adapting to where you get stuck.