Not sure which lesson to do first? Find your level and get a plan →

Statistics · Grade 6-8

Scatter Plot

A scatter plot graphs each individual as a single dot at the coordinates of its two numerical measurements, so the cloud of dots reveals whether the two variables move together.


Quick Answer

A scatter plot graphs each individual as a single dot at the coordinates of its two numerical measurements, so the cloud of dots reveals whether the two variables move together. Make one when every individual has two measurements, like each day's temperature and the number of ice cream cones sold that day, and you want to see the relationship before calculating anything. The tilt of the cloud shows direction, its tightness shows strength, and dots far from the rest are outliers worth checking. The explanatory variable goes on the horizontal axis and the response on the vertical axis.

Why This Matters

A scatter plot is where any study of two measurements should start, because it shows what a single number hides: a curve instead of a line, two separate clusters, or one outlier dragging the trend. Skip it, and you can fit a straight line to a curved pattern, or report a correlation that one strange point created.

Intuitive Explanation

An ice cream stand records two numbers each day: the afternoon temperature and how many cones it sold. On Monday it was 24°C and 61 cones, so Monday becomes one dot, 24 along the bottom and 61 up the side. After six days there are six dots, and reading them from left to right, they climb: hotter days sold more cones. A table of the same twelve numbers does not show that nearly as fast.

Read four things from any scatter plot. Direction: do the dots rise or fall from left to right? Form: do they follow a straight line, a curve, or separate clusters? Strength: do they hug the pattern tightly or spread loosely around it? Outliers: is any dot far from the rest? A student who studied 2 hours and scored 98, when the others near 2 hours scored about 65, deserves a second look, not a quiet delete.

A scatter plot needs two numbers for each individual. If you record each student's favorite color and test score, color is a category, and putting red at 1, blue at 2 and green at 3 on an axis is arbitrary: reorder the colors and the apparent trend changes. For a category against a number, compare the groups side by side with dot plots or box plots; for counts of categories alone, use a bar graph.

Think of it like this. Picture pinning each student to a corkboard: across by hours of sleep, up by quiz score. Step back, and the shape the pins make, a rising band, a falling band or a scattered mess, shows the relationship at a glance. The pins show where students landed, not why they landed there.

Core idea. Plot each individual once, at the point set by its two measurements; the shape of the cloud, meaning its direction, form, strength and outliers, is the relationship between the two variables, visible before any calculation.


When to Use

Make a scatter plot when the data come in pairs of numbers for the same individuals, such as height and arm span or hours practiced and score, and you want to see whether and how they are related. It comes before computing a correlation or drawing a line of best fit. If one variable is a category, use side-by-side dot plots or box plots; if you only have counts of categories, use a bar graph; if one variable is time, a line graph usually reads better.

Tip. Ask: Does each individual give two numbers, to be plotted as one unconnected dot so the shape of the cloud shows the relationship?

How to Recognize It

Before drawing a scatter plot, check how many numbers were recorded for each individual and whether the data follow one thing over time.

  1. Was each person, day or object measured on two numerical variables?

    Yes means each one becomes a dot at (x,y). If only one number was recorded for each, such as test scores alone, a Dot Plot or histogram shows how those values spread.

  2. Are both measurements numbers rather than categories like favorite color?

    Yes lets both go on number-line axes. If one is a category, its values have no order along an axis, so compare the groups with side-by-side dot plots; if both are categories, use Two-Way Tables.

  3. Do the data come from many different individuals rather than one thing measured again and again over time?

    Yes fits unconnected dots. If you are tracking one plant's height every week, a Line Graph joins the points in time order to show the change.

  4. Do you only need to show and describe the cloud, not predict a new value?

    Yes is the scatter plot's whole task. If you must estimate the score for 4.5 hours of study, draw a Line of Best Fit through the cloud and read the estimate from it.

Scatter Plot vs Line Graph vs Dot Plot vs Correlation

All four can be drawn with dots on a grid, so they get confused. Ask how many numbers each dot carries and whether the points follow one thing through time.

Scatter Plot

Meaning
A scatter plot shows each individual as one unconnected dot placed by two numerical measurements.
Key test
Two numbers per dot; the order means nothing, so dots are never joined.
Example
Each student is one dot at (hours of sleep, quiz score), such as (8, 9).

Line Graph

Meaning
A line graph shows how one quantity changes over time, with the points joined in time order.
Key test
Time runs along the bottom, and neighboring points are connected.
Example
A puppy's weight each month for a year, the points joined into one rising line.

Dot Plot

Meaning
A dot plot stacks one dot per data value above a number line to show how one variable is spread.
Key test
One number per dot; dots pile up in columns over the values.
Example
Quiz scores for 20 students, with four dots stacked above the score 7.

Correlation

Meaning
Correlation is the description of the trend a scatter plot shows: positive, negative or none, and how tight.
Key test
It is a statement about the dots, not a graph.
Example
The sleep and quiz-score dots climb from left to right, a positive correlation.

Formula & Notation

How to read it: The horizontal axis (x) shows the independent (explanatory) variable; the vertical axis (y) shows the dependent (response) variable. Each dot represents one observation at coordinates (xi,yi).

Worked Examples

Example 1 — Temperature and cones sold

Easy

An ice cream stand records six days as (temperature in °C, cones sold): (18, 40), (21, 52), (24, 61), (27, 75), (30, 82), (33, 95). Set up a scatter plot and describe the relationship.
Show the steps and the answer
  1. Put temperature on the horizontal axis and cones on the vertical axis.

    Temperature is the explanatory variable: the weather helps explain sales, not the other way around.

  2. Choose scales that fit the data: temperature from 15 to 35 in steps of 5; cones from 0 to 100 in steps of 20.

    Each axis needs equal steps and a range that covers every value, or the shape of the cloud gets distorted.

  3. Plot each day as one dot, for example (24, 61): across to 24, up to 61.

    Each dot is one day measured twice, so six days give exactly six dots.

  4. Describe the cloud: cones rise by 12, 9, 14, 7 and 13 for each 3-degree step, about 95−4033−18=5515≈3.7 cones per degree.

    Steady increases mean a positive, roughly straight, fairly strong pattern, and no dot sits far from the others.

Answer: A strong positive, roughly linear relationship with no outliers: about 3.7 more cones sold for each extra degree.

Takeaway: After plotting, describe direction, form, strength and outliers, in that order.

Example 2 — A category is not an axis

Standard

Leo records favorite color and test score for 12 students. Red: 70, 80, 75, 85. Blue: 60, 90, 80, 70. Green: 65, 85, 78, 72. He wants a scatter plot with red at 1, blue at 2 and green at 3 on the x-axis. Is a scatter plot the right display? What do the data show?
Show the steps and the answer
  1. Check the variable types: color is a category; score is a number.

    A scatter plot needs two numerical measurements, because positions on both axes must mean amounts.

  2. Test Leo's axis: ordering the colors green, red, blue would change any uphill or downhill look.

    Numbers like 1, 2, 3 for colors are labels, so any slope in the plot would come from the order Leo chose.

  3. Compare the groups instead: red 3104=77.5, blue 3004=75, green 3004=75.

    For a number across categories, side-by-side dot plots and group means are the fair comparison.

Answer: No; color is categorical, so use side-by-side dot plots. The group means are 77.5, 75 and 75, with heavy overlap, so color shows no clear link to score.

Takeaway: Use a scatter plot only when both axes measure amounts; a category against a number calls for a group comparison.

Example 3 — Find the mistake: deleting an outlier

Application

Zoe plots hours studied against quiz score for six students: (1, 60), (2, 65), (3, 72), (4, 78), (5, 84), (2, 98). She writes: "The dot at (2, 98) must be a mistake, so I deleted it. Now the plot shows a perfect pattern." The teacher's records confirm the 98 is real. Find the mistake and correct it.
Show the steps and the answer
  1. Describe the other five dots: scores rise from 60 to 84 over 4 hours, about 84−604=6 points per hour.

    The main pattern is clear and positive, so the extra dot does not need to be removed for the trend to show.

  2. Measure how unusual (2, 98) is: the trend predicts about 60+6=66 at 2 hours, so 98 is about 32 points above it.

    A point that far from the pattern is an outlier, and it carries information, for example a student who already knew the material.

  3. Keep the dot and report it separately from the trend.

    A real value cannot be deleted just because it spoils the pattern; doing so makes study time look like it explains more than it does.

Answer: Keep the point. The plot shows a positive trend of about 6 points per hour studied, plus one real outlier at (2, 98) that the trend does not explain.

Takeaway: Investigate outliers and report them; only remove a point when it is shown to be an error.

Where You'll See It

  • Height vs. arm span in class

    Each classmate is one dot positioned by height (x) and arm span (y); the upward cloud of dots shows that taller kids tend to have longer arm spans.

  • Hours practiced vs. game score

    Plotting practice hours against video-game high scores puts each player as a dot, and an upward trend in the dots suggests more practice goes with higher scores.

  • Temperature vs. ice cream cones sold

    Each day is a dot placed by that day's temperature and cones sold; the rising pattern of dots reveals the relationship between heat and sales.

Common Mistakes

  1. The slip: It doesn't matter which variable goes on the x-axis.

    Why it's tempting: Both are just numbers, so the axes feel interchangeable.

    Instead: Put the explanatory (independent) variable on x and the response (dependent) variable on y. Swapping them mislabels which variable is being used to predict the other.

  2. The slip: An upward trend in the dots proves one variable causes the other.

    Why it's tempting: A clear visual slope looks like cause and effect.

    Instead: A scatter plot shows association only. A lurking variable can produce the trend; causation needs a designed experiment.

  3. The slip: A stray dot far from the pattern is just a mistake to ignore.

    Why it's tempting: Outliers look like errors that should be deleted for a cleaner picture.

    Instead: Outliers can be real and can strongly pull the line of best fit and correlation. Investigate them rather than silently dropping them.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. Five people's (hours of sleep, reaction time in ms) are (5, 350), (6, 320), (7, 290), (8, 270), (9, 250). Describe the direction and form of the scatter plot.

    Hint: As sleep increases, what happens to reaction time, and do the points follow a line or a curve?

  2. Eight students' (study hours, test score) are (1, 55), (2, 62), (3, 68), (4, 70), (5, 79), (6, 84), (2, 95), (7, 90). Which point is an outlier, and why?

    Hint: Compare each score with the scores of students who studied about as long.

  3. You study how fertilizer amount (g) affects plant height (cm). Which variable goes on each axis? Where do you plot a plant given 15 g that grew 42 cm?

    Hint: The explanatory variable goes on the horizontal axis.

  4. A scatter plot shows children's (age in years, height in cm): (4, 102), (6, 115), (8, 128), (10, 138), (12, 149). Use the trend to estimate the height of a 9-year-old.

    Hint: Age 9 lies halfway between two ages in the data.

  5. For 25 students you know each one's favourite sport and number of siblings. Can you show these two variables in a scatter plot?

    Hint: A scatter plot needs two numbers for each person.

  6. A scatter plot of 20 towns shows that towns with more libraries also have more crimes. Tia says, "Building libraries causes crime." Correct her.

    Hint: What could make a town have both more libraries and more crime?

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

Should I connect the dots on a scatter plot?

No. Each dot is a separate individual, and the order you list them in means nothing, so joining them draws a zigzag with no meaning. Connecting points makes sense only when they track one thing over time, which is a line graph. To show the trend, draw one straight line through the middle of the cloud instead of joining dot to dot.

What if two students have exactly the same pair of numbers?

Their dots land on the same spot, and the plot hides one of them, so it looks as if there are fewer data than there really are. You can draw a slightly bigger dot, write a small 2 beside it, or nudge one dot a tiny bit sideways, a trick statisticians call jittering. With rounded data like whole hours of sleep, overlaps are common, so watch for them.

Do the axes on a scatter plot have to start at zero?

No. Unlike a bar graph, where bar length must start from zero, a scatter plot can start each axis near the smallest value so the cloud fills the space. If heights run from 140 to 180 cm, an axis starting at 0 squashes every dot into one corner. Mark a small zigzag break on the axis to show you skipped ahead, and use the same scales when you compare two plots.

What does it mean when the dots form two separate clumps?

Clumps often mean the data mix two different kinds of individuals. Plot height against weight for a school that includes first graders and eighth graders, and you may see one clump low on the left and another high on the right. Each clump on its own can tell a different story from the combined picture, so ask what makes the groups different before describing one overall trend.

Learning Path

← Before

No prerequisites
Scatter Plot

You are here

No earlier statistics concept is required, only plotting points (x,y) on a coordinate grid. The scatter plot adds the idea that each point is one individual measured twice, so the shape of the cloud is the relationship. Next, Correlation describes that shape's direction and strength, and Line of Best Fit draws a trend line through the cloud for making predictions.

See Also