Best examples of scatter plot examples with correlation coefficient in real data

Real-world examples of scatter plot examples with correlation coefficient

Let’s start where most people learn this concept: plotting two variables and asking, “Do they move together?” These real examples of scatter plots with correlation coefficients show how that plays out across different fields.

Example 1: Height vs. Weight (Moderate Positive Correlation)

Imagine a data set of 200 adults from a health survey. On the x-axis you plot height (in inches), and on the y-axis weight (in pounds). Each point is one person.

What you see: taller people tend to weigh more, but not in a perfectly straight line. Some shorter people weigh more than taller people, and there’s plenty of spread.

If you compute Pearson’s correlation coefficient, you might get something like r ≈ 0.65.

How to read that:

• The scatter plot shows an upward trend: as height increases, average weight increases.
• The correlation coefficient of 0.65 says this is a moderate to strong positive relationship, but far from perfect.
• You can visually confirm it: points cluster around an upward-sloping line, but they’re not locked onto it.

This is one of the best examples of scatter plot examples with correlation coefficient because it’s intuitive. Most people expect the relationship to be positive, but they also understand there are many other factors: muscle mass, diet, age, and health conditions.

For real health data sources similar to this, you can explore the CDC’s National Health and Nutrition Examination Survey (NHANES) data sets at cdc.gov.

Example 2: Study Time vs. Exam Score (Strong Positive Correlation)

Now picture a college class of 100 students. On the x-axis: hours spent studying for the final exam. On the y-axis: exam score (0–100).

What you might see in the scatter plot:

• Students who studied fewer than 2 hours cluster below 70.
• Students who studied 5–10 hours cluster between 75 and 95.
• A few outliers: someone who barely studied and still scored high, and someone who studied a lot but still scored low.

A typical correlation coefficient here might be r ≈ 0.80.

How to interpret this example of scatter plot examples with correlation coefficient:

• The scatter plot has a tight upward trend; more study time usually means a higher score.
• The correlation coefficient near 0.8 indicates a strong positive correlation.
• The outliers remind you that correlation is not destiny: test anxiety, prior knowledge, and question difficulty all matter.

For education research that uses similar analyses, you can look at materials from Harvard’s Graduate School of Education at gse.harvard.edu.

Example 3: Advertising Spend vs. Sales Revenue (Real Business Data)

Switching to business, consider a retail company tracking monthly data over two years. On the x-axis: monthly advertising spend (in dollars). On the y-axis: monthly sales revenue.

What the scatter plot shows:

• As ad spend rises from \(5,000 to \)50,000, revenue generally rises.
• The relationship is positive, but there’s noise: seasonal patterns, promotions, and economic conditions.

You might calculate r ≈ 0.55.

How to read this in context:

• The scatter plot suggests a moderate positive correlation between ad spend and revenue.
• A correlation coefficient around 0.55 tells you there is a relationship, but it’s not the only driver of sales.
• In 2024–2025, marketers often combine this kind of scatter plot with digital analytics (for example, click-through rates or conversion rates) to refine campaigns.

This is one of the best examples of scatter plot examples with correlation coefficient in business: it shows why companies can’t just throw money at ads and expect perfectly predictable returns.

Example 4: Hours of Exercise vs. Resting Heart Rate (Negative Correlation)

Now let’s flip the direction. Suppose you have data from a wellness program with 150 participants. On the x-axis: average hours of moderate exercise per week. On the y-axis: resting heart rate (beats per minute).

What the scatter plot looks like:

• People who exercise 0–1 hours per week cluster around 75–85 bpm.
• People who exercise 4–6 hours per week cluster around 60–70 bpm.
• There’s a downward trend, but not perfect.

The correlation coefficient might be r ≈ −0.60.

Interpreting this real example:

• The scatter plot shows a negative relationship: more exercise is associated with a lower resting heart rate.
• The correlation coefficient of −0.60 indicates a moderate to strong negative correlation.
• This does not prove that exercise alone caused the lower heart rate; genetics, age, and health status also matter.

If you want to see real cardiology data and guidelines that often rely on this kind of analysis, check resources from the National Institutes of Health at nih.gov and the Mayo Clinic at mayoclinic.org.

Example 5: Age vs. Reaction Time (Positive Correlation in Aging Research)

In aging research, one common topic is how reaction time changes with age. Imagine a study with 300 participants aged 20–80. On the x-axis: age. On the y-axis: reaction time in milliseconds.

What the scatter plot usually reveals:

• Younger adults (20–30) cluster at lower reaction times (faster responses).
• Older adults (60–80) cluster at higher reaction times (slower responses).
• There’s plenty of overlap, but a visible upward trend.

The correlation coefficient might be r ≈ 0.45.

Why this is a useful example of scatter plot examples with correlation coefficient:

• The scatter plot shows a positive trend: as age increases, reaction time tends to increase.
• An r of 0.45 is a moderate positive correlation, not a perfect predictor.
• This helps researchers frame realistic expectations: aging tends to slow reaction time, but individual differences are large.

Aging and cognitive research using similar methods can be found through the National Institute on Aging at nia.nih.gov.

Example 6: Screen Time vs. Sleep Duration in Teens (Weak to Moderate Negative Correlation)

Public health researchers have been looking hard at screen time and sleep, especially among teenagers, through 2024–2025. Imagine a survey of 500 high school students. On the x-axis: average daily recreational screen time (hours). On the y-axis: average nightly sleep duration (hours).

The scatter plot might show:

• Teens with 1–2 hours of screen time often sleep 8–9 hours.
• Teens with 5–7 hours of screen time more often sleep 5–7 hours.
• There’s lots of scatter, but a visible downward slope.

A typical correlation coefficient could be r ≈ −0.35.

How to interpret this real example:

• The scatter plot suggests a weak to moderate negative correlation: more screen time is associated with less sleep.
• The correlation coefficient is not very large, which fits the reality that many other factors affect sleep (stress, homework, family routines, and so on).
• Public health recommendations use this kind of pattern to inform guidelines, not to claim a perfect one-to-one effect.

For related data and reports, check the CDC’s adolescent health resources at cdc.gov.

Example 7: Temperature vs. Ice Cream Sales (Classic Strong Positive Correlation)

Here’s a classic textbook-style business example that still shows up in modern analytics. Imagine a food truck tracking daily data over a summer. On the x-axis: average daily temperature (°F). On the y-axis: number of ice cream cones sold.

The scatter plot pattern:

• On cool days (60–70°F), sales are modest.
• As temperature rises to 80–95°F, sales rise sharply.
• The points line up closely along an upward-sloping curve or line.

A correlation coefficient might be r ≈ 0.85.

Why this is one of the best examples of scatter plot examples with correlation coefficient:

• The strong positive correlation is visually obvious.
• It’s a clean illustration that correlation does not imply causation in the simple sense: temperature doesn’t “force” people to buy ice cream, but it strongly influences behavior.
• It also hints at nonlinearity; beyond a certain temperature, people may avoid going outside, which could flatten or bend the relationship.

Example 8: No Relationship – Shoe Size vs. Math Score (Near Zero Correlation)

To round things out, consider a nonsense pair of variables: shoe size and math test scores among middle school students.

In the scatter plot:

• Points are scattered randomly, with no clear upward or downward trend.
• Larger shoe sizes don’t consistently match higher or lower scores.

The correlation coefficient here is often r ≈ 0.00 or very close to zero.

This is a helpful example of scatter plot examples with correlation coefficient because it shows what no linear relationship looks like:

• The scatter plot is a cloud of points.
• The correlation coefficient near zero confirms that there is no consistent linear pattern.
• It’s a reminder that not all pairs of variables are meaningfully related, even if you can always calculate a correlation.

How to read scatter plot examples with correlation coefficient like a data pro

When you look at these examples of scatter plot examples with correlation coefficient, there are a few habits that separate careful analysts from people who just chase big numbers.

Look at the plot before the number

The correlation coefficient is tempting because it’s a single, tidy number. But every statistician will tell you: always check the scatter plot first.

Here’s why:

• A correlation of 0.8 could come from a nice straight-line pattern or from a weird curved pattern that Pearson’s r doesn’t fully capture.
• Outliers can dramatically inflate or deflate the correlation coefficient.
• Two data sets can have the same correlation but very different shapes in their scatter plots.

The classic illustration of this idea is Anscombe’s quartet, a set of four data sets with identical correlation coefficients but very different scatter plots. You can read more about it in many statistics courses hosted on university sites, including materials from UCLA and other .edu statistics resources.

Match the correlation coefficient to the story

In the best examples of scatter plot examples with correlation coefficient, the number supports a story you can defend:

• A moderate correlation (around 0.4–0.6) often shows up in social science and health data where many factors are at play.
• Very high correlations (above 0.9) in messy real-world data can be a red flag for data issues or variables that are basically measuring the same thing twice.
• Very low correlations do not mean “no relationship ever” — they mean no strong linear relationship in that data set.

When you interpret real examples, always ask:

• Does the direction (positive or negative) match what you’d expect from domain knowledge?
• Is the strength plausible given all the other factors that could influence the outcome?
• Are outliers driving the correlation?

Remember: correlation is about association, not cause

Every one of the examples of scatter plot examples with correlation coefficient above can tempt you into saying things like “screen time causes poor sleep” or “exercise causes lower heart rate.” The honest statement is softer:

In this data set, higher screen time is associated with shorter sleep duration.

Causation usually requires:

• Experimental manipulation (like randomized controlled trials), or
• Very careful observational designs and controls.

Public health agencies, including the CDC and NIH, repeatedly emphasize this distinction in their reports. When you see real examples in those documents, you’ll notice careful language around association vs. cause.

FAQ: common questions about scatter plots and correlation

Q: Can you give a simple example of a scatter plot with correlation coefficient that beginners understand?
A: One of the easiest examples of scatter plot examples with correlation coefficient for beginners is study time vs. exam score. You plot each student’s study hours and score, then calculate the correlation coefficient. The upward trend in the scatter plot and a correlation around 0.7–0.8 line up with common sense: more studying tends to be linked with higher scores, but not perfectly.

Q: What are some real examples of negative correlation in scatter plots?
A: Real examples include exercise hours vs. resting heart rate, screen time vs. sleep duration, and price vs. quantity demanded in basic economics. In each scatter plot, as the x-variable increases, the y-variable tends to decrease, and the correlation coefficient is negative.

Q: Is a correlation coefficient of 0.3 worth paying attention to?
A: Yes, especially in fields like psychology, education, or public health. A correlation of 0.3 can represent a meaningful, though modest, linear association. The scatter plot will show a loose trend rather than a tight line. In large data sets, even r = 0.3 can be statistically significant and practically important.

Q: How many data points do I need for a reliable example of scatter plot with correlation coefficient?
A: Technically you can compute a correlation with as few as two or three points, but that’s not useful. In practice, analysts often want at least a few dozen data points before trusting the correlation, and hundreds or thousands when making serious policy or business decisions. The scatter plot helps you judge whether the pattern looks stable or dominated by a few outliers.

Q: Can two variables have a strong relationship but a low correlation coefficient?
A: Yes. If the relationship is nonlinear (for example, a U-shape or a curve), Pearson’s correlation can be near zero even when the scatter plot shows a clear pattern. That’s another reason real examples of scatter plot examples with correlation coefficient always start with a visual check. If the pattern is curved, you may need a different model or a transformed variable.

When you look back at these examples of scatter plot examples with correlation coefficient — from health and education to business and everyday behavior — the pattern is consistent: the scatter plot shows you the shape and quirks of the relationship, and the correlation coefficient gives you a compact summary. You need both to tell an honest story about your data.