What Is an XY Graph? Axes, Plotting and Examples

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-10-03

An x y graph is a chart that plots pairs of numbers as points on a grid, one value on the horizontal axis and one on the vertical axis. Each point shows a single observation, and the overall pattern of points shows how the two variables relate. This article explains the axes, the plotting steps, and how to read the result.

Quick Answer

An x y graph (also called an XY scatter plot) places each data pair at a coordinate $(x, y)$ on a two-axis grid [1].
The horizontal axis is the x-axis. The vertical axis is the y-axis [2].
The independent variable usually goes on the x-axis and the dependent variable on the y-axis [3].
Plotting a point means finding its x value along the horizontal axis, then moving up or down to its y value.
Clustering or a linear shape in the points suggests the two variables are correlated [1].

What an XY Graph Means

In plain terms, an x y graph is a picture of paired data. You have two measurements for each item, and each item becomes one dot. A student's study hours and exam score form a pair. A month and its carbon dioxide reading form a pair. Plot enough pairs and the dots reveal a trend, a cluster, or no pattern at all.

The precise definition is narrower. An XY scatter plot displays a collection of points where the value of one variable sets the position on the horizontal axis and the value of the other variable sets the position on the vertical axis [3]. Each marker represents one data point, and every data point must carry two pieces of data, its x coordinate and its y coordinate [1]. When one variable is systematically changed by the other, that control variable is the independent variable and is customarily plotted on the horizontal axis, while the measured or dependent variable goes on the vertical axis [3]. If no dependent variable exists, either variable can go on either axis, and the plot shows only the degree of correlation, not causation [3].

How It Works

The mechanism is a coordinate system. Every point is written as an ordered pair.

$$(x, y)$$

$x$ is the value read along the horizontal axis.
$y$ is the value read along the vertical axis.

To place a point, start at the origin where the axes meet, move right by the x value, then move up by the y value. A point at $(4, 72)$ sits four units right and seventy-two units up.

Once the points are on the grid, you can summarize the relationship with two numbers. The Pearson correlation $r$ measures how tightly the points follow a straight line:

$$r = \frac{\text{cov}(x, y)}{s_x \, s_y}$$

$\text{cov}(x, y)$ is the covariance, the average tendency of $x$ and $y$ to move together.
$s_x$ is the sample standard deviation of the x values.
$s_y$ is the sample standard deviation of the y values.

You can also fit a least-squares line, $y = b_0 + b_1 x$, where $b_1$ is the slope and $b_0$ is the intercept. The slope tells you how much $y$ changes for a one-unit change in $x$.

Worked Example

Six students reported their weekly study hours and their exam scores. The table below holds the raw pairs.

study_hours (x)	exam_score (y)
1	52
2	58
3	65
4	72
5	80
6	88

Here is the walkthrough with the computed values.

Number of points: $n = 6$.
Mean of x: $\text{mean}_x = (1 + 2 + 3 + 4 + 5 + 6) / 6 = 3.5000$.
Mean of y: $\text{mean}_y = (52 + 58 + 65 + 72 + 80 + 88) / 6 = 69.1667$.
Sample standard deviation of x: $s_x = 1.8708$.
Sample standard deviation of y: $s_y = 13.5413$.
Covariance: $\text{cov} = 25.3000$.
Pearson correlation: $r = 25.3000 / (1.8708 \times 13.5413) = 0.9987$.
Slope: $b_1 = 25.3000 / 3.5000 = 7.2286$.
Intercept: $b_0 = 69.1667 - 7.2286 \times 3.5000 = 43.8667$.
Predicted score at 4 hours: $y = 43.8667 + 7.2286 \times 4 = 72.7810$.

The same numbers come out of a short Python script.

import numpy as np
hours = np.array([1, 2, 3, 4, 5, 6])
scores = np.array([52, 58, 65, 72, 80, 88])
r = np.corrcoef(hours, scores)[0, 1]
slope, intercept = np.polyfit(hours, scores, 1)
print(f"r = {r:.4f}, slope = {slope:.4f}, intercept = {intercept:.4f}")

Output:

r = 0.9987, slope = 7.2286, intercept = 43.8667

The scatter plot of these six pairs with the least-squares line $y = 43.87 + 7.23x$ and $r = 1.00$ shows the points sitting almost exactly on the line. If you want to build the same chart yourself, this step-by-step guide to creating an XY graph in Excel walks through the menu choices.

How to Interpret It

Read the shape first, then the numbers.

Direction. Points rising left to right mean a positive relationship. Points falling mean a negative one.
Strength. Tight clustering along a line means a strong relationship. A loose cloud means a weak one. The correlation $r$ runs from $-1$ to $1$, and $0.9987$ in the example is about as tight as real data gets.
Shape. A curve means the relationship is not linear, even if $r$ looks respectable. A residual plot is the standard way to check this, and this guide to interpreting residual plots covers the patterns to look for.
Outliers. A single point far from the rest can pull the line and inflate or deflate $r$. Check whether it is a data error before trusting the fit.

If the points cluster or bunch in a certain configuration, for example if they tend to form the shape of a line, that indicates the two sets of data are correlated in some way [1].

When to Use It (and when not to)

Use an x y graph when both variables are continuous and you want to compare two sets of values for each item [1]. Time is a continuous variable, so a scatter plot is the logical choice when you plot a measurement against time [4]. Bivariate graphs also let you see trends and patterns in large volumes of data without sorting through a cumbersome table [2].

Do not use an x y graph when one variable is categorical. If one set of data falls into categories with one entry per category, such as a statistic that happens every year or for every age group, that data is better served as categories on a different chart type [1]. A column graph does not work when the horizontal variable is continuous, because it treats that variable as categorical [4].

XY Graph vs Line Graph

These two look similar and behave differently. The distinction is how the horizontal axis spaces its values.

Feature	XY graph (scatter)	Line graph
Horizontal axis	Continuous, proportional spacing	Discrete, equal spacing between labels
Best for	Two continuous variables	Values tracked across equal intervals
Effect of uneven x values	Preserved on the scale	Compressed to equal gaps
Typical use	Correlation, regression	Trends over time

The difference matters. Consider the pairs (1,2), (4,8), (5,10), (8,16), (9,18). This is a linear relationship where $y = 2x$. An XY plot distributes the x values proportionally along a linear scale. A line graph spaces them equally, so the intervals from 1 to 4 and 5 to 8 look the same length as the intervals from 4 to 5 and 8 to 9, even though they are three times as long, which misrepresents the relationship [4]. If you are unsure which chart fits your data, this comparison of charts versus graphs sets out the trade-offs.

Common Mistakes

Swapping the axes. Putting the dependent variable on the x-axis reverses the story. Keep the independent variable horizontal and the dependent variable vertical [3].
Using a line graph for uneven x values. A line graph treats the independent variable as discrete and spaces the labels equally, which distorts the slope [4]. Use an XY scatter instead.
Plotting categories on the x-axis. Categories belong on a bar chart, not a scatter plot [1]. Save the XY graph for continuous variables.
Reading causation from a tight cluster. A high correlation shows the variables move together, not that one causes the other [3].
Ignoring outliers. One extreme point can dominate the slope and the correlation. Inspect it before you report the fit.
Forgetting that software needs instructions. XY charts can be arranged in various configurations, so many tools cannot build them automatically and need you to define how the data should be used [1].

Limitations

An x y graph shows association, not cause. When no dependent variable exists, the plot illustrates only the degree of correlation between two variables [3]. A strong $r$ value is compatible with a third variable driving both, or with pure coincidence in a small sample.

The plot is also only as good as the axes. A truncated y-axis that starts at 50 instead of 0 exaggerates small differences. Unequal axis scales can make a gentle slope look steep. And with more than two variables, a single XY graph cannot show the full picture, so you need color, faceting, or a different method entirely.

Frequently Asked Questions

What is the difference between the x-axis and the y-axis?

The x-axis is the horizontal line and the y-axis is the vertical line [2]. Each point on the graph is located by its x value along the horizontal axis and its y value along the vertical axis [3]. By convention, the independent variable goes on x and the dependent variable goes on y.

Which variable goes on the x-axis?

The independent, or control, variable goes on the horizontal axis, and the measured or dependent variable goes on the vertical axis [3]. If you are plotting exam score against study hours, study hours is the control variable and belongs on x. When neither variable depends on the other, either can go on either axis [3].

How do I plot a point on an x y graph?

Find the x value on the horizontal axis, then move vertically to the y value and mark the spot. For the pair $(4, 72)$, move four units right and seventy-two units up. Repeat for every pair in your data set, then look at the shape the points form.

Can an x y graph show more than one series?

Yes. XY charts can have more than one series, and the data points in one series all share the same marker style [1]. Using different markers or colors per group lets you compare two or more sets of pairs on the same axes.

What does a straight line of points mean?

It means the two variables are strongly correlated and the relationship is close to linear [1]. In the worked example, the six points produced $r = 0.9987$, which is nearly a perfect straight line. Check a residual plot before treating the line as a reliable model.