Quality Improvement Zone

A scatter plot is a graph that helps you explore whether there appears to be a relationship between two variables. By plotting pairs of observations on a chart, it can help identify patterns, test assumptions, and generate ideas for further investigation.

Scatter plots can be used when teams want to understand how different parts of a system may be related. For example, you might explore whether:

• staffing levels are related to waiting times
• training attendance is related to confidence scores
• patient satisfaction is related to appointment delays
• smoking prevalence is related to low birth weight

Each point on the chart represents a pair of observations. The horizontal (X) axis usually contains the factor you think may be related to the outcome, and the vertical (Y) axis usually contains the measure you are interested in understanding.

Scatter plots help teams explore whether two variables appear to be related. They are particularly useful when investigating assumptions, understanding systems and identifying questions that may require further investigation.

Using a scatter plot can help you:

• explore possible relationships between variables
• test assumptions and beliefs using data
• generate hypotheses for further investigation
• identify patterns that may warrant additional analysis
• communicate relationships visually
• demonstrate when there is no evidence of a relationship between variables

Sometimes a scatter plot is valuable when it shows no apparent relationship, particularly when strong assumptions exist that the variables should be connected.

Scatter plots help teams investigate questions about possible relationships within a system. The goal is to explore patterns and generate learning rather than prove cause and effect.

1. Define the question you want to explore

Start with a clear question or hypothesis.

For example:

• Does appointment availability relate to patient satisfaction?
• Is staff attendance related to service performance?
• Does waiting time influence referral completion rates?

2. Identify the two variables

Decide which two variables you want to compare. In many cases:

• the X axis contains the factor you think may be related to the outcome
• the Y axis contains the outcome or measure you want to better understand

3. Gather paired data

Collect data in pairs so that each observation contains values for both variables. The values should relate to the same person, service, location, time period or event.

4. Create the scatter plot

Enter the paired data into two columns in Excel and create a scatter plot using the chart tools. By default, the first column is plotted on the X axis and the second column on the Y axis.

5. Look for patterns

Review the chart and consider whether the points suggest a relationship between the variables.

Common patterns include:

• positive relationship: as one variable increases, the other tends to increase
• negative relationship: as one variable increases, the other tends to decrease
• no apparent relationship: there is no clear pattern in the data
• curved or non-linear relationship: the variables appear related, but not in a straight-line pattern

6. Consider additional features

When interpreting the chart, look for:

• Strong relationships, where points cluster closely around a visible pattern
• Weak relationships, where the pattern is less clear
• Clusters, which may suggest different groups within the data
• Outliers, which may represent unusual observations that warrant further investigation

7. Consider alternative explanations

A relationship shown on a scatter plot does not prove that one variable causes the other. The relationship may be influenced by other factors, or it may be coincidental.

Scatter plots are best used to generate questions and hypotheses for further investigation rather than proving cause and effect.

8. Use the findings to inform improvement

Use the insights from the scatter plot to deepen your understanding of the system and identify areas for further exploration. Findings may prompt additional investigation using other tools.

Positive relationship

If the points tend to move upwards from left to right, this suggests that higher values of one variable are associated with higher values of the other variable.

Negative relationship

If the points tend to move downwards from left to right, this suggests that higher values of one variable are associated with lower values of the other variable.

No apparent relationship

If there is no clear pattern in the data points, there is no evidence of a relationship between the variables.

Important note

Finding a relationship between two variables does not prove that one causes the other. Scatter plots help identify potential relationships and support learning about a system, but further investigation is often needed to understand what is driving the observed pattern.