A scatterplot is a two-axis graph that shows how two quantitative variables relate by displaying each observation as a single plotted point.
What Is a Scatterplot?
In social science research, a scatterplot is a graph that helps visualize the relationship between two numerical variables. Each dot on the graph represents a single case, such as a person, group, or unit of analysis. One variable is plotted on the horizontal axis (x-axis), and the other is plotted on the vertical axis (y-axis). The position of each point reflects the values of both variables for that case.
Scatterplots are also known as scatter diagrams or scattergrams. While the terms are sometimes used interchangeably, “scatterplot” is the most common name in data analysis. These visual tools are especially useful for identifying patterns, testing assumptions, and spotting outliers before conducting more formal statistical procedures.
Why Scatterplots Matter in Social Science
Social science researchers often want to explore how two variables relate. For instance, they may ask whether more education is associated with higher income or whether stress levels decrease with age. Scatterplots allow researchers to inspect these relationships quickly and visually. Instead of relying solely on numbers or summary statistics, a scatterplot presents the data as a full picture.
This kind of visual inspection offers several benefits:
- It helps determine whether a linear or nonlinear relationship exists.
- It shows the strength and direction of the relationship.
- It highlights any unusual or extreme cases (outliers).
- It can suggest which variables may be predictive of others.
In short, scatterplots serve as both an exploratory and confirmatory step in social science research.
Key Elements of a Scatterplot
Axes
The x-axis typically represents the independent variable, the one presumed to influence or predict change. The y-axis represents the dependent variable, the outcome being measured.
Example: If a researcher is examining how study hours affect exam scores, hours would be on the x-axis and scores on the y-axis.
Data Points
Each point on the scatterplot represents a single observation. If a survey includes 150 participants, the scatterplot will contain 150 points. The location of each point depends on the values of both variables for that person or unit.
Pattern or Shape
The arrangement of points reveals the type of relationship:
- A rising pattern suggests a positive relationship.
- A falling pattern indicates a negative relationship.
- A scattered or flat pattern shows no clear relationship.
Some scatterplots also include a line of best fit or trend line to highlight the overall direction of the data.
Types of Relationships Visible in a Scatterplot
Positive (Direct) Relationship
As the value of the x-variable increases, the value of the y-variable also increases. The dots form an upward-sloping pattern from left to right.
Example: A political scientist might find a positive relationship between political knowledge and likelihood of voting.
Negative (Inverse) Relationship
As the x-variable increases, the y-variable decreases. The dots form a downward-sloping pattern.
Example: A psychologist might find a negative relationship between daily stress levels and quality of sleep.
No Apparent Relationship
If the dots are widely scattered without any upward or downward trend, it suggests little or no relationship between the two variables.
Example: A sociologist might find no relationship between shoe size and job satisfaction.
Curvilinear Relationship
This occurs when the relationship is nonlinear—such as a U-shape or an inverted U-shape. Scatterplots make it easy to spot these more complex relationships.
Example: Moderate levels of anxiety might improve performance, but both very low and very high anxiety levels reduce performance.
When to Use a Scatterplot
Scatterplots are used when both variables are numerical—either interval or ratio level—and when a researcher wants to examine the association between them.
Common uses include:
- Exploring relationships before running statistical tests like correlation or regression.
- Checking for linearity.
- Identifying outliers.
- Detecting clusters or patterns.
- Visualizing model predictions.
They are often used during data exploration and in presenting findings visually to others.
Example from Social Science Research
Imagine a criminologist is interested in the relationship between unemployment rates and crime rates in different neighborhoods. They gather data from 50 neighborhoods, record the unemployment rate (x-axis) and the number of crimes per 1,000 residents (y-axis), and plot the results.
If the scatterplot shows that neighborhoods with higher unemployment tend to have higher crime rates, the researcher can then test that relationship more formally using correlation or regression analysis. The scatterplot guides their expectations and helps spot anything unusual, like a high-crime neighborhood with low unemployment.
Differences Between Scatterplot and Related Terms
While scatterplot, scattergram, and scatter diagram refer to the same type of graph, “scatterplot” is more common in statistical software and academic texts. Some researchers use “scattergram” in teaching or presentations, but the differences are mostly in terminology, not in function.
Strengths of Using Scatterplots
- Clarity: They make patterns easy to see, even for non-specialists.
- Transparency: They show the entire dataset rather than summarizing it.
- Exploratory power: They help determine if variables are related before applying statistics.
- Diagnostic use: Scatterplots help assess assumptions such as linearity or homoscedasticity in regression analysis.
Limitations of Scatterplots
- Limited to two variables: They can only show bivariate relationships.
- Not suited for categorical data: Both variables must be numerical.
- Overplotting in large datasets: Too many points may clutter the graph and obscure patterns.
- Do not imply causation: A visible trend does not mean one variable causes changes in the other.
Best Practices for Creating Scatterplots
Label Clearly
Both axes should be labeled with variable names and units of measurement if applicable.
Choose Appropriate Scale
Avoid compressing or stretching the axes in a way that distorts the data. Use a consistent and logical scale.
Add a Trend Line When Useful
A line of best fit helps the viewer quickly grasp the general direction of the relationship.
Consider Color or Grouping
In more advanced versions, scatterplots can include color-coded points to show subgroups (e.g., by gender or region) or include multiple scatterplots side by side.
Watch for Outliers
Visually check for data points that fall far from the general pattern. Outliers may indicate important exceptions or possible data entry errors.
Examples from Social Science Fields
Sociology
A scatterplot might show the relationship between neighborhood diversity and trust in local institutions.
Psychology
Psychologists may plot perceived stress levels against hours of sleep to explore how stress impacts rest.
Education
Education researchers might explore how attendance relates to test scores by plotting total absences against exam performance.
Political Science
A scatterplot could show how media consumption relates to political knowledge or civic engagement.
Criminology
Researchers might use scatterplots to explore how population density relates to property crime rates.
Conclusion
Scatterplots are vital tools in social science research. By mapping one variable against another using simple plotted points, researchers can visually assess relationships, detect outliers, and decide how to proceed with analysis. Whether the goal is exploration, diagnosis, or communication, scatterplots offer an accessible and effective way to see how variables move together—or don’t. Their simplicity and visual power make them a core part of the research process.
Meta description:
Glossary Return to Doc's Research Glossary
Last Modified: 03/27/2025