How to visualize survey results: a guide for high school researchers

This post answers a specific question: how do you turn raw survey data into clear, accurate visualizations that belong in a published research paper? It is written for high school students who have already collected survey responses and need to present their findings in a way that meets academic standards. By the end of this guide, you will know which chart types to use for which data, what errors to avoid, and how to write about your visuals in a results section. If your survey-based research is ready for peer review, the Princeton Journal of Pre-Collegiate Research publishes original student work across all academic disciplines.

Introduction

Survey data is among the most commonly misrepresented data type in student research papers. The American Statistical Association has noted that chart choice errors, including using pie charts for continuous data or bar charts for ordinal scales without acknowledging their limitations, are a leading cause of misleading quantitative reporting in non-specialist publications. Knowing how to visualize survey results is not a design skill. It is an analytical one. The chart you choose determines what your reader concludes. A wrong choice does not just look unprofessional; it misrepresents your findings. This guide addresses the specific decisions high school researchers face when moving from raw responses to publication-ready figures.

How do you visualize survey results in a research paper?

To visualize survey results accurately, match your chart type to your data type: use bar charts for nominal or ordinal categorical data, histograms for continuous numerical distributions, and Likert scale stacked bar charts for attitude or agreement questions. Each figure must include a descriptive title, labeled axes, and a reference in the body text. Most high school survey datasets are best represented with three to five focused figures rather than one chart per question.

The core principle is that your visualization must communicate one clear finding per figure. Researchers who try to show everything in a single chart produce visuals that readers cannot interpret. Before you open any graphing tool, sort your survey questions by data type. This step determines every chart decision that follows.

Survey questions generally produce three types of data. Nominal data has no natural order, for example, questions asking respondents to select their school subject. Ordinal data has a meaningful order but unequal intervals, for example, a satisfaction scale from “very dissatisfied” to “very satisfied.” Continuous or interval data has equal spacing between values, for example, a question asking for a numeric score from 0 to 100.

For nominal data, use a horizontal bar chart. Bar charts allow direct comparison across categories and remain readable when category labels are long. Pie charts are appropriate only when you have fewer than five categories and you are showing proportions of a whole. Even then, bar charts are generally preferred in academic contexts because they allow more precise visual comparison.

For ordinal Likert-scale data, the stacked bar chart is the standard in social science research. A stacked bar chart shows the full distribution of responses across each scale point in a single figure. It is particularly useful when you are comparing responses across two or more groups, for example, comparing agreement levels between students in different grade years.

For continuous data, use a histogram rather than a bar chart. A histogram shows the frequency distribution of a numeric variable. The bars touch each other, signaling that the variable is continuous. Bar charts have gaps between bars, signaling discrete categories. Using a bar chart for continuous data is one of the most common errors in student research papers, and reviewers notice it.

Every figure in your paper must be numbered sequentially and given a descriptive caption below the figure. For example: “Figure 2. Distribution of reported weekly study hours among survey respondents (n = 84).” Every figure must also be referenced explicitly in your results section. A figure that is not mentioned in the text does not belong in the paper. For guidance on writing the prose that accompanies your figures, the post on how to write a results section that shows findings covers this in detail.

What makes a survey visualization publishable rather than just presentable?

A publishable visualization does more than display data accurately. It supports the specific claim you are making in your results section. This distinction separates a strong class project from a paper that meets journal standards.

When a reviewer reads your results section, they look at each figure and ask two questions: does this figure show what the author says it shows, and does the figure reveal anything the author has not addressed? If the answer to the second question is yes, the paper will receive a revision request. This means your figures must be fully interpreted, not just presented.

Full interpretation means stating the direction and magnitude of the pattern you observe, not just describing that a pattern exists. Instead of writing “Figure 3 shows that most respondents agreed with the statement,” write “Figure 3 shows that 67 percent of respondents agreed or strongly agreed with the statement, compared with 12 percent who disagreed or strongly disagreed.” The specific percentages give the reader something verifiable. The comparison between the two ends of the scale gives the finding analytical weight.

A publishable figure also acknowledges its sample size. Every caption should include the n value for the data shown. If a subgroup analysis shows a pattern among a subset of respondents, the caption must reflect the smaller n. A pattern observed in 12 out of 14 respondents in a subgroup is not the same as a pattern observed in 58 out of 84 respondents overall, and the figure must make that visible.

For students whose research involves survey design and data collection, the post on how to design a survey for a research study addresses how question structure affects the type of data you collect and, by extension, the visualizations available to you.

What are the most common mistakes students make when visualizing survey results?

The most common errors in student survey visualizations are chart type mismatches, missing axis labels, truncated y-axes, and presenting every question as a separate figure regardless of relevance. Each of these errors is distinct, correctable, and frequently cited in peer review feedback.

The first error is using a pie chart for ordinal or continuous data. Pie charts show part-to-whole relationships for nominal categories. When students use them for Likert scale data, they obscure the ordered nature of the responses. A reviewer reading a pie chart for a five-point agreement scale cannot assess whether responses clustered at the positive or negative end. A stacked bar chart preserves that information. The fix is to audit every pie chart in your paper and ask whether the variable has a natural order. If it does, replace the pie chart.

The second error is omitting axis labels or units. A bar chart showing response frequencies without a labeled y-axis forces the reader to guess whether the values are raw counts or percentages. This is not a minor formatting issue. It is an interpretive failure. Every axis must be labeled, including the unit of measurement where applicable.

The third error is truncating the y-axis to exaggerate differences. Starting a bar chart’s y-axis at a value above zero makes small differences appear large. In academic publishing, this is treated as a form of misrepresentation, even when unintentional. The y-axis for frequency or percentage data must begin at zero.

The fourth error is including a figure for every survey question. Not every question produces a finding worth visualizing. If a question shows an even or unremarkable distribution that does not relate to your research question, it belongs in a supplementary table, not in a figure. Reviewers interpret an excess of figures as a sign that the author has not identified what the data actually shows. Select figures that support a specific claim in your results section. For context on how unexpected or null findings should be handled, the post on what to do if research results do not support your hypothesis is directly relevant.

How to visualize survey results for a research paper, step by step

1. Classify every survey question by data type. Label each question as nominal, ordinal, or continuous before opening any graphing software. This classification determines your chart type for every figure.

2. Identify which questions are directly relevant to your research question. Only questions that address your hypothesis or central inquiry should appear as figures. Others may be summarised in a table.

3. Select the correct chart type for each relevant question. Nominal: bar chart. Ordinal Likert: stacked bar chart. Continuous: histogram. Proportional nominal with four or fewer categories: pie chart, used sparingly.

4. Build each figure with labeled axes, a descriptive caption, and the n value. Captions go below the figure. Axis labels include units where applicable. The y-axis begins at zero for all frequency and percentage data.

5. Write a corresponding sentence in your results section for each figure. Reference the figure by number. State the specific finding, including direction and magnitude. Do not leave a figure unaddressed in the text.

6. Review each figure against your research question. Ask whether the figure supports, complicates, or is irrelevant to your central argument. Remove figures that do not contribute to the argument.

7. Check your figures against the submission guidelines for the journal you are targeting. Review the submission guidelines at princeton-jpcr.org/submit for PJPCR’s figure formatting requirements before finalising your manuscript.

PJPCR publishes original research across all academic disciplines, including survey-based studies in the social sciences, psychology, and public health. If your survey research is complete and your findings are ready for peer review, review the submission guidelines at princeton-jpcr.org/submit.

Frequently asked questions about how to visualize survey results

What is a Likert scale and how should it be visualized?

A Likert scale is an ordinal rating scale, typically with five or seven points, used to measure attitudes or agreement. The standard format ranges from “strongly disagree” to “strongly agree.” Likert scale data should be visualized using a stacked bar chart, which preserves the ordered nature of the responses and allows comparison across groups or questions. Avoid pie charts for Likert data, as they remove the ordering information that makes the scale meaningful.

How many figures should a high school research paper include for survey data?

Most survey-based research papers at the high school level include between three and six figures. The number is determined by how many questions directly address the research question, not by the total number of survey items. Including a figure for every question is a common error that reviewers flag. A focused set of well-interpreted figures is more credible than a large set of underanalysed ones. Each figure must be referenced and interpreted in the results section.

Do I need statistical software to visualize survey results?

Statistical software is not required to produce publication-quality survey visualizations. Microsoft Excel, Google Sheets, and Canva all support the chart types used in most student research papers, including bar charts, histograms, and stacked bar charts. What matters is that the chart type matches the data type, the axes are correctly labeled, and the y-axis begins at zero for frequency data. Software choice does not affect the quality of a visualization; accuracy and clarity do.

What makes a survey visualization strong enough to be published in an academic journal?

A publishable survey visualization accurately represents the data, uses the correct chart type for the variable, includes a descriptive caption with the sample size, and is directly connected to a claim in the results section. Reviewers assess whether the figure could mislead a reader, whether the axes are honest, and whether the author has interpreted the finding rather than simply displayed it. A figure that requires the reader to guess what it shows will not pass peer review. For examples of published student research, browse the published issues at princeton-jpcr.org.

What kinds of survey-based research does PJPCR publish, and is the peer review process rigorous?

PJPCR publishes original survey-based research across disciplines including psychology, sociology, public health, education, and political science. Submissions are evaluated through a rigorous peer review process conducted by qualified reviewers. The journal does not guarantee acceptance; papers are assessed on the originality of the research question, the appropriateness of the methodology, and the accuracy of the analysis and presentation. The standard review timeline is 2 to 3 months. A fast-track option is available for students who need a quicker turnaround.

Conclusion

Visualizing survey results accurately requires three decisions made in sequence: classify your data type, select the matching chart type, and interpret each figure explicitly in your results section. The most consequential errors, chart type mismatches, unlabeled axes, and truncated y-axes, are all correctable before submission. A paper with three precise, well-interpreted figures is stronger than one with twelve unanalysed charts. If your survey-based research is complete and your findings are clearly presented, submit it to PJPCR. Review the submission guidelines at princeton-jpcr.org/submit to confirm your manuscript meets the journal’s requirements before you submit.