Presenting survey results accurately and clearly is just as important as how you conduct the survey. Presented well, the results of a survey are informative and enlightening. But poor presentation can confound the study and threaten your credibility as a researcher. Bar charts are easy-to-interpret representations of survey data. A bar chart can compare frequency of responses to Likert items, which measure respondents’ levels of agreement or disagreement with an issue. A typical Likert scale includes the responses, “strongly agree,” “agree,” “neither agree nor disagree,” “disagree” and “strongly disagree.”
Gather your data. Bar charts represent discrete values, not percentages. Use the raw sums of responses to each item for your bar chart.
Create a bar chart for each individual item. Each survey question should be represented in its own chart. One bar corresponds to each possible response: “strongly agree,” “agree,” “neither agree nor disagree,” “disagree” and “strongly disagree.
Make a frequency scale on one axis. Always include zero and label periodic intervals.
List the possible responses on the other axis. This is where the bars will begin. Bar charts can be situated horizontally or vertically, so the axes are interchangeable.
Create a bar for each response that represents the frequency of that response.
Label each bar chart with the substance of the Likert item measured. The exact wording of the question is preferable.
Repeat this process for each Likert item. If possible, maintain the same frequency scale for each chart so that, when viewed side-by-side, the charts measure responses equally.
About the Author
Emily Jarvis is a graduate of University of Georgia Grady College of Journalism. Her articles have appeared in "Southern Distinction Magazine" and "The Red & Black." Jarvis holds a Bachelor of Arts in magazine journalism and a Master of Arts in journalism.