Skip to main content icon/video/no-internet

Bar Graphs

Bar graphs (also called bar charts) are a type of data visualization in which data points are represented by rectangular bars. Typically, the bars extend vertically from the bottom of the x-axis up to the data value, which is plotted along the y-axis; thus, the height of the bar (physical magnitude) is analogous to the numerical magnitude of the data point; bars may also be horizontal with length representing magnitude. Each data point is either labeled along the x-axis or referenced in a legend in a separate location near the graph. This entry discusses how bar graphs are used, factors that affect comprehension of bar graphs, and implications for educational research.

Bar graphs are often used to communicate scientific results, qualitative trends, and statistical analyses, such as main effects and interactions. Because data points are aligned along a common axis, bar graphs can facilitate comparison between individual data points; for instance, a user can quickly assess whether the data points are the same or different. Additionally, a user can easily compare differences between data points by judging the relative sizes of height gaps between bars. Global patterns, such as linear trends, are also salient in bar graphs. Thus, bar graphs are generally better for visualizing qualitative data patterns than exact values, which are more easily extracted from tables. In Figure 1, a simple bar chart shows the percentage of times different classroom assessment formats were used by a sample of teachers.

Figure 1 Percentage of times for different classroom assessment formats

Figure

Factors That Affect Comprehension

The visual features of a bar graph, such how the bars are organized and how far apart the bars are from one another, can affect comprehension. Bar graph comprehension is often facilitated when the bars are grouped in various ways; for instance, bars that are clustered together along the x-axis or that have the same color are perceived as belonging to the same group and viewers are more likely to make comparisons within rather than across such groups. Additionally, larger effects (i.e., larger height differences between bars) are more salient and easier to perceive.

Because bars are usually spatially segregated in a bar graph, it is typically easier to compare discrete values than continuous values, for which line graphs are better suited. However, Jeff Zacks, Ellen Levy, Barbara Tversky, and Diane Schiano found that discrete height judgments of bars are subject to bias from neighboring elements in bar graphs (such as the presence and height ratio of nearby bars).

Additionally, the knowledge that the user brings to the graph can affect comprehension of bar graphs. One type of knowledge that affects comprehension is graphical literacy, which involves having basic knowledge about how graphs should look and what they are used for. Having graphical literacy would include, for example, knowing that independent or categorical variables are represented along the x-axis and/or legend, whereas the dependent variable is represented along the y-axis. Graph expertise is especially helpful because it can allow the user to compare data more efficiently through the use of mental manipulation and perceptual shortcuts.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading