Comparing data is a fundamental aspect of data analysis and visualization. In isolation a data point has little to no meaning, and only by comparing it with other data can we start to derive some understanding about what the significance of the data. We can extend this to understanding relative magnitudes and differences between data and begin to identify patterns or trends. Comparing data also facilitates decision making by understanding different scenarios and possible outcomes. Most data visualizations can become functions of comparison by duplication. So, for example, the box and whisker plot seen below by itself is a visualization of dispersion and range, but by creating two or more plots we can gain understanding through comparison.
Please see below for more charts that are frequently used to compare data:
A Vertical bar chart B. Horizontal bar chart
Usage Notes: Bar charts are used to compare magnitudes of categorical data. Bar charts are often presented vertically, though a a horizontal chart may be preferable when the label is very long, when included in a dashboard where the direction or reading is from left to right, or when the data represent ranked categories. Can be used to visualize distribution as well. A vertical bar chart is also known as a column chart and some authorities strictly distinguish between column and bar charts
Also see: Histogram, Stacked Bar chart Bullet chart, Marimekko chart, Multi-set bar chart, Span chart, Radial bar chart
Usage Notes: Bubble charts are a type of scatter plot used to compare three dimensional data. Two values are represented by the x and y and a third value is represented by a dot of varying size representing the magnitude of the z-value.
Also see: Scatterplot, dot density map.
Usage Notes: A Bullet chart is primarily used for comparing a performance measure. The components of the bullet graph are an achievement bar in the middle that represents the data, surrounding this is a comparison indicator such as a previous performance, and a perpendicular bar called the target.
See also: Bar chart with reference line, thermometer chart
Usage Notes : Heat Maps are used to visualize and compare three continuous or categorical variables. Heat map are especially useful for comparing ordered data that ranges in sequences such as time or temperature, etc. The Heat map organizes observations in a matrix according to two variables and visualizes a third one with color gradients.
Usage Notes: Line charts are useful for comparing continuous trends most often over time. The lines graph is a cartesian coordinate plot chart with the points being connected to indicate the order of the trend and illustrate more clearly the changes in magnitude between one observation and the next. Each dataset is visualized in different colors or line styles. For comparisons of categorical or discrete data bar charts are preferable.
Usage Notes: The Marimekko, also known as a 'Mekko' chart, is designed to visualize data across three dimensions. These dimensions can encompass a blend of either two numeric and one categorical variable or two categorical and one numeric variable. Mekko charts excel at comparing categorical data, especially when the number of categories is limited. When the x-axis is a numeric variable like a percentage the widths of the column can be varied to represent proportion.
See also: tree graph, stacked bar chart.
Usage Notes: A Multi-set car chart, also called a Grouped bar chart, is useful for comparing multiple categories segmented by another variable. The x-axis typically represents a categorical variable, further divided into several sub-categories. The y-axis represents a numeric value.
See also: Bar chart, Stacked bar chart
Usage Notes: This is sometimes mistakenly called a radial bar chart, and is similar to the Radial Column Chart both of which, see below.
Usage notes: Population pyramids are used to represent and compare typically two demographic variables, one numeric and one categorical, for a specific time. Population pyramids are most often seen in visualizations of age by sex. A third variable can be incorporated with gradient shading of the individual population cohorts (e.g. females aged 45-50, and average income).
See also: Diverging bar chart, Bar chart, Stacked bar chart, histogram.
Usage notes: The Radial Column chart is useful for representing directional data where the value is represented at the compass degree position such as wind direction or other geographical orientations; periodic or cyclical data representing time of day month or year, etc, or sequential data where the sequence can be logically ordered around the circularity of the graph. The radial column chart is more compact than the bar chart and is useful for highlighting differences in values, especially when visualized with a polar grid. This chart is very similar in appearance to the Nightengale Rose Chart (which, see).
See also: Bar chart, Nightengale rose chart
Usage Notes: The Span chart is useful for highlighting and comparing durations and ranges. While it is arguably most often used for representing time durations, it can also be used for visualizing the range of values between minimum and maximum levels across various categories or groups. This visualization can be illustrative in scenarios where understanding the extent and overlap of ranges is important. The span chart can also be oriented vertically. The span chart is also known as Range chart, Floating bar chart, Difference chart, High-Low chart.
See also: Gantt chart.
Usage Notes: Stacked area charts are used to compare part ot whole relationships. Each layer in a stacked area chart represents a variable and the area of the layer represents the variable's total value. The x-axis typically represents a continuous or interval vairable like time, and the y-axis represents both the cumulative total of all variables and the individual value of specific variables. This chart is useful for comparing relative proportions of individual variables to each other and to the total value of the dataset, where all values are positive.
See also: area chart, stacked bar chart.
Usage Notes: Stacked bar charts are used to compare magnitudes of categorical data. Bar charts are often presented vertically, though a a horizontal chart may be preferable when the label is very long, when included in a dashboard where the direction or reading is from left to right, or when the data represent ranked categories. Can be used to visualize distribution as well.
See also: Histogram, Stacked bar chart Bullet chart, Marimekko chart, Multi-set bar chart, Span chart, Radial bar chart, Population pyramid.
Usage Notes: Word clouds are used to visualize a collection of words and shows their frequency by size and/or color. More precise visualizations can be accomplished with other types of charts
