Research Guides: Data Visualization: Choosing the Best Visualization Type

Introduction

Different disciplines have different standards and approaches when it comes to visualization of data. As you do your literature search and discover research in your topic of interest, you may notice different kinds of graphics used to describe data. In many cases, the most effective data visualization approaches can be determined when you think about what kinds of data that you have, and how much information that you want to share. The table below outlines some common graphing or visualization strategies depending on the type of data you may be working with.

If you need help choosing the appropriate visualization type for your data, here are some resources to help:

Chart Chooser Diagram
A flow chart to help you decide what visualization type to use based on what you want to show and the data you have
Data to Viz
A decision tree that leads you the most appropriate graph for you data and further links to the code to build it and lists common caveats you should avoid
The Data Visualization Catalogue
Catalogue of different data visualizations that can be searched by function or type, links to pages with more detailed information
Tableau: Visual Best Practices - Chart Choice Overview of charts types in Tableau and their appropriate usages

Visualization Types

Data Types	Graphic Types
Quantitative Data	Histograms Histograms allow a reader to view the frequency distribution for a single quantitative variable class="collapse" id="collapseEX1"> Source: Palmer Penguins Data Set Two Variables Scatter Plots Scatter plots can be used to view the relationship between two measurements. Useful when working with up to hundreds of observations. class="collapse" id="collapseEX2"> Source: College Tuition Data Line Plots Line plots allow you to see the trend or linear relationship between two measurements (e.g. change over time) class="collapse" id="collapseEX3"> Source: Incarceration Trends Dataset
Qualitative Data	Word Frequency Word clouds use size to represent the frequency of different words in a particular text. class="collapse" id="collapseEX4"> Source: Text scraped from this research guide Timelines Depiction of events that occur over time class="collapse" id="collapseEX5"> Source: War Data Pie Charts Pie charts are best used for depicting the proportion of qualitative responses relative to a whole Caution: If there are too many response types, or if multiple different responses appear in similar proportions, it can be difficult to visualize differences between responses. class="collapse" id="collapseEX6"> Source: War Data Mind Maps Depiction of an arrangement of ideas or logical flow Cladograms Depiction of the similarity or difference between a variety of qualitative categories.
Mix of Quantitative and Qualitative Data	Bar and Column Charts These plots are frequently used to compare measurements collected between different groups. Caution: Bar and column plots may not show the spread or range of data which could significantly impact how the data can be interpreted. class="collapse" id="collapseEX9"> Source: Berkeley Lab Energy Data Boxplots A very common visualization method that depicts basic information on the spread of a quantitative measurement for different categories. class="collapse" id="collapseEX10"> Source: Palmer Penguins Data Set Dot Plots Useful visualization that depicts every measurement grouped by their respective categories. Useful when you have relatively fewer measurements in each category. Violin Plots Visualization depicting the frequency of a quantitative measurement separated into categories. Useful when you have many measurements in each category. class="collapse" id="collapseEX12"> Source: Palmer Penguins Data Set
Complex Data Visualizations	Three Variables Heatmaps Visualization depicting the quantitative measurement intensity between two related qualitative categories. Can be paired with cladograms to view and organize categorical relationships. Density Plots Visualization depicting the measurement intensity between three quantitative variables on a two-dimensional plot. Most useful when working with thousands or tens of thousands of measurements. Many Variables Principal Component Analysis A qualitative depiction of the primary factors impacting similarity between observations with many simultaneously measured variables in a two dimensional plot. Useful when working with dozens, hundreds or thousands of quantitative measurements across multiple variables.