Graphical displays help present data in a way that allows the reader to visualize trends and observe differences in means that may not be as obvious if only the data is presented in a table. Varying types of graphical displays exist, with each being used for different purposes. Below are several commonly used graphical displays.
Histograms - Column bar charts that show frequency distributions. Each measurement is placed into a bin (a range of values). Often, the extreme values of a variable occur infrequently while the most common values occur somewhere between the extremes.
Pie Charts - Circular representations of relative (percentage) frequency distributions. The more frequent a value falls within a bin, the larger the portion of the circle.
Scatter Charts - Individual data points are placed on an x/y grid to show the the type of relationship that may exist between two the variables. If a trend appears when the data points are plotted, the data may suggest a relationship between the variables. If no logical trend appears, then it is unlikely that a relationship exists between the variables. The relationship can either be linear or nonlinear in nature.
Column Charts - Column bar charts that plot the mean values for classes of data. One standard deviation, 1 SD, is usually added to each column to show the variability of the data. The difference between the heights of the bars can suggest if the differences between the classes is significant (large enough) or not.
Box Plots - Box plots (box and whisker plots) are visual representations of the descriptive statistics, providing a better picture on the distribution and spread of the data. Outliers, values that lie outside a set range based on quartiles, can be identified and possibly excluded from the analysis.