Analysis Tools

Analysis

This tool will generate graphs from a dataset you provide. While each type of graph has some parameters specific to it, the parameters under All Graphs apply to any type graph you choose. Note that some will be set by default, but replaced if you provide an alternative. A template file to help you generate properly formatted datasets can be dowloaded here.

All Graphs

Title:

If the test results are saved, this will be the title of the page that displays them.

Paste Data:

Dataset copied from Excel with the appropriate format.

Variable(s): *

The column number(s) of the continuous variable(s) to be graphed. Where applicable, this designates the Y-axis. If two or more variables are specified (e.g., 1-3,5,8), they will be drawn in different graphs. For multiple series to appear in one graph, organize the dataset such that all of the numbers for one graph are in a single column, with a second column containing grouping codes for each value.

Image dimensions

This determines the size of images that are produced. The units are approximately inches, though they will appear smaller on many monitors.

Height:

Width:

Bar Graph

Bar graphs are used to show how a continuous variable varies across groups of a categorical variable. For example, a bar graph could show how the species richness varies at three different sites.

Create bar graph

Factor:

Specify the column containing the categorical variable. Please note, you can only specify one column here; if you are making multiple boxplots from multiple variables in your dataset, this column will be treated as the grouping variable for all of those variables.

Apply Function:

What computation should be applied to the data within each group? If your "groups" are single values already, please select sum.

Draw Error Bars

Use this only if graphing an average; it will show the standard error associated with that average.

Title:

Main title to be displayed at the top of the graph.

Pie Chart

Pie charts show the size of some value (e.g., sum or average) for each of several groups as a fraction of the total for all groups. In this way pie charts are very similar to bar graphs.

Pie charts are particularly effective when at least one "slice" is much bigger than the others. Because pie charts show relative data, they should not be used when it is necessary to compare one chart with another.

Create pie graph

Factor:

The column in your dataset that designates which groups the numeric data are in. Please note, you can only specify one column here; if you are making multiple boxplots from multiple variables in your dataset, this column will be treated as the grouping variable for all of those variables.

Title for Graph:

Main title to be displayed at the top of the graph(s).

Apply Function:

What computation should be applied to the data within each group? If you only have one value per group, please select sum.

Histogram

A specific type of bar graph for showing the distribution of a single variable. The x-axis has subsets of the data range (classes), and the y-axis has the frequency that data points fall into each class. See the Wikipedia article for more information (e.g., on the different formulas for determining class breaks).

If you are at the beginning stages of data analysis, plot a histogram (and/or boxplot) to assess the structure of the data (e.g., if there are outliers, if there is high/low variability, if the data fall equally on either side of the mean, etc.). Take particular note of whether or not data are "normally dstribued" (i.e., follow a bell curve). Many statistical tests requires data to be normal to function properly (e.g., the t-test).

Create histogram

Show labels

Class Breaks

The number of classes determines how large the bins are when computing frequencies. The default method is Sturges' rule.

Determined by a Formula:

Sturges' rule (suboptimal if n<30)

Scott's choice

Freedman-Diaconis' choice

Use the following number of classes:

This number may not be followed exactly.

Title for Graph:

Main title to be displayed at the top of the graph(s).

Boxplot

Boxplots are useful for showing the distribution of multiple datasets or groups within a datset side by side (that is, if the numbers are on the same scale). Like histograms, they are helpful for assessing variability, symmetry, and identifying outliers.

Boxplots (like histograms) help you visualize how data are distributed. The bold line in the center of the box is the median, the box itself represents the interquartile range (half of the datapoints fall within the box), and the whiskers extend to 1.5x the interqurtile range. Values outside the whiskers can be considered outliers. A boxplot is similar to a histogram, but shows the bird's eye view.

Create boxplot

Factor:

If you are graphing several groups together (e.g., different experimental treatments), specify the column containing the categorical variable. Please note, you can only specify one column here; if you are making multiple boxplots from multiple variables in your dataset, this column will be treated as the grouping variable for all of those variables.

Legend Position:

A legend will only appear when Factor is set.

Title:

Main title to be displayed at the top of the graph.

Scatter Plot

Plots a continuous variable against another continuous variable (a Y-variable vs. an X-variable). Use this when the order of values is not important (you may want to make a line graph if it is).

Each variable you list under All Graphs will be a Y-variable, and appear in a separate graph. The variables specified here are the corresponding X-variables. Please be sure the number of entries for Y and X match (e.g., 1,2,3 for Y and 4,5,5 for X; yes, a variable can be used more than once). If you have more than one factors/groups in your Y-variables, and want to plot each with a different color, designate the column containing these groups under Factor (only one column for this). Note that all groups are assumed to have the same units; it is not possible to graph multiple Y-variables against the same X-variable on different scales with this tool.

Create scatterplot

X-Axis Variable:

The column number(s) of the continuous variable(s) to be used for the x-axes (see instructions above).

Factor:

If you are graphing several groups together (e.g., data from each of several species), specify the column containing the categorical variable. Please note, you can only specify one column here; if you are making multiple boxplots from multiple variables in your dataset, this column will be treated as the grouping variable for all of those variables.

Legend Position:

A legend will only appear when Factor is set.

Title:

Main title to be displayed at the top of the graph.