Launch the Chart platform by selecting Graph > Chart.
•

Up to two X variables, which appear on the xaxis in the same order that you assign them in the launch window.

•

As many Y variables (statistics) as you want. If the data is already summarized, select Data as the statistics option.

Select up to two X variables whose levels are categories on the xaxis. The Chart platform produces a bar for each level or combination of levels of the X variables. If you do not specify an X variable, the chart has a bar for each row in the data table.


For one or more Grouping variables, independent results are produced for each level or combination of levels of the grouping variables. These results appear in the same report window, but in separate plots. See Use Grouping Variables.


Select the type of chart that you want. Available chart types are Bar Chart, Line Chart, Pie Chart, Needle Chart, and Point Chart. Selecting a chart controls which of the other options are available.


This option is available only for Line Chart.


Adds error bars when the Mean statistic is selected for at least one Y variable and at least one X variable is assigned. This option is not selected by default.
This option is available for Line Chart and Bar Chart, and additional options are added to the Chart launch window. See Adding Error Bars.


Sets the specific quantile when the Quantiles statistic is selected for at least one Y variable. The default value is 25. Specify a different quantile:

You can plot the raw data for Y variables, or you can plot as many statistics as you want on the yaxis. The Statistics menu in the Chart launch window lists the available statistics. To specify the yaxis, highlight one or more numeric columns in the Select Columns list and select from the list of statistics. If all the statistics requested are counting statistics (for example, N) for the same column, that column is used as the category variable.
The available statistics in the Chart platform are described in the following table. They are the same as those computed by statistical platforms in the Analyze menu and the Summary command in the Tables menu.
The value of each row in the data table when there is no categorical variable. If there is a categorical variable, Data produces a point plot within the variable’s levels.


The number of nonmissing values. Also used to compute statistics when there is no column assigned as a weight variable. The Chart platform shows N for each level of a categorical variable.


The percentage of the total number of rows represented by each level of the Categories, X, Levels variable. If summary statistics are requested on a continuous variable, then the % of Total equals the proportion of the sum represented by each level of the Categories, X, Levels variable.


The sum of all values in a column assigned as Weight. Also used instead of N to compute other statistics. Chart shows the sum of the weight variable for each level of a categorical variable.


The standard error of the mean of each level of a categorical variable. It is the standard deviation, Std Dev, divided by the square root of N for each level. If a column is assigned a weight variable, then the denominator is the square root of the sum of the weights.


Divides a data set so that n% of the data is below the nth quantile. To compute a specific quantile, enter the quantile value in the box located in the lower left of the Chart launch window before requesting Quantile from the menu.

You can assign zero, one, or two X variables whose levels are categories on the xaxis. The Chart platform produces a bar (or a needle, or a pie slice, and so on) for each level or combination of levels of the X variables. If you do not specify any X variable, the chart has a bar for each row in the data table.
If you do not specify a variable for categories, most statistics produce a bar (or a needle, or a pie slice, and so on) for each observation in the data table. This is useful when your data is already summarized. In that case, you usually specify Data as the statistic to plot. Each bar reflects the value of the Y variable. See Plot a Single Statistic, for an example.


If there are multiple statistics, the Overlay option is checked by default, and the Y variables (statistics) are plotted on the same chart for each level of the grouping variable. However, the levels of the grouping variable cannot be overlaid into the same plot frame. For example, if the levels of your grouping variable are Male and Female, the Overlay option cannot be used to combine the two Male and Female graphs into one graph. To see that type of result, use Categories, X, Levels instead of Grouping variables.
Error bars are available when the Mean statistic is selected for at least one Y variable, and at least one X variable is assigned. Error Bars are not available for pie charts. Selecting Add Error Bars to Mean causes additional options to appear in the Chart launch window.