Review Descriptive Statistics - Within Tab

Select the Within tab of the Review Descriptive Statistics dialog box to access the options described here. Use these options to review a comprehensive collection of within group descriptive statistics.

Pooled within-groups covariances & correlations

Click this button to produce a spreadsheet with the pooled within-groups covariances for all variables, and another spreadsheet with the pooled within-groups correlations. A particularly useful Graph of Input Data for these spreadsheets is the scatter plot, which will be computed to reflect the pooled within-groups correlations. Specifically, the scatter plot will show the individual values in terms of their deviations from the respective group mean. In this plot, the variable on the horizontal axis actually represents deviations of the respective cases from the means of the variable in the respective groups to which the cases belong; in short, the two axes in the plot represent within-groups deviation scores (from the respective means).

Means & number of cases

Click this button to produce a spreadsheet with the means for each variable for each group; the last column of the spreadsheet will contain the number of cases in each group and the last row of the spreadsheet will contain the overall means (collapsed across all groups, i.e., the weighted means). A particularly useful Graph of Input Data for this spreadsheet is the histogram, which can be produced for all cases or for selected groups only. In all histograms, the normal distribution will be superimposed over the observed distribution.

Note: histogram for selected groups. To produce a histogram for a selected group, highlight the desired variable (column) and the desired group (row). Then right-click on that cell and select Graph of Input Data from the shortcut menu. In the Graph of Input Data menu, select the Histogram ... Normal fit option to produce this graph.

Note: histogram for all groups combined. To produce a histogram for all groups combined, highlight the desired variable (column) and highlight the last row (labeled All Grps). Then right-click on that cell and select Graph of Input Data from the shortcut menu. In the Graph of Input Data menu, select the Histogram ... Normal fit option to produce this graph.

Within-groups standard deviations

Click this button to produce a spreadsheet with the standard deviations for each variable for each group; the last column of the spreadsheet will contain the number of cases in each group and the last row of the spreadsheet will contain the overall standard deviations for all groups.

Categorized histogram by group

Click this button to produce a categorized histogram for a selected variable. A standard variable selection dialog box will first be displayed, in which you select the variable. This graph shows the distribution of the variable in the different groups.

Box plot of means by group

Click this button to produce a box and whisker plot for a selected variable. A standard variable selection dialog box will first be displayed, in which you select the variable. Next the Box-Whisker Type dialog will be displayed, in which you select the type of box and whisker plot. This graph shows the mean or median, and different measures of dispersion for the variable (e.g., standard deviation, quartile range) in the different groups.  For details concerning the median and quartiles, see the Computation of Percentiles option in the Options dialog box - Analyses/Graphs: Limits tab. For more information, see Percentile Calculation Options in Statistica.

Categorized scatterplot by group

Click this button to produce a categorized scatterplot for two selected variables, by group. A standard variable selection dialog box will first be displayed, in which you select the two variables.

Categorized probability plot by group

Click this button to produce a categorized probability plot for a selected variable, by group. A standard variable selection dialog box will first be displayed, in which you select the variable. The normal probability plot provides a quick check of whether or not the respective variable is normally distributed in the groups. If so, the points in the plot will approximate a straight line.