Computing summary statistics for columns

This tutorial covers the steps for computing summary statistics for columns of data in StatCrunch. To begin, load the Exam scores data set, which will be used throughout this tutorial. This data set contains only one column of data containing 23 exam grades for an introductory Statistics course. To compute summary statistics using rows of data, see Computing summary statistics for rows.

Computing column statistics

To compute summary statistics for the scores in the Exam 2 column, choose the Stat > Summary Stats > Columns menu option. Select the Exam 2 column and click Compute!. By default, StatCrunch produces a table containing the following eleven summary statistics for the data: sample size, mean, variance, standard deviation, standard error, median, range, minimum, maximum, first quartile, and third quartile.

Customizing statistics displayed

Under the Statistics option, StatCrunch allows for the selection of which statistics to display. There are a number of statistics such as the interquartile range, kurtosis and skewness which can be added to the default listing or one can trim down the default listing to a smaller set of statistics. For example, in the window containing the summary statistics output above, choose Options > Edit to reopen the dialog window. Select n, Mean, and Std. dev. and click Compute!. The resulting summary statistics output shown below contains only the statistics selected.

Adding percentiles to the output

Beyond the list of statistics given, StatCrunch allows for the calculation of a user defined set of percentiles under the Percentiles option. To illustrate, in the window containing the above summary statistics output, choose Options > Edit to reopen the dialog window. The percentiles to be computed must be between 0 and 100 and separated by commas. In this case, enter 10, 90 under Percentiles and click Compute!. Values for the 10th and 90th percentiles of the scores now appear to the right of the other statistics that are computed.

Creating a custom statistic

StatCrunch can also calculate other user defined statistics that are not included the under Statistics listing. This ability is found under the Other statistics option. From the previous resulting summary statistics, choose Options > Edit to reopen the dialog window. The Other statistic option allows for the use of a wide array of StatCrunch functions to be applied to the column of data which is labeled x for the purposes of the calculation. Clicking on the blue Build on the right will open a custom expression to assist in the creation of the formula for the statistic. For this demonstration though, simply enter std(x)*std(x) under Other statistic which is equivalent to a variable's variance. Highlight Variance under the Statistics option as a way to compare and click Compute!. The custom statistic in the resulting summary statistics table has the exact same value as the variance of the exam scores.

Always Learning
Pearson