Statistics and Parameters

When a graphic element is selected in the Edit Properties Of list, you can choose a statistical summary function for the graphic element. The result of the statistic determines the position of the graphic elements on the y axis. In a 2-D chart, the statistic is calculated for each value on the x axis. In a 3-D chart, it is calculated for the intersection of values on the x axis and z axis. For example, assume that you are creating a bar chart with gender as the x-axis variable and Count as the statistic. One bar is positioned at Male and one at Female. The height of the bars (the position of the bars on the y axis) represents the number (count) of cases in each category. If you were using the same statistic with a point (marker) graphic element, the points would be positioned where the top of the bars would have been.

There are two types of statistical summary functions. The distinction is important because it determines whether you need a variable in the y-axis drop zone.

• Functions that do not require a summary variable. These are functions that do not require a y-axis variable. All count and percentage statistics are in this category. These statistics are available when there is no variable in the y-axis drop zone. If you want to use one of these statistics and there is a variable in the y-axis drop zone, you first need to remove the variable from the drop zone.
• Functions that do require a summary variable. These are functions that do require a y-axis variable. For example, the Mean function requires a variable on which the mean is calculated. There is also the special Value function, which is used to display the variable's raw value without any summary. These statistics are available when there is a variable in the y-axis drop zone.

If there is a question mark (?) next to statistic's name in the Element Properties tab, you need to specify a parameter by clicking the Set Parameters button. This parameter controls how the summary function is calculated. Not all statistics use parameters.

Summary functions that do not require a summary variable

Histogram. The number of cases in bins (groups) of x-axis variable values. Histogram is available only when the x-axis variable is a scale variable. The parameters specify the size of the bins (using a width or overall number) and an anchor for one of the bins. See the topic Histogram Parameters for more information.

Histogram Percent. The percentage of cases in bins (groups) of x-axis variable values. Histogram Percent is available only when the x-axis variable is a scale variable. The parameters specify the size of the bins (using a width or overall number), an anchor for one of the bins, and the base (denominator) used to calculate the percentage. See the topic Histogram Parameters for more information.

Count. The number of cases with particular x-axis and z-axis (if available) values.

Cumulative Count. The cumulative number of cases with particular x-axis and z-axis (if available) values. Each graphic element shows the number of cases with particular x-axis and z-axis (if available) values plus the total number of cases for every graphic element drawn previously on the same axis in a 2-D chart and drawn previously on both axes in a 3-D chart.

Percentage. The percentage of cases with particular x-axis and z-axis (if available) values. The parameters specify the base (denominator) used to calculate the percentage. See the topic Percentage Base for more information.

Cumulative Percentage. The cumulative percentage of cases with particular x-axis and z-axis (if available) values. Each graphic element shows the percentage of cases with particular x-axis and z-axis (if available) values plus the total percentage of cases for every graphic element drawn previously on the same axis in a 2-D chart and drawn previously on both axes in a 3-D chart. You cannot change the base (denominator) used to calculate the cumulative percentage. The base is always the total count of all cases across all panels.

Responses. The number of responses with particular x-axis and z-axis (if available) multiple response set values. This function is useful only when the x axis and z axis (if available) include multiple response set variables.

Cumulative Responses. The cumulative number of responses with particular x-axis and z-axis (if available) multiple response set values. This function is useful only when the x axis and z axis (if available) include multiple response set variables. Each graphic element shows the number of responses with particular x-axis and z-axis (if available) values plus the total number of responses for every graphic element drawn previously on the same axis in a 2-D chart and drawn previously on both axes in a 3-D chart.

Response Percentage. The percentage of responses with particular x-axis and z-axis (if available) multiple response set values. This function is useful only when the x axis and z axis (if available) include multiple response set variables. The parameters specify the base (denominator) used to calculate the percentage. See the topic Percentage Base for more information.

Cumulative Response Percentage. The cumulative percentage of responses with particular x-axis and z-axis (if available) multiple response set values. This function is useful only when the x axis and z axis (if available) include multiple response set variables. Each graphic element shows the percentage of responses with particular x-axis and z-axis (if available) values plus the total percentage of responses for every graphic element drawn previously on the same axis in a 2-D chart and drawn previously on both axes in a 3-D chart. You cannot change the base (denominator) used to calculate the cumulative response percentage. The base is always the total count of all responses.

Summary functions that do require a summary variable

Value. The value of the y-axis variable for cases with particular x-axis, y-axis, and z-axis (if available) values. You might use this statistic if your data were already aggregated, or if you are plotting every case, as in a scatterplot. If you create a bar chart that looks as if it is segmented, you probably should use a summary statistic instead. The apparent segmentation is the result of bars being drawn on top of each other because there are multiple cases with the same value.

Mean. The mean of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Median. The median of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Group Median. The grouped median of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Mode. The mode of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Minimum. The minimum of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Maximum. The maximum of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Valid N. The number of valid cases for the y-axis variable is calculated for particular x-axis and z-axis (if available) values.

Sum. The sum of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Cumulative Sum. The cumulative sum of the y-axis variable for cases with particular x-axis and z-axis (if available) values. Each graphic element shows the sum of the y-axis variable for cases with particular x-axis and z-axis (if available) values plus the total sum of the y-axis variable for every graphic element drawn previously on the same axis in a 2-D chart and drawn previously on both axes in a 3-D chart.

Standard deviation. The standard deviation of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Variance. The variance of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Percentile. The specified percentile of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

G Percentile. The specified group percentile of the y-axis variable is calculated for cases with particular x-axis and z-axis (if available) values.

Percentage Less Than. The percentage of cases for which the value of the y-axis variable is less than the number you specify is calculated for cases with particular x-axis and z-axis (if available) values.

Percentage Greater Than. The percentage of cases for which the value of the y-axis variable is greater than the number you specify is calculated for cases with particular x-axis and z-axis (if available) values.

Number Less Than. The number of cases for which the value of the y-axis variable is less than the number you specify is calculated for cases with particular x-axis and z-axis (if available) values.

Number Greater Than. The number of cases for which the value of the y-axis variable is greater than the number you specify is calculated for cases with particular x-axis and z-axis (if available) values.

Percentage in Range. The percentage of cases for which the value of the y-axis variable is greater than the minimum and less than the maximum for cases with particular x-axis and z-axis (if available) values.

Number in Range. The number of cases for which the value of the y-axis variable is greater than the minimum and less than the maximum for cases with particular x-axis and z-axis (if available) values.