Editing graphic element properties

The graphic element properties allow you to change the statistic associated with a graphic element and to modify properties specific to a graphic element type (for example, changing the shape of a bar graphic element from a rectangle to an I-beam).

  1. Click the Element Properties tab in the side bar of the Chart Builder. (If the side bar is not displayed, click the button in the upper right corner of the Chart Builder to display the side bar.)
  2. In the Edit Properties Of list, select an item that identifies a graphic element. An example would be Bar1.
  3. Choose statistic options and, if available, error bar options. Descriptions of options follow.
  4. Choose graphic element-specific options. For example, if the graphic element is a bar, you would be able to choose the shape of the bars, changing them to I-beams if necessary. Descriptions of graphic element-specific options follow.

Note: There is at least one button to the right of the Edit Properties Of list. The delete button allows you to delete a graphic element. See the topic Removing elements from the canvas for more information. The arrow buttons appear for dual y-axis charts and control the order in which the elements are drawn in the chart. Select an element and use the arrows to change the order. The last element is drawn on top of the other elements. For information about creating dual y-axis charts, see Dual y-axis Charts .

General options

Assigned Axis. For dual axis charts, you can specify the axis to which the element is assigned. For example, if you have a bar showing mean values, you might assign the bar to Y-Axis2 so that the mean values appear on the second y axis.

Statistic. Specify a statistic that determines the position of the graphic element on the y axis (for example, the height of the bars in a bar chart or the location of a point in a summary point plot). The statistic is calculated for the displayed variable. For some charts (like high-low-close and summaries of separate variables), there may be more than one variable, each of which can have different statistics. For details about the available functions and parameters, see Statistics and Parameters.

Display normal curve. Superimpose over the histogram a normal curve with the same mean and standard deviation as the data. This is available only when the statistic is Histogram or Histogram Percent.

Display error bars. For charts that summarize a categorical variable, you can choose to display error bars. Error bars are available only for Mean, Median, Count, and Percentage.

Error bars represent. For Mean, you can choose to display confidence intervals around category means, plus and minus n times the variable standard deviation, or plus and minus n times the standard error of the mean. For Median, Count, and Percentage, only confidence intervals are available.

All confidence intervals are individual intervals, with coverage levels (based on the specified alpha) applying to individual categories, not to the set of all categories in the chart simultaneously. For Count and Percentage, the intervals are based on binomial intervals for the proportions of each category of the total over all categories, appropriately rescaled to the count or percentage metric. Note that the confidence intervals for Count, Median, and Percentage are generally not symmetric around the statistic.

Bar element-specific options

Bar style. If the graphic element is a bar, you can choose to change its shape.

Point element-specific options

Stack identical values. If the graphic element is a point and a variable in one of the 2-D axis zones is categorical, you can choose to force points with the same x and y values to be stacked next to each other. Otherwise, the points would completely overlap each other, making them indistinguishable. This behavior is also known as dodging.

Linear Fit Lines. For simple scatterplots, grouped scatterplots, and scatterplot matrixes, you can include linear fit lines.

Display vertical drop lines between points. When the points are grouped and the chart is summarizing a categorical x-axis variable, you can choose to display a line that connects the points in each x-axis category.

Plot shape. For 1-D dot plots, you can specify a shape for the dot plot:

  • Asymmetric. This is the default and, in most cases, the best option. Points are stacked on the x axis.
  • Symmetric. The points are centered around the vertical midpoint of the graph (symmetric).
  • Flat. Points are drawn on the vertical midpoint of the graph without any position modification.

Y-X Pairs. For overlay scatterplots, displays the pairs of variables. A point group is drawn for each pair. You can delete a pair by selecting it and clicking the delete button. You can also use the arrow buttons to control the order in which the pairs are drawn in the chart. The pair listed last is drawn on top of the previous pairs. For information about creating an overlay scatterplot, see Scatterplots and dot plots .

Scattermatrix element-specific options

Matrix Variables. Displays the variables in the scatterplot matrix. You can delete a variable by selecting it and clicking the delete button. You can also use the arrow buttons to control the order of the variables in the matrix. The order may be useful for emphasizing varying degrees of correlation among the variables. For example, you may want to put highly correlated variables together at the top of the list.

Line, area, and difference area element-specific options

Interpolation. If the graphic element is a line or an area, you can choose interpolation options. These determine how the line or area is drawn through (interpolated through) data values.

  • Straight. Connects the data values in succession by straight lines.
  • Step. Draws a horizontal line through each data value, and connects the horizontal lines with vertical lines.
  • Jump. Like step interpolation, draws a horizontal line through each data value, but the horizontal lines are not connected.
  • Location. For step and jump interpolation, indicates the data value's position on the horizontal line.
  • Interpolate through missing values. Forces the line or area to continue through missing values. Otherwise, the line or area breaks at missing values.

Pyramid element-specific options

Note: The Scale Variable Distribution Options are available when the Distribution Variable is a scale variable. The Categorical Variable Distribution Options are available when the variable is categorical.

Display normal curve. Superimpose over each histogram a normal curve with the same mean and standard deviation as the data.

Anchor Bin. Specify the starting value of a bin. If the value is less than the lowest data value, this is the starting value of the first bin. By default, the first bin will include the lowest data value. The anchor is set so that bin boundaries are at good values. You can change the default so that the first bin includes values that are not in the dataset. For example, you might want the values of 0–5 to be included even though the lowest value in the dataset is 6.

Note: If the variable being binned is a date, the starting value must be a date literal in the date format specified for the variable on the Variable View tab in the Data Editor (for example, 01/01/2001).

Bin Sizes. Change the size of the bins. You can either specify the number of bins or the width of each bin. The width also affects the number of bins. For example, if the axis range is 0–100 and you specify the width as 5, there will be 20 bins. The greater the number of bins, the more detailed the histogram is. However, more bins may make the histogram too irregular to see the shape of the distribution.

Note: If the variable being binned is a date, the units used for the width are days. Therefore, specifying a width of 30 indicates 30 days.

Display error bars. Display error bars representing the confidence interval for the count. The specified alpha must greater than or equal to 50 and less than 100.

High-Low-Close element-specific options

Bar style. Choose a shape for the bar that shows the high and low values.

Point style for close. Choose a shape for the point that shows the close value.