Frequency Distribution and Dialog Box
SPSS: Grouped Frequency Distribution
FIRST STEP: Under the Transform menu, choose Visual Binning… This command assists you in creating a new variable that groups the data. You will then use the new variable to create a grouped frequency distribution.
• From the Variables list box, click on wt (weight) and then on the arrow to move it to the Variable to Band list box. Click Continue.
• Select wt in the left box. Near the top of this dialog box, enter a name for your new variable (such as wt_groups) in the “Binned Variable” box
(cannot have any spaces in the name).
• Near the lower right, click Make Cutpoints…
• We are going to make Equal Width Intervals, which is the default selection in this dialog box. You have to fill in 2 of the 3 fields; for our purposes, fill in “Number of Cutpoints” and “Width”.
• As discussed above, generally 10 to 15 intervals works well. The Number of Cutpoints = [number of intervals – 1]. (Why? * ) Thus, if we want 10 intervals, we’ll enter 9 in the “Number of Cutpoints” box.
• For the “Width” of each interval: (a) find the difference between the lowest and highest score in your data (you can see these values in the background dialog box behind the active dialog box); (b) divide the difference by the number of intervals (in this example, [122.7 – 65] / 10 =
5.77); and (c) round up to the whole number (6.0). Enter that number as the interval Width.
*
Because one cutpoint makes 2 categories. For example, if you have one apple and you "cut" it in half (one cutpoint), you have 2 apple halves. If you cut one of those halves (so now you have a total of two cuts to the apple), you'll have 3 pieces of apple. Make another cut (3 total cuts to the apple) and you'll have 4 pieces of apple, and so on.
So, you will always have one more category than you have cutpoints.
• Click in the First Cutpoint Location box, which should fill in automatically.
In this example, the value 65.00 automatically appears.
FIRST STEP: Under the Transform menu, choose Visual Binning… This command assists you in creating a new variable that groups the data. You will then use the new variable to create a grouped frequency distribution.
• From the Variables list box, click on wt (weight) and then on the arrow to move it to the Variable to Band list box. Click Continue.
• Select wt in the left box. Near the top of this dialog box, enter a name for your new variable (such as wt_groups) in the “Binned Variable” box
(cannot have any spaces in the name).
• Near the lower right, click Make Cutpoints…
• We are going to make Equal Width Intervals, which is the default selection in this dialog box. You have to fill in 2 of the 3 fields; for our purposes, fill in “Number of Cutpoints” and “Width”.
• As discussed above, generally 10 to 15 intervals works well. The Number of Cutpoints = [number of intervals – 1]. (Why? * ) Thus, if we want 10 intervals, we’ll enter 9 in the “Number of Cutpoints” box.
• For the “Width” of each interval: (a) find the difference between the lowest and highest score in your data (you can see these values in the background dialog box behind the active dialog box); (b) divide the difference by the number of intervals (in this example, [122.7 – 65] / 10 =
5.77); and (c) round up to the whole number (6.0). Enter that number as the interval Width.
*
Because one cutpoint makes 2 categories. For example, if you have one apple and you "cut" it in half (one cutpoint), you have 2 apple halves. If you cut one of those halves (so now you have a total of two cuts to the apple), you'll have 3 pieces of apple. Make another cut (3 total cuts to the apple) and you'll have 4 pieces of apple, and so on.
So, you will always have one more category than you have cutpoints.
• Click in the First Cutpoint Location box, which should fill in automatically.
In this example, the value 65.00 automatically appears.