Scrambler Guide

The scrambler window, all ready to go!

The basics

Here is the super-short guide on how to use the current version of the scrambler plugin…

  • Get data you want to analyze.
  • Choose scrambler from the Plugins menu in CODAP.
  • Prepare your dataset for scrambling:
    • Make a measure (a new attribute with a formula) that describes the effect you’re studying.
    • Drag it left in the table (or up in the case card) so that it’s at a higher level in the hierarchy.
  • If there is anything wrong with the way you have prepared the table, you’ll get a message with help.
  • To specify the dataset and the attribute you want to scramble, drag the attribute into the scrambler plugin. (In the picture, we’re set up to scramble Gender in the heights dataset).
  • Adjust the number and then click the buttons to create as many “scrambles” as you wish, up to 1000. (Usually, 400 is plenty.)
  • A table of measures from the scrambling appears. Analyze these to see if your original data seem unusual compared to the scrambled data.

You can try all this yourself in the live example below. The data shows some 13-year-olds, with Gender and their Height in centimeters. Make a graph of Height against Gender and put the mean height on the graph. Which gender is taller?

This example is all set up for scrambling with a measure called dMeanHeight, which has already been dragged to the left (just as you’re supposed to do :).

As you can see, the actual difference in mean heights is 5.87 cm.

Is there a significant gender difference in height? Click the 10x or 1x buttons to scramble the data and collect new values for dMeanHeight…and see what you can figure out!

(Want a task? That document is set up to compare 13-year-olds. Make it compare 10-year-olds!)

Want more information? More detail? Click a topic on the left.