Repeated Measures To-Do Activity

Goals:

  • Review the options for dealing with pretest scores in a repeated measures study design

Activity Guide:

  1. Let’s do a similar activity as we did in Module 11 to compute a t-test comparing the Wise Use of a classroom technology with standard best practices. But this time, instead of comparing 2 different classes with entirely different students, this time let’s use just one class, and give them a pretest before they start the technology-enhanced unit. In fact, let’s use this gradebook that shows (4 Homework assignments and) the Unit #1 test . It also has a column that shows their pretest scores before the Unit started. Here is that Excel file (xls) with the pretest column added (there is also a “gain score” column (Column I) added, which is just each Unit test score minus the pretest score).
  2. For those of you choosing not to use Excel, here are the test scores:
    1. Pretest scores:70 77 71 68 74 70 63 71 72 73 72 68 72 75 75 77 74 72 72 78 76 74 77 76 80
    2. Unit test scores:92 98 97 87 59 72 91 99 88 94 94 87 94 89 95 92 94 93 98 97 92 88 97 88 99
  3. Again, let’s just look at the scores to see what we notice. All 25 of our students took both tests, so we don’t need to worry about anyone who missed either or both of the tests (missing data). Remember little Jon bombed the Unit test, but now we see on the pretest, he actually did better! Something really must be going on with Jon on that Unit test, right? We need to pull him aside and inquire about what’s happening, if we haven’t already done that. The average (mean) score on the pretest was 73.1, while remember the average on the Unit test was 91.0.
  4. As you did in Module 11, let’s use the free online t-test tool to look for a statistical difference between the pretest performance and the grades on the final Unit 1 test. This time we’ll use the paired t-test option, also known as dependent samples.
    1. Copy/Paste the pretest scores into one of the Enter Data for Group # windows and paste the Unit test scores into the other. Again it does not matter which is which, but typically the 2nd set will be subtracted from the first set, so I’d suggest putting the pretest scores in as “Enter Data for Group 2.” That would make the Unit test scores Group 1. It really does not matter whether the t statistic is positive or negative here.
    2. Click the blue “Find t and p values” button and scroll to the bottom. WOW! These guys really improved from pretest to post test.
    3. Let’s discuss this. Do you think it was the wise integration of technology that really made the difference? Could it have been something else? What else? Is there anything else in that report of interest, such as the Variance or the standard deviations that could tell us anything? Just sayin’. Let’s discuss it.

 

Back to Challenge Cycle