# Repeated measures pre tests and ppst tests

I'm currently working on the effectiveness of a card game I developed. Lets say that I let my students play this card game for 5 consecutive days then in that 5 days period I provide them pre tests and post tests to know if the card game has an impact on their learning with the topic we are currently studying. What statistical treatment will I use to measure if there is a significant difference between those pre tests and post tests? I'm familiar with the classical paired t-test, but it only include 1 pair pre test and post test ... should I get the mean of all pre tests and mean of all post tests then use the results to perform paired t test?

• One possibility is to use all five days. Block design: Students are blocks, Days 1-5 are main effect. Similar to this Q & A with Judge = Student, Wine = Day. // ANOVA (or Friedman) depending on whether (or not) you believe test scores are normal. // Could also use t test as you suggest, but that may lose some information. The generalization of 'pair' is 'block'. – BruceET Sep 19 '18 at 18:34
• Thanks for the reply sir, but Im worrying that each day has pre test and pos test results... somehow its not paralle or maybe I cannot see the connection to the one you propose... – rosa Sep 19 '18 at 20:12
• Or maybe sir are you trying to tell me that what I am going to put inside of each main effect are the differences of post test and pre test?please enlighten me – rosa Sep 19 '18 at 20:43
• That is new information. What did you hope to discover by doing pre and post-test each day? What else did you hope to discover by testing on five different days? // How about editing full disclosure of data and objectives into your question? That way others won't have to read comments to figure out what you're asking. – BruceET Sep 19 '18 at 21:25
• Pre and post tests of what exactly? – Carl Sep 20 '18 at 1:27