15May/102
What is the best statistical test to determine whether a computer program really helped my students improve?
Students were assigned randomly to an experimental group to work with a reading comprehension tutorial program. At the end of the unit, all students were tested, but I need to see if the test scores are significantly higher or if it's just random sampling error. I haven't taken a statistics course in years... What is the best test of significance for the mean scores?
May 15th, 2010 - 14:46
A t-test might do what you want. Would be better to consult a applied statistician.
May 15th, 2010 - 15:41
maybe you can use analysis of variance. i’m not a statistician, though.