Question

A manufacturer of tennis rackets would like to test their new racket. The company sponsors 10 tennis players who will be playing at Wimbledon. To test the racket, they have each player serve 30 times with their old racket. Then they have each player serve 30 times with the new racket. They compare the percentage of good serves with the two rackets. Identify the flaw(s) in this experiment and suggest some improvements.

Answer 1

This is a repeated measures design where each player is exposed to each treatment condition (type of racket). Hence, there are multiple (In this case, 2) measurements involved with each subject. This type of design have some disadvantages, one of them is the order effects. The %age of good serves could decrease due to other factors like fatigue, or increase due to experience. Hence, there are other confounding factors or order effects which come into play when we tell the players to serve 30 times with the old racket first and 30 times with the new racket first. Hence, this flaw should be taken care of in this type of design.

To reduce this problem, we can use randomization to divide the players into two groups OR allow time between the treatments.