Hi, what analysis is appropriate for the following research?

Question

data collected from an experiment to determine whether prior information about a team’s aggressive reputation impacts on the way in which referees perceive the intent to harm when two players are challenging for the ball. A group of referees was shown a series of incidents on video from Argentinean Football League games in which opposing players challenged for the ball. All incidents were taken from a match involving a team that played predominantly in blue (blue team) and a team that played predominantly in white (white team). In a balanced design 30 participants were randomly assigned to one of 2 groups (experimental or control). The information given to both groups was identical apart from the experimental group being primed (via written and oral instructions) that the ‘blue team’ had an aggressive reputation. Following this, each participant watched the same 20 incidents, randomly presented, and responded to each incident indicating whether or not a foul had occurred. These 20 incidents comprised 10 incidents in which the blue team challenged for the ball and 10 incidents in which the ‘white team’ challenged for the ball. For each incident participants were asked to rate the aggressive intent of the player challenging for the ball on a 7 point Likert type scale where 1 = ‘no aggressive intent’, 4 = ‘some aggressive intent’ and 7 = ‘a great deal of aggressive intent’. The data were averaged to provide a single score of ‘aggressive intent’ for each condition (challenge by blue or white player). Explore whether there is a difference in the groups’ perceptions of aggressive intent across both conditions in which the players challenged for the ball.

Answer 1

Hi Hossein,

 

If I understand your research question, your Null Hypothesis is that there is no difference between the groups's perception of aggressive intent by the players in blue and the players in white, while your alternative hypothesis is that there is a difference between between the groups's perception of aggressive intent by the players in blue and the players in white. If that is indeed your question, then you need a type of test that will measure the difference in central tendency's between the scores of the different groups, blue and white. The initial temptation is to try an independent sample t-test; however the data is really ordinal and not interval data, since the we have no way of knowing whether the difference between a 1 and a 2 on the aggressive scale is the same and an increase from 6 to 7. Therefore my recommendation would be to use a non-parametric version of the t-test, which is the Mann-Whitney U Test. This is a non-parametric version of the t-test that uses ordinal variables.

 

If you need any further help, please let me know.

 

Walter