The use of both regression tree and mixed ANOVA techniques in the interpretation of repeated speed measures belonging to an-on road experiment

A data set composed by repeated speed measurements belonging to an on-road experiment was inquired into this study by using the regression tree and the ANOVA mixed model techniques. The on-road experiment consisted in the repetition of a driving test six times in six different days for a sample of 20 drivers on a rural road in the province of Bari, Italy. In order to explain the changes in the response variable speed, the drivers’ behavioral differences, the road geometry (in terms of sight distance) and the time (in terms of the six days of testing) were used as variables. All the three chosen variables are responsible for improvements in the tree growing and they have significant effects on speed according to the mixed oneway ANOVA. Some considerations about the results of the two analyses were given in this paper, together with an analysis of the interactions between the variables based on them. In more detail, some interesting differences between the behavior of aggressive drivers and the behavior of prudent drivers were highlighted.