

Comment Info: =================

General Comment:1. The 2007 NHTSA analysis of ESC effectiveness states that ESC was first available on some luxury models in 1997 while this study states (section 1.2) that ESC was first available on Mercedes and BMW vehicles in 1987. Was ESC introduced in 1987 or 1997?

2. In addition to the studies discussed in section 1.3, a number of other studies have evaluated the effectiveness of ESC. See references to those studies in the attachment.

3. There are a few typing errors in the years found in the fourth column, titled ?Years after ESC included?, of Table 1 and Table 2.

4. Some studies have compared crash rates before and after the introduction of ESC or used logistic regression instead of using contingency tables to calculate risk ratios. The reasons for not comparing the crash rates before and after the introduction of ESC were discussed in section 2.1, but there is no mention of logistic regression. The advantages and limitations of using contingency tables to compute risk ratios in comparison to logistic regression could be discussed in the report.

5. The use of contingency tables to compute risk ratios does not allow for the inclusion of confounding variables in the analysis. The inclusion of the two model years before and the two model years after the introduction of ESC only for each vehicle model allows controlling for the effect of changes in vehicles models. Possible confounding variables such as the age of the vehicle, the year of the collision and driver effects, such as age and sex, are not taken into account in the estimate of ESC effectiveness however.

6. An analysis of the causes of the increased similarity of effectiveness across cars and LTVs is mentioned in section 3.2 but no details are given. More details on that analysis would be welcome, so the reader gets a better understanding of what was done.

