A cross-sectional study considers one or several similar sites with and without a treatment applied to them and analytically compares the performance measure(s) between them. This method can be used when there is not many instances of locations where a specific treatment is applied and the data is available over a descent time period. If the data is available a before-after study would be preferred.
It should be noted that in a cross-sectional study, the effects of a treatment found by comparing the crashes in sites with and without the treatment may have causes other than the treatment and that is why care must be taken in choosing the sites to be similar in their functions, geometry and traffic characteristics.
[To be continued]