Prediction of drivers' recognition of posted signs in five Arab countries.

The effects of six predictor variables, drivers' age, monthly income, years of education, sex, marital status, and nationality on perception of traffic signs were modeled through regression techniques. The populations sampled here were from Kuwait, United Arab Emirates, Bahrain, Qatar, and Oman. A total of 28 posted regulatory and warning signs were investigated. A questionnaire especially prepared for this study was distributed to 9,000 drivers in the five states, out of which 4,774 responded (53%). The drivers were selected according to stratified choice-based random sampling technique. Two prediction models were developed for assessment of drivers' recognition of regulatory and warning signs.