A Multivariate Test-Retest Approach for Verifying Self-Reported Speeding Behaviours among Taxi Drivers in Tehran

The aim of this study is to assess the repeated responses and verification of surveys using a one-way multivariate test-retest process with a four-week interval between measures for ۱۲۰ taxi drivers in Tehran. The only independent variable includes the questioning stage. The items of a scenario-based questionnaire were considered as dependent variables. The null hypothesis posits no significant difference. The results were addressed using two testing approaches: variance testing and average testing. The results of the multivariate test statistics showed no significant variance difference in the answers between the first and the second test. The average testing including the contrast test also showed such a conclusion for the average measures in two stages. This means that factors which motivate speeding amongst taxi drivers were stable over a four-week period. In all these cases, the p-values were greater than ۰.۰۵, which indicates the non-significance of the difference between the responses in two stages, and as a result, accepting the null hypothesis. The added power in multivariate analysis results from the correlation between outcome variables and the use of cross-products.

کلیدواژه ها:

Reliability ، Manova ، Survey validation ، Self-reporting ، Theory of planned behaviour (TPB) ، Driver Behaviour Questionnaire (DBQ)