In order to permit statistical comparisons between the two samples, the data sets from the two separate samples were merged together on like questions. The sample versions (Version 1 for Safety Belt Usage and Version 2 for Child Safety Seats) were cross-tabulated with each of the survey questions that had been asked in an equivalent fashion in the two samples. A chi-square test was conducted for each of these cross-tabulations to test for the independence of the samples.
An exact test of independence was calculated to test the differences between the two samples. Pearson's chi-square is a widely used statistic to test the hypothesis that the row and column variables are independent. It is calculated by summing over all cells the squared residuals divided by the expected frequencies. The calculated chi-square is compared to the critical points of the theoretical chi-square distribution to produce an estimate of how likely (or unlikely) this calculated value is, indicating if the two variables are in fact independent. This probability is also known as the observed significance level of the test. If the probability is small (greater than 0.05), the hypothesis that the two variables are independent is rejected.
No statistically significant difference (at the .05 level) was found between the two samples on most of the vehicle characteristics (e.g., type of vehicle driven most often, airbags in vehicle, location of airbags, type of seatbelts in vehicle), driver behaviors (e.g., drive everyday, ever injured in crash), or demographic characteristics (White, Black, Hispanic, gender, marital status, interview conducted in Spanish).
There are, nonetheless, a limited set of differences large enough to be
statistically significant with samples of this size. The proportion of drivers
who always wear their shoulder belts is slightly higher in Version 1 (84.1%)
than in Version 2 (82.1%), as is the proportion that always wears lap belts
(83.2%-81.2%). Although the combined use (always) of any seatbelt is also
slightly higher in Version 1 (84.6%) compared to Version 2 (82.8%), when
the measure is limited to those who have always worn their seatbelt in the
past year the difference (70.8%-69.4%) is no longer statistically significant.
The proportion of the sample with at least some college education (57.8%-55.4%)
and incomes of $75,000 or more (25.0%-22.1%) is also slightly higher in Version
1 than in Version 2. These demographic differences in the two samples are
consistent with the slight differences in seatbelt usage in the two samples.