Investigation of Special Subgroups

The construction of injury probability models can include the whole data set, as presented above, or can be performed using subgroups of the data set. In view of biomechanical differences [19-21], a splitting of the population into subgroups depending on age (e.g., 4-17, 18-64, 65-h) could help detect possible injury risk factors specific to particular age groups and improve the model quality. 

Upper bumper ref. line— vert. UBRL oidas 4.0 0.001 2.886 1.630 5.112 

Lower bumper reference line—vert. LBRL ,GIDAS 9.2 0.024 0.433 0.209 0.896 

As a result, the further splitting of the whole data set seems to reduce power and produce selection effects, which distort the results and make a sound interpretation nearly impossible. The statistical problems become also evident in very large confidence intervals. The method elaborated in the preceding sections can therefore not be recommended for small empirical data sets. 

The corresponding results for PCDS are given in Table 5.28. As many of the groups presented have low case numbers, similar effects as presented above for the GIDAS data become evident for the PCDS data set. The large confidence intervals are an indication for the instability of the models. A further interpretation of the results does not seem feasible and is therefore not given. 

---