This report documents a study of how CRS features, labels, and instructions contribute to CRS installation errors. Task 1A focused on assessing different physical features of 16 convertible CRS, while Task 1B evaluated baseline and modified versions of labels and instructions for two CRS models. For each phase of testing, 32 subjects were recruited based on their education level (high or low) and experience with installing CRS (none or experienced). Each subject was asked to perform four child restraint installations in a 2006 Pontiac G6 sedan. An 18-month-old CRABI anthropometric test device (ATD) weighing 25 lb was used for all installations. Each subject installed two CRS forward-facing (FF), one with LATCH and one with seatbelt, and two CRS rear-facing (RF), one with LATCH and one with the seatbelt. For Task 1A, each subject installed four of the sixteen convertible CRS. For Task 1B, each subject installed a Graco ComfortSport and Evenflo Titan twice, where each CRS had one of eight alternate instruction manuals and one of eight alternate label conditions. After each installation, the experimenter evaluated 42 factors for each installation, such as choice of belt routing path, tightness of installation, and harness snugness. Analyses used linear mixed models to identify CRS installation outcomes associated with CRS features or label/ instruction type. LATCH connector type, LATCH belt adjustor type, and the presence of belt lockoffs are associated with the tightness of the CRS installation. The type of harness shoulder height adjuster is associated with the rate of achieving a snug harness. Correct tether use is associated with the tether storage method. In general, subject assessments of the ease of use of CRS features are not highly correlated with the quality of their installation, suggesting a need for feedback with incorrect installations. The results of testing alternative labels and instructions indicate that no alternative condition significantly improved CRS installation compared to baseline conditions across all potential installation errors that were evaluated. An unintended consequence of using “combined” labels (incorporating all of the variations recommended by human factors experts that were tested individually in other conditions) is that subjects were less likely to use the CRS manual. The most promising alternative manual is a video version. Neither the labels nor manuals with improved graphics showed substantial benefit compared to baseline, even though graphics-based manuals and labels score well using the ISO and NHTSA ease-of-use rating system. The effects of varying labels and manuals on installation error are small compared to the effects of different CRS designs. The data from this study provide quantitative assessments of some CRS features that are associated with reductions in CRS installation errors. This information can be used to update the NHTSA CRS ease-of-use rating system to account for recently developed CRS features. Based on human factors recommendations, suggestions are made for assessing “clear” labels and instructions and modifying FMVSS 213 requirements to facilitate clearer labels and instructions. (Author/publisher)
Samenvatting