Task 2: Analysis and Recommendations
Under Task 1 of contract No. QA96001001, an annotated bibliography concerned with the anthropometry of people with disabilities, and its applications to the design of facilities, workspaces, and equipment, was completed. It appears in this report as an appendix. The objective of Task 2 is to assimilate the information gathered in Task l, to identify further anthropometric research needed to update guidelines and standards for accessible design, and to recommend the means of carrying out such studies.
The bibliography compiled in Task 1, while by no means exhaustive, incorporates a large body of anthropometric data on more than 11,000 persons of every age and a wide variety of disabilities. Unfortunately, most of the studies were conducted on specialized populations, many of them foreign. Dimension definitions and measurement techniques vary from study to study and, in many cases, samples were very small. In a recently published review of the anthropometry of people with disabilities (Kumar, 1997), A. Goswami examined six international studies of people with lower limb disorders and discovered that, for a combined total of 58 body size descriptors measured in the studies, not a single dimension was found in common. Goswami also could not find a single study that attempted to standardize either body landmarking or measurement procedures. These and similar findings are illustrative of the current state of affairs in regard to anthropometry of this group of individuals. Thus, while there is a great deal of existing anthropometric data, any attempt to combine them into a useful database would be futile.
Examination of the literature further reveals virtual unanimity among experts in the field regarding the undesirability of applying data from non-disabled populations to the design of equipment and spaces intended to accommodate populations with the full range of abilities and disabilities. This would be true of data from any non-disabled population, but is exacerbated in the U.S. by the fact that most existing anthropometric data on U.S. adults comes from military personnel. So poor is the status of applied anthropometry on U.S. civilians that the last major survey containing significant data applicable to design was completed in 1962 (Stoudt et al., 1965). This nationwide stratified random sample of men and women measured 14 anthropometric dimensions that can be used for the design of workspaces. Since that time three other large civilian surveys have been conducted but none contain anthropometric dimensions useful in design. As a result, many texts and guidebooks intended for interior and product designers in the U.S. are based on body size information collected from highly fit military populations.
The most recent of such comprehensive studies was conducted on U.S. Army personnel (Gordon et al., 1989). Although over 200 dimensions were measured on a group of 9,000 ethnically diverse soldiers, the data from this survey lack the range of variability found in the population of interest here. To examine this contention, a comparison was made for workspace dimensions from the Army survey and a compilation of seven separate studies of people with disabilities. Using the coefficient of variation (CV) as the statistic to compare the degree of variability, Table 1 presents the differences for 11 variables. The CV is a dimensionless statistic expressed in percent, so comparisons across dimensions that vary in magnitude are still valid.
Coefficients of Variation (CV) for Selected Anthropometric Dimensions: Several Samples of Persons with Disabilities and U.S. Army Males
|DIMEN-SION||PERSONS WITH DISABIL-ITIES1||ARMY||DIFFERENCE|
|Min and Max CV||No. of Studies||Average CV|
|Shoulder Height, Sitting||6.9‒11.0||4||9.3||5.0||4.3|
1. Disabled data computed from means and standard deviations given by Goswami (Kumar,1997).
2. Sitting height value used.
3. Biacromial Breadth value used
4. Forearm-Forearm Breadth value used.
Except for weight, the group with disabilities shows as much as twice the variability of the non-disabled sample in some cases. One reason for this result is, of course, the great number of disabilities, which, in turn, can cause a wide variety of changes in body size, posture, and function. This has led many investigators to a second finding relevant to Task 2: Anthropometric data obtained from individuals with a particular disability should not be used to draw up designs and standards for individuals with different disabilities. Nor are they applicable to a general U.S. population of people with disabilities.
The principal way to achieve good design is through the application of anthropometric data. In order to be effective, however, the data must not only be appropriate to the design at hand but must also be descriptive of the target user population. As noted above, much if not all the anthropometry so far collected on groups with disabilities involves specialized populations (Damon and Stoudt, 1963; Goswami et al., 1987; Molenbroek, 1987), and therefore has limited application for federal agencies that must concern themselves with the general U.S. population of individuals with a wide variety of disabilities.