Anthropometry for Persons with Disabilities

Measuring Techniques

The Body Size Descriptors

These measurements serve as basic population descriptors and are applied in the design of workspaces and the physical environment, as well as the sizing of personal items and equipment. Except for body weight, this group of measurements is made up of simple point-to-point distances in one or another of the principal body axes and some geometrically more complex circumferences and surface contours; they are typically obtained manually using an anthropometer, measuring tape, and a variety of special calipers. Modern technology currently provides alternate ways to obtain accurate and reliable data of this type. Among them are the Faro Arm (Faro Technologies, Inc.) which is a portable coordinate measuring system. It consists of a probe on the end of a 6 degree-of-freedom arm, which is linked to a laptop computer. The user touches the probe on a body landmark, presses a button, and the location of the point in three-dimensional space is recorded automatically in the computer software. Software later allows the calculation of point-to-point distances and other dimensions. Such a device might be useful here because some subjects in the proposed study may not be able to assume the rigid standardized postures often used in traditional anthropometry, and the Faro Arm probe might be able to access some critical body areas difficult to reach in seated subjects.

Reach

For reach and field-of-vision measurements, Air Force methodology is, once more, instructive. Since 1990, investigators at Wright-Patterson Air Force Base have been engaged in testing accommodation of aviators seated in the cockpits of a variety of aircraft. (Kennedy and Zehner, unpublished) In many ways, the problems presented by this project are similar to those faced by the Access Board and by designers of workspaces intended to accommodate wheelchair users. Among the seven major areas of accommodation this long-running AF project is specifically concerned with is "hand reach to, and actuation of, controls."

The functional reach dimensions listed above can all be measured in the traditional way by keeping the back, shoulder, and buttocks against the back of the seat and stretching the arm along a scaled wall chart (to the thumbtip, to the forefinger resting on the pad of the thumb, or the tip of the middle finger). Alternatively, the Faro Arm might be touched to the wall or reference plane (possibly the back of the chair), and then touched to the tip of the finger.

Air Force investigators take arm reach measurements one step farther, in that they measure arm reach in three "zones." Reach Zone 1 requires that the operator's shoulders be fully restrained by harnesses with the pilot held against the seat back by the inertial reel. Zone 2 requires use of the harness, but the operator is free to move his/her shoulders and torso forward and to the sides to a comfortable limit permitted by the total restraint system. Zone 3 specifies that the inertia reel be unlocked and the shoulders and torso permitted to move forward and to the sides as necessary for maximum reaches. Though it is not altogether clear that these kinds of distinctions should be made in conducting reach measurements on people using wheelchairs, there is certainly the possibility that the principle will be relevant.

Field of Vision

This is another area critical to cockpit accommodation. Air Force anthropologists measure maximum upward and downward lines of sight, forward and to the sides using a carpenter's inclinometer fitted with a sight tube to measure visual angle. The sight tube is equipped with cross hairs at each end. An Abney Level can also be used.

Strength

The ability of wheelchair users to operate equipment in work and living spaces depends not only on reach but also on sufficient hand strength to grasp and manipulate controls. The design and placement of grab bars are also guided by strength capabilities, chiefly in the hands and arms. Strength can be measured in a number of ways with pushing, pulling and twisting perhaps the most relevant to the present case. ADAAG standards currently specify that door opening and operation of assorted other control mechanism require no more than 5 pounds of pushing or pulling force, for example. Strain gauges that can be instrumented for direct computer readout, are probably the means of choice for taking these measurements.

Wheelchair/User Measurements

Accommodation and accessibility standards for individuals using mobility aids are worse than useless unless they take into account the wheelchair and its user as a single unit. Measurements from and to the most protruding points, whether they be located on the chair or on the user, are not difficult to make using either traditional manual instruments or a Faro Arm. The difficulty arises in the multiplicity of chairs and scooters on the market today. Investigators undertaking to make such measurements would have to do some research to determine at least the largest of such mobility aids and/or those with the highest seats in order to obtain results useful in creating guidelines for accessibility. One source of such information is a 1995 study (KRW Inc.) conducted for the U.S. Architectural and Transportation Barriers Compliance Board which incorporates a listing of more than 125 models of scooters and power chairs along with their lengths, widths, wheel base lengths, and seat heights.

From a sampling point of view, these chairs are very challenging. Ordinarily, one designs to accommodate a certain percentage of the population, or designs to a specific value (95th percentile forward arm reach, for example). This point is determined not only by the total range of variability, but by the frequency. Thus the relative number of certain chair types is very important. For example if a very large chair were infrequently purchased and used, it would have little effect on the value of the 95th percentile. However, if a very large chair were purchased often, then it would have the effect of raising the 95th percentile, and in turn, changing the design target. Thus it is not enough to know that the range in chair height is 30 to 39 inches. We would need to know the effective numbers of the chairs at various heights in the population of wheelchair users.

This need not be especially complicated, particularly for the pilot study. By measuring people in their chairs, one would automatically get a random sample of the chairs that people buy, in the approximate frequency in which they are seen in the population. In creating the sampling plan and subject acquisition plan, one would exercise caution to make sure that no bias in chair type is introduced. An example of one such bias might be conducting a pilot test in a geographic area where a certain type of chair is more readily available. If it develops in the pilot study that chair variability cannot be accommodated in this way, then the follow-up full study would have to include chair type as a parameter in the sampling plan. This would introduce complexity, however, and should be avoided if at all possible.

Measurer's Handbook

Crucial to achieving the second goal of the proposed survey is the creation of a measurer's handbook that would serve to ensure that future studies produce data that could be used to expand the original database. Such a handbook should include clearly worded definitions of the dimensions measured, detailed descriptions of the methods used to measure them, landmark descriptions would also be included, and illustrations to enhance the measurement descriptions. An example page is shown in Figure 1.

Data Entry

We have found through experience that an online data entry and editing system dramatically reduces the amount of observer error present in the final data set. The system we use was developed for the ANSUR survey and has been used extensively since that time. In this software, measured values are entered into a laptop computer, and are checked for reasonableness as they are entered. A suspicious value is flagged, and can be remeasured while the subject is still available. In this way, many types of measurement error can be reduced. The process is documented in detail in Churchill et al., (1988). We recommend such a system for any data collection effort for a population of wheelchair users.

Establish Allowable Error

Because anthropometric data are used in the design of workspaces, and equipment, excessive error in the data can result in badly designed workspaces and unsuitable products. Observer error is a fact of life in almost any scientific endeavor. Though it cannot be eliminated entirely, it can be considerably reduced.

Error analysis of anthropometric data is usually done after the data collection has been completed. While this gives the user of the data the information necessary to judge the effects of error on his/her use of the data, it does not allow observer error information to be used during data collection to improve the quality of data collection. The approach used in the Army's 1987‒1988 anthropometric survey and the one recommended here was to establish an allowable observer error for each dimension.

Standards for allowable error are established by a team of expert anthropologists conducting repeat measurements of the selected dimensions, and analyzing the inter- and intraobserver differences. Error allowances will differ: larger ones will be established for functional reach measurements, for example, than for breadths which tend, on the whole, to be more easily repeatable.

Allowable errors are used for two purposes. They are first used during the initial training period as an indicator that measurers have successfully learned their tasks. Team members make practice measurements on a group of subjects to learn their assigned dimensions. Intraobserver and interobserver error results are calculated regularly to assess the ability of each measurer to repeat measurements within fixed limitations, and the ability of each pair of measurers to achieve interobserver consistency.