Center for Inclusive Design and Environmental Access RERC Logo

The Anthropometrics of Disability

Prepared for the U.S. Access Board

caption next to image says, Design professionals, consumer advocates and government officials concerned with accessible and universal design are seeking more reliable data on anthropometrics in a form that they can use more effectively. Actual image shows a researcher taking measurements using a Farro Arm.

Rehabilitation Engineering Research Center on Universal Design

School of Architecture and Planning

University at Buffalo

The State University of New York

Buffalo, NY 14214-3087

The Anthropometrics of Disability: An International Workshop

Version 1.0

REPORT PREPARED FOR: U.S. ACCESS BOARD, WASHINGTON, DC

Authors:

Edward Steinfeld

James Lenker

VictorPaquet

February 7, 2002

This report is available on the World Wide Web:

http://www.ap.buffalo.edu/idea

TABLE OF CONTENTS

The Anthropometrics of Disability: An International Workshop ............. 1

Preface .................................................................................... 8

Disclaimer ................................................................................ 9

Introduction .............................................................................. 10

Background ........................................................................... 10

Understanding Anthropometry ................................................ 12

Summary of Papers ................................................................. 14

Overviews ........................................................................... 15

Structural Anthropometry & Reach .......................................... 16

Functional Anthropometry ...................................................... 17

Sampling ............................................................................ 18

Integrating Research with Design ............................................ 20

New Research Directions ....................................................... 21

PREFACE

RERC on Universal Design

Anthropometry of people with disabilities is receiving an increasing amount of attention. This Workshop brought together about 40 invited experts from many fields to share information and ideas and to discuss the state of the art in this cross-disciplinary area of knowledge. The goal was to generate new ideas and approaches about data collection, analysis, computer modeling and effective use of anthropometric data in the design of environments and products. An important long-range objective was to identify ways to improve the knowledge bases used in accessibility codes so that they can more accurately reflect the functional abilities and needs of people with disabilities.

Proceedings were prepared for use in the Workshop. This document is intended to summarize both the papers and the discussion that occurred among those present. More formal publications of the full papers and versions of this document are planned.

We wish to personally thank our Co-Organizer, the RERC on Ergonomic Solutions for Employment, our staff who helped to organize the Workshop, and all the Participants who contributed papers and participated in the discussions. In particular, thank you to Lois Thibault of the U.S. Access Board and William Peterson of NIDRR for identifying the need for this event and providing the financial support necessary to make it happen.

Edward Steinfeld Signature

Edward Steinfeld, Arch. D.
Director
RERC on Universal Design at Buffalo

DISCLAIMER

This report was developed with funding from grant H133E99005 from the U.S. Department of Education. The contents do not necessarily represent the policy of the Department of Education and readers should not assume any endorsement by the Federal government.

National Institute for Disability and Rehabilitation Research logo

INTRODUCTION

BACKGROUND

“The Anthropometrics of Disability: An International Workshop” was a three-day conference held in Buffalo, NY between May 31 and June 2, 2001. It was attended by about 40 invited participants and featured 20 presentations by researchers and practitioners from around the world who are experts in the areas field of anthropometry and anthropometric issues related to people with disabilities.

The broad underlying purpose for this workshop was captured in the forward to the conference proceedings:

Design professionals, consumer advocates and government officials concerned with accessible design and universal design are seeking more reliable data on anthropometrics in a form that they can use more effectively. Rehabilitation practitioners are looking for new approaches to measuring people with disabilities in order to fit them with seating systems, wheelchairs and other assistive devices. In the wheeled mobility community, researchers, manufacturers and designers are working on international standards for wheelchairs and other related products. Anthropometric data is a key piece of information for that work. In the field of ergonomics, researchers and practitioners need better data to implement interventions in the workplace, vehicles and other settings.

The format of the workshop was a series of paper presentation and discussion sessions. Papers were prepared prior to the Workshop and distributed to participants in printed form. Presentations were timed to insure that about 50% of the time available would be devoted to discussion. All participants submitted written recommendations based on the discussion at each paper session. A final summary discussion was also held to identify key priorities and develop an agenda for future work.

This report is designed to summarize the papers, discussion and recommendations and to communicate it to a diverse audience. To make this report more accessible to the lay reader, a brief overview of contemporary anthropometry is provided. The first section in the main body of the report includes short summaries of the papers and the key points presented in every paper session. Next, we summarize the discussion and recommendations made by participants to advance the field. The last section, the Executive Summary, provides an overview of the entire workshop, presents a proposed agenda for action to follow through with many of the recommendations and identifies specific implications for the two sponsors, the U.S. Access Board and the National Institute of Disability and Rehabilitation Research.

The entire report was designed to provide information as concisely as possible, given the complexity of the topic and variety of viewpoints represented. For those not interested in the details and technical issues, we recommend reading the remainder of the Introduction and then skipping to the Executive Summary. A volume of proceedings has also been produced for those who want to obtain the full text of the papers presented. This document is available for download on the website of the RERC on Universal Design at Buffalo. (http://www.ap.buffalo.edu/rercud/).

The Workshop was underwritten by the US Access Board, with generous support from the National Institute on Disability and Rehabilitation Research through the Rehabilitation Engineering Research Center on Universal Design at Buffalo and the Rehabilitation Engineering Research Center on Workplace Ergonomics. We thank all the participants for the effort they put into their papers, their enthusiasm and interest at the Workshop and their interest in expanding this effort with other activities in the future.

UNDERSTANDING ANTHROPOMETRY

Anthropometry is the measurement and analysis of body characteristics, including stature, sizes of body parts and the space in which the body functions, e.g. reach limits and clearances for movement. Anthropometry is important for a variety of professional disciplines. It is used extensively in product design, particularly when usability and safety require a close fit between people and their environment. For example, detailed anthropometric studies should be completed in automobile design to make sure that people with a range of statures can have an adequate field of view through windows and mirrors, reach the pedals and hand controls, fit comfortably in the seats and not be exposed to avoidable safety risks in collisions. In architectural and interior design, anthropometry is the basis for many code requirements including those related to accessibility and fire safety. For example, the width of wheelchairs and their occupants is a key determinant of minimum clearances and spaces for reserved seating. In the rehabilitation fields, anthropometry is very important for design of assistive technology. It comes into play in design of mass produced AT as well as in the custom fabrication of adaptive equipment. Because of the large variability in body and ability characteristics of many people with disabilities, the anthropometry of disability poses some major challenges and the need for departures from conventional practices.

Anthropometry is, at first glance, a relatively simple scientific endeavor. However, when it is better understood, like all science, it has many complexities. For example, one of the most difficult problems is identifying where to take measurements of body parts. The first step in this process is defining landmarks on the human body that can be reliably identified by data collection personnel. Variations in selection of landmarks can make a significant difference in results between studies, which challenge those who must interpret the information when making design decisions or decisions about design standards for products or environments. What bony protuberance does one use to mark the location of the shoulder joint? How do you estimate the position of internal points like the hip socket from measurements made on the surface of the body? How do you measure soft body tissue that deforms when pressure is applied to it like the posterior? These are not easy problems to solve.

There are a great many other issues that add to the complexity of anthropometry and influence the validity and accuracy of this work. Some include:

· What measurements should be taken

· What are the most accurate and efficient measurement methods

· Posture to be used when taking measurements

· Who to select for measurement when everyone in a target population cannot be measured

· How many people are needed to represent the target population

· How to recruit participants including how much to pay them and how to get them to the research site

· How to measure sensitive parts of the body

· What type(s) of clothing should be worn during measurement

Although different terms are used, there are two main types of anthropometry:

Structural anthropometry (often called “static anthropometry”) – measurement of the body at rest including overall measurements like total stature and weight, measurements of links or circumferences like wrist to elbow, knee to hip, circumference of head and measurement of specific landmarks in reference to some other point, like the floor, e.g. eye height. Static anthropometry includes the measurement of assistive devices like canes and wheelchairs either alone or in relationship to the body.

Functional anthropometry (often called “dynamic anthropometry”) – measurement of the body in motion like the reach envelope of seated work, the movement of body parts in relationship to one another or the space required to turn a wheelchair. Functional anthropometry includes the measurement of assistive devices or other objects used by people as they move. It also includes measures of strength (e.g., grip or pull strength) that depend on the characteristics of a task (e.g., direction in which something is pulled or length of time that force must be held).

Today there is an increasing emphasis on functional anthropometry because it has been established that structural measurements alone cannot fully predict human performance in real world settings where the body is usually in motion or under stress of some sort. This has added another level of complexity to anthropometry because free, or unloaded, movement is not always sufficient to capture the nature of performance in real world tasks. In a great number of tasks, we lift weight and apply force to objects and world a round us, sometimes in complex ways. Even in free movement like walking or seemingly static conditions like standing or sitting for long periods of time, understanding the forces on joints and tissue may provide new information leading to improved safety, increased comfort and increased independence. Human movement analysis, or kinematics, and the study of forces on the body, or kinetics, need to be integrated with anthropometry to obtain a full understanding of human performance in functional tasks.

SUMMARY OF PAPERS

The 20 papers addressed topics in 7 related areas:

· Overview and Summary of Needs – the big picture of
anthropometry and its relationship to design

· Structural Anthropometry and Reach Capabilities – measurements of body size and body parts including consideration of clothing and assistive technology used by an individual (e.g. wheelchairs)

· Functional Anthropometry – measurement of the body in motion or strength as used in performing everyday activities

· Human Modeling Based on Anthropometric Data – computer modeling of the human body and its utilization of anthropometric databases

· Sampling – methods and approaches for selecting people to participate in anthropometric surveys

· Design approaches – methods for utilization of anthropometric data in design

· New Directions and Alternative Methods – emerging research methods that could provide useful information for improving our knowledge of anthropometry and its relationship to life with a disability

The papers were presented in groups of 2 -4, with each group relating to one of the 7 topical areas. The presentations were immediately followed by discussion. The summaries below describe the content of the papers and the important points made by the authors.

1. OVERVIEWS

Four papers provided an overview of the major issues that need to be addressed in the field of anthropometry and disability. These papers demonstrated the complex nature of anthropometry and the gap that exists between the need for knowledge and accurate reliable databases.

Lois Thibault provided the perspective of a government administrator who needs anthropometric data to establish policies and standards for design of the built environment. John Roebuck summarized needs for development of accurate and comprehensive static anthropometric measurement methods as well as computer-based models. He identified many of the unique difficulties of this work with populations that have disabilities. J. Mark Porter described approaches to using human modeling in design and suggested that such models be used to create models of individuals rather than using aggregated summaries of isolated anthropometric variables that are more difficult for designers to use. Doug Hobson described efforts to develop standardized postural descriptions for wheelchair users and outlined the benefits of standardization in measurement methods.

Summary of key points:

· Available databases are not providing the information that policy makers and designers need in a form that is easy to use.

· Conventional anthropometry is not adequate for improving our knowledge of the anthropometrics of disability.

· Computer human modeling offers great promise in helping to improve our knowledge as well as providing information in a form that designers and policy makers can utilize more readily.

· Task-based models of individual people with disabilities performing specific tasks could be a very useful resource in the absence of comprehensive human models based on data collected from large samples.

· Standardization in measurement approaches can provide significant benefits by providing a common ground for policy makers, researchers, rehabilitation practitioners, computer modelers, code developers and designers.

· Contemporary issues in the regulation of accessibility emphasize cognitive issues, aging and usability of information technology.

5. SAMPLING

Two papers were presented on sampling for anthropometric studies of people with disabilities.

Some background is important in understanding these issues. Historically, anthropometric data has mostly come from military samples yet research shows that there are real differences compared to the civilian population. Thus sampling from the broader population is critical to developing databases that are useful for the design of the everyday environment. Conventional sampling strategies select study participants using proportionate strategies, e.g. if African Americans constitute 20% of the population, then 20% of the sample will be made up of African Americans. Typically, design based on anthropometrics tries to accommodate 95% of the target population but most people with severe disabilities fall outside that level when proportionate sampling schemes of the total population are used. This means that disability will be very underrepresented in conventional anthropometric studies. Some studies purposely exclude people with disabilities and older people to keep the results unaffected by “outlying cases” or people who have widely divergent abilities and characteristics.

Bruce Bradtmiller reviewed the basic issues of sampling in anthropometric research. He described the key problems in sampling related to disability and outlined the related challenges of future research. James Lenker and Victor Paquet presented two alternatives to proportionate sampling strategies for people with disabilities and compared their benefits and limitations.

Summary of important points:

· Proportionate sampling does not provide the best results when the anthropometric characteristics of groups in the population are very different, e.g. since there are many more people with arthritis than people with Cerebral Palsy, data based on a proportionate sample of people with disabilities framed by cause of disability will not represent the needs of people with CP very well.

· Independent sampling of groups can help to identify and compare the characteristics of people from small groups; however, it then becomes difficult to define the characteristics of the entire target population, i.e. there will be several 95^thpercentile values.

· In anthropometry of people with disabilities, we should reevaluate the confidence level typically used in determining the size of samples (95%) because of the high cost of recruitment and time in data collection; there is no magic in the 95% target.

· For people with disabilities, using a lower confidence level may be reasonable because one can assume that many individuals will need custom design solutions.

· Sampling frames based on cause/source of disability do not reflect the functional variation within any category and the dynamic nature of disability over time.

· Sampling on type of mobility aid used would represent the variation in functional ability and be relatively easy to implement.

· Stratification on type of mobility aid used is complicated by people who use more than one device or who may use a device only part of the time or in certain limited settings.

· Sampling based on functional independence in activities of daily living could be implemented with well established scales; this would make it easier to apply findings in design for diverse populations and would be consistent with the universal design perspective–a continuum of needs based on functional ability.

· Stratification based on functional independence requires screening participants prior to selection using trained professionals.

· Defining functional independence is difficult since it is mediated by the environment in which a person lives.

SUMMARY OF DISCUSSION & RECOMMENDATIONS

The papers were meant as the point of departure for discussion of the issues surrounding this type of research and generating recommendations. Each presenter advanced specific points of view and recommendations. From there, all participants joined in to discuss the themes, facilitated by a session moderator. Discussion occurred after each group of papers was presented so that it usually addressed issues common to each set and often expanded on themes established earlier in previous sessions. At each session of the workshop, about half the time available was devoted to discussion. Two different note takers recorded the points made. In addition to the verbal discussion, participants were asked to submit written recommendations for research, practice and policy for each discussion period. They wrote their recommendations during or immediately after the discussion on the papers. About 130 individual recommendations were submitted. Clearly, participants took that task very seriously. At the end of the Workshop, a wrap up session focused on a summary and final discussion.

In order to summarize the m ain points of the discussion and the recommendations in a concise and meaningful way, we used content analysis to group them into 10 main themes:

1. 1. Increasing standardization in methods

2. 2. Increasing use of functional approaches to research

3. 3. Better organization of knowledge

4. 4. Identifying research priorities

5. 5. Developing computer simulation models

6. 6. Incorporation behavioral and social factors

7. 7. Improving sampling

8. 8. Improving cost effectiveness of research

9. 9. Improving data collection methods
10.Increasing communications and dialogue

Each one of these themes will be discussed below by summarizing the discussion related to them, reporting specific recommendations and elaborating on the recommendations with further observations.

1. INCREASING STANDARDIZATION AND CONSENSUS IN RESEARCH AND PRACTICE

Much discussion focused on the limitations of previous research. In older studies, the methods used were so different from study to study that it is difficult to interpret or utilize the results. In particular, since most previous studies used small samples, the differences in methods make it impossible to combine findings to create a database that allows a higher level of generalizability. By standardizing methods, the results of small or focused studies could be pooled to establish larger databases. Moreover, it would be possible to compare the results of studies completed with different populations, e.g., different disabilities, different age groups, etc. Other recommendations focused on the limitations or even undesirability of using standard postures and clothing for this group. Unorthodox ways of standardizing posture and clothing may need to be devised. Finally, some recommendations focused on

With respect to improving standardization, specific recommendations include:

· Develop an ISO standard on how to measure people with disabilities. This standard should identify which body landmarks should be measured and define each landmark. It could also establish guidelines on how to measure people, e.g. use of standard, “normal” or “resting” postures, use of restraints or body bracing during reaching trials, use of two or one hand for measuring reach limits, etc. The definitions should be easy to understand and measure with the tools available. It was also noted that encouraging policies based on an ISO process is problematic because of the difficulty in getting everyone to agree. There may also be many culturally based differences, especially between developed and developing societies. Before starting an ISO process, it was recommended that a good document with terminology and proposed standards be developed first.

· Determine the target population for anthropometric studies through collaboration of researchers and government funding institutions. Researchers need to have some sense of the target population for standards and policy in order to establish sampling plans. By defining a profile for samples through consensus, researchers will have a better idea of what populations need to be recruited and sponsors will have realistic expectations about the generalizability of results for the application of data.

· There is a need to develop simple tools to make the measurements, for example, easy to use measurement devices, pre-developed protocols and databases. This will avoid the need to re-invent data collection methodologies and data management systems. The methods developed should be affordable so that many researchers can adopt them rather than just one or two specialized centers. This will facilitate their use in clinical practice as well.

· Government can take the initiative to fund a “meta -database” through which individual studies could be combined.

These recommendations focused on the need to be cautious in standardizing research methods:

· Avoid reliance on so called “standard postures.” People with disabilities often cannot maintain a “standard” position or may not be able to hold it for long. It may be desirable to measure people in several positions for each task.

· A standard method for taking clothing into account is needed, especially in functional studies where clothing may affect range of movement. Other problems with this population are that many individuals may not be able to wear “standard” articles of clothing, may be more reluctant to be measured nude or partially nude, and may take a long time to change into clothing provided at the research site.

Several participants suggested the use of standardization to improve the application of anthropometric research to practice. These two recommendations summarize these perspectives:

· Government agencies should strive to develop “performance standards” rather than specification standards for design. Such policies could include requirements or options (e.g. “equivalent facilitation”) to use anthropometric and ergonomic tools like computer models in order to accommodate a certain percentage of the target populations. It could also include the development of consensus based testing protocols that can be used for evaluating the accessibility of products.

· Develop a discussion group to standardize the methods in order to tie anthropometric databases to computer aided design (CAD). This could lead to a manual that could be used by researchers and human modelers.

2. INCREASING USE OF FUNCTIONAL APPROACHES TO RESEARCH

The discussion centered on the need to put more emphasis on functional anthropometry than structural anthropometry although the participants recognized the importance of structural anthropometry in order to understand the relationship between body (and assistive device) size and function. While many of the recommendations focused on the general issues of functional anthropometry, there were also a number of very specific recommendations about methodology with a view toward developing research methods that reflect reality more closely.

Recommendations about functional anthropometry in general include:

· Functional anthropometry includes more than simple movement. There is a psychological component. For example, maximum reach may not be reasonable to expect on an everyday basis, especially for tasks that are frequently repeated. An acceptable reach limit for a task would be a more realistic functional approach. This may vary based on the context of the task, e.g. the frequency at which it has to be done, expectations of others, rewards.

· It is important to develop data for specific tasks and population groups. This includes different reaching tasks, wheelchair maneuvers and viewing tasks like reading a computer screen. It also includes understanding the differences between age groups, impairments, sexes and cultural groups. For example, there may be a need to use alternative landmarks for different disability groups. Another issue is defining reach in terms of one hand, two hands or either. Many people need two hands to accomplish something others can do with one. Should reach envelopes be derived separately for one handed or two handed reach or could preferred reach be used as the basis for measurement?

· The way people with disabilities perform tasks and activities in the real world should be studied, not just laboratory based behavior. Ideally this would come prior to the design of laboratory based research so that results can be more relevant to practice.

· Data collected on limited standardized tasks should be related to a wide range of specific activities. For example, data on two handed lifting of an object could be applied to a variety of tasks like lifting something out of an oven, from a refrigerator, from a table or from a work desk. Identifying generic tasks associated with Activities of Daily Living (ADL) could be used as a basis for task-based functional anthropometry.

· In studies of wheelchair users, it is important to investigate how people actually use a wheel chair in different places and under different conditions. The method of transfer used, for example, may vary based on the characteristics of the environment due to preference and adaptation to constraints. Assisted transfers need to be studied as well as independent transfers.

· The type of wheeled mobility device used plays an important role in functional anthropometry. Studies comparing usability of environments by the same people in different types of devices could identify limitations due to the type used, e.g. scooters, as well as identify directions for improvement of different types. In interpreting findings, it is important to consider the appropriateness of devices used by individuals and to avoid basing recommendations on technology that is obsolete or ill conceived.

· The impact of pushing practices and rim design should be
investigated as part of functional studies of wheelchair
maneuvering.

Specific recommendations for methodology include:

· Strength is a critical variable in reaching ability. It is important to understand the relationship of weight of object being moved and strength to task performance.

· Free reach may not be too useful to measure because people are always using or grasping objects while they reach. However, even free reach must overcome gravity so strength always plays a role. This suggests the need to measure the body and body segment mass to understand reach in a biomechanical context.

· Research should take into consideration “functional” reach posture. For example, static and relaxed positions may be more appropriate for measuring functional reach than extended postures. Head movement in functional tasks is often overlooked but it can play an important role, especially if the task involves eye-hand coordination.

· Studies of reaching abilities and grasping tasks for people with disabilities should accommodate different methods of reaching, including two handed reach as well as one handed reach, the use of one hand to brace the other or the use of a wheelchair to stabilize the body while reaching.

· The fact that people may not be fully independent should be recognized in research. When selecting samples or reporting results, independent populations should be separated from dependent populations.

3. BETTER ORGANIZATION OF KNOWLEDGE

Discussion on this theme focused on organizing knowledge with a view toward making it more useable in design practice. Participants were concerned with the organization of the existing knowledge base as well as new knowledge that will be generated by future research. The recommendations include approaches to organizing research data and findings as well as the form of design criteria.

Recommendations included ideas for organizing data for improving our knowledge base and access to research findings:

· Existing technical standards are rigid and incomplete. Performance criteria could be developed for design based on person, environment, tasks and level of assistance to be provided. This approach would be richer and more useful.

· There is a need to review current literature and research as well as past comparisons of existing studies. While Bruce Bradtmiller completed a comparative report, it would be very useful to hold a seminar where the existing work could be discussed further by a broader group of experts that includes designers, human modelers and others.

· A database library should be organized to collect all the data from old and new studies in one place and make it available to others in a consistent and easy to use form. NIST is working on a web-based data-structure for anthropometry which could be a good model on how the government could be involved in setting up a shared database.

· We need to use existing data, best practice information and expert opinion to augment codes and allow exceptions. Organizing data in a form that could be mined for different populations and tasks would help to identify the gaps in our knowledge and frame a research agenda. The Access Board has begun to do some things like this, e.g. the Children’s Guidelines and the Recreation Facilities Guidelines.

· Existing databases should be reanalyzed. For example, the New ANSI Standards Project completed in the late ‘70’s collected data on wheelchair size that was not analyzed in detail. The data was also not analyzed to study age related differences in performance.

· A multi-site initiative using the NIH model would facilitate interdisciplinary perspectives and be useful to evaluate what data would be included in a shared archive or database.

· Other disciplines should be contacted and involved in this area of research. We need to find ways of contacting those groups that would be interested and have something to contribute. One of the most important to involve is the design community to be sure that they will utilize whatever new information is developed.

· Build anthropometric models and automated code checking programs based on anthropometrics into current design tools, e.g. AutoCAD.

· Investigate sociological resources. Perhaps a compilation or bibliography on relevant literature would be a first step.

· Assemble data on accidents involving people who use wheelchairs using an epidemiological approach. Such data would be very useful for understanding the impact of standards and codes in real life.

· There were also several recommendations about how knowledge could be more effectively applied in everyday practice:

· Knowledge from anthropometric research should be integrated into clinical practice in wheelchair design and fitting. Although the emphasis has been on design of the environment and products, research results could improve wheelchair design as well indirectly making the environment more usable.

· Interactive interfaces could be developed to enable non-experts to access and use databases effectively. The AIS system developed at TU Delft is an example. Testing shows that students use it more effectively than data in books.

· A reasoning model for designers should be developed to help them go beyond code requirements. The codes and standards should only be guides. The reasoning model could help designers provide an authoritative rationale for departing from standards (e.g. equivalent facilitation).

· Policy incentives could improve the utilization of knowledge. For example, government purchasing policies could encourage designers to use anthropometric/ergonomic tools as a means to accommodate a given percentage of the population.

· There needs to be more discussion on the most effective way to present and collect anthropometric data in a way that will be useful to designers. This is a major issue of concern among researchers. How can we increase demand for research products? One possibility is to make using anthropometric interesting and entertaining.

Another set of recommendations focused on improving data analysis in research to make use of results more effective:

· Data analysis strategies are needed before large-scale surveys are implemented. Otherwise, the data may not be usable.

· In analysis of data, we should not rely on the normal distribution model. We need to develop new strategies for analysis that take into consideration the great variation in the population of people with disabilities and the fact that there will be small numbers of people in each multidimensional sampling category.

· Averages are meaningless in design for disabilities. Report findings in percentiles and also individual performance. This allows users of the data to determine who is being accommodated and, where the aggregated data is inadequate or not detailed enough, the individual data can be used in human modeling simulations of individuals performing specific tasks.

· Cumulative normal distributions could be more effective than percentiles for code development because they identify clearly which part of the population is included; however, such an approach implies an assumption of normality, which might not be an accurate way to represent the range of body sizes and abilities of those who have disabilities.

· Explore how static anthropometry could be used to predict baseline functional measures. This would help us adapt static data, which is easier to collect, for application to functional tasks.

· Develop 3 -D equivalents of percentiles in order to make it easy to generate recommendations for 3 -D data.