Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 What New Knowledge Is Needed? On the bash of the workshop, position papers, and the existing literature, the Apes of knowledge needed to advance human fac- tors research ~d engineering for the elderly can be enumerated and described. For the most part the presang needs are not esoteric theoretical ad~ces or ascientific breskthroughs~ but merely the collection of systematic data where none exist. Moreover, no fun- damental advances in in~restigative methodology are required, but rather the effective application of already demonstrated research took. ~ essence, the foBow~g fire major cI"ses of knowledge are needed. DIl3TR~RU]?IONAI DATA ON TAS1~13, 131TUAlqONS, AND ACTIVITD:S It wiD be important to gather more information about the way older people allocate their time to various activities. Specifically, what ~ needed are detailed t~me-eample, observational, and survey data on what people of pervious ages do all day and why. For example, with respect to transportation, we need to know how older people get Tom one location to another; where, when, and on what kinds of roam older people drive; "d what type of location information they use. Similar information ~ required for other activities such as daily living tasks sod recreational activities. These data should be collected at the appropriate level. For example, the application of t~k-analytic techniques, such as those used by Faletti (1984) in 59
60 RUMAN FACTORS RESEARCH NEEDS FOR AN AGING POPULATION the study of meal preparation, would be useful to collect initial de- scriptive data on activity patterns and demands. In Faletti's study, videotapes were made of a sample of people actually preparing meals in their homes. The tapes were then analyzed in detail In order to understand the demands (e.g., movements, grips, postures) associ- ated with meal preparation tasks. Data of this type can be used to develop models of task performance and to set the stage for more systematic laboratory or s~mulation-based analyses of human factors issues associated with various activities. Such data can be used, for example, to identify activities, equipment, or environmental supports that wiD serve to unprove task performance. P1tOB[EM DATA In addition to understanding patterns of activity performance, we need to understand which tasks are problematic for c'lder adults and the types and frequency of difficulties associated with these ac- tivities. The data collection methods previously cited would provide valuable information on these questions. However, some answers Knight be obtained more economically by abstracting those activities associated with frequent problems and by studying their execution under controlled laboratory conditions. For example, we know that using stairs is typically problematic for older adults. It might prove useful to study stair climbing ~ a laboratory setting where variables such as stair design features and lighting could be systematically varied. This would adore the use of instrumentation and protocols to identify the exact sources of visual, motor, or cognitive difficulty. Also, it might be beneficis} to compare demands across tasks so that representative tasks could be selected for more detailed anal- ysis. This would allow us to generalize with respect to functional assessment and the application of intervention strategies. 1lUNCTIONAL NORMS In many aspects of human factors work, data on the distribu- tion of human characteristics across populations ~ a fundamental tool. For example, ~ order to design controls such as those found on appliances or in bathtubs/showers or to determine the apprm priate height and depth of shelves or cabinets, we need such data as comfortable reach dimensions, hand dunensions, and hand grip strength for a broad range of users. Similar data are needed on sen- sory, perceptual, motor, and cognitive functions. Currently, there is
WHAT NEW KNOWLEDGE IS N-FtF1nED? 61 very little dynamic anthropometr~c data available on an age-specific basis, especially for older cohorts (Stoudt, 1987~. The utility of normative data on adult populations is twofold. Fit, this type of data serves as a bash for general analysm and design and aBows us to design for the whole population. To account for the entire range of the population, we need to understand to what extent aging changes the distributions of ~ ous charactermtics. To date, characteristics of the older population have been largely ig- nored in engmeenog design, "d thus the elderly often have difficulty manipulating the designed environment. For this reason many older adults currently need some type of assistance to complete living ac- tivities. Secondly, these data would be useful in explicit analysm "d design to address the specific proble}ns associated with aging, as, for example, ~ the design of amistive devices. In addition to normative functional data for older people, data on the incidence of specific disabilities, dysfunctions, and diseases are needed. For example, it ~ important to know what percentage of older adults are afflicted with arthritis and how this affects their functional capabilities. Collecting normative data ~ prunarily a matter of systemati- cally sampling populations of different ages and obtaining apprm priate measurements. In deciding what measurements to make, it is important to be guided by information on the frequency with which activities and difficulties performing the activities occur. For example, if opening jars is a Sequent activity and causes frequent difficulty, then the strength distribution of the related twang mo- tion should be a priority for data collection. As indicated, this type of information can be gathered from task-analysis and time-sample data. When collecting distributional data, it is important to recognize that the elderly, as a group, are heterogeneous; the changes with aging are highly variable both within and across individual. This implies that reported data should represent the variabilities at each age as wed as averages. It also unplies that it ~ unportant to under- stand what functions and abilities increase with age as well as those that decline. FOCUSED HUMAN FACTORS litESEARCH ISSUES Ultimately, we need answers to specific scientific and practical
62 RUMAN FACTORS R~513ARC~ NEEDS FOR AN AGING POPULATION questions about identified hurrah factors and aging problems. Sur- vey and t~k-~ysm Formation will fib in gaps In our ex~t~g knowledge base ard win identify the problems and some of their parameters. However, solutions to these problems win ~ most cases require additional, more narrowly direct research. For example, it is known that older people have difficulty learning computer-based text editing and that they tend to have increasing difficulty with complex cognitive tasks. But to understand the relationship be- tween these To pieces of information and to fired a way to make computer tex~editing problerna easy for the elclerly, we need to know which aspects of the tex~edit~g task are especially demon cling for them. similarly, we currently know that certain types of automo- bile accidents are more common among older people, and we know that older people have, on average, deficiencies in a variety of visual and motor abilities. Yet we still lack information regaling which aspects of driving cause stress to age-related abilities. Similar issues can be ramed about stairways. It is not the intention or purpme of this report to provide an exhaustive list of research questions but to suggest some examples for initial research. As we begin to collect more detailed information on the activity patterns of older people, additional areas of specific research will be identified. The routes for de iiing tenth specific problems, once identified, may be differentiated as remediation (eliminating the dysfunction or the environmental problem); compensation (fining some alternative way to achieve the same goab); adaptation (restricting one's beha~r- ior pattern to deal with the problem despite its difficulties); and accommodation (learning to substitute other goad ~d activities for those that are unachievable). AD of these solution paths may involve significant humor factors research, analysis, add design. DESIGN PRINCIPLES AND EXAMPLES Knowledge that would help reduce the number of problems that the elderly face would cons~t of actual solutions to a representative sample of design problems. If several important t~k-related prom letrm ~n~rolv~ng perceptual difficulties (e.g., reduced night vision for driving); cognitive difficulties (e.g., reduced ability to perform com- plex tasks such as text editing); ~d motor problems (e.g., street crossing) could be reduced by systematic efforts at analyze and re- design, both the methodological examples ~d the substantive design techniques would be likely to generalize to many other problems. A
W]IAT NEW KNOWLEDGE IS NF!F!nED, 63 few such successes might also serve to stimulate greater efforts on the part of researchers and design ere. The two most important contnbuto" to good human factors de- sign are task versus ability analysm add iterative design. By iterative desigm we mean a process in which design ideas are implemented in an easily modifiable prototype or simulation; tested on a represents tire sample of actual users; modified according to obeer~red successes, failures, and~mpremions; ardtr~ed again. Su~design methods have been found to be extremely effective and cost efficient, especially in the development of computer-based systems where the flexibility and power of the machine make extensive adaptation to the user fe~i- ble (Gould, Boies, I,e~ry, Richards, and Schoonard, 1987; Landauer, 1988; Good, Whiteside, Wilson, and Jones, 1984~. Fourfold to ten- fold improvements ~ speed and accuracy resulting from a few weeks of effort are common with this approach. The essential ingredient is aformative evaluations to guide design early in the invention ~d development phase. Given that the evaluation must Solve people with the same characteristics ~ the eventual users, and given that most systems, products, =d environments are used by a wide range of users, older adults must be included in such iterative testing.
