Usability of an adaptive computer assistant that improves self-care and health literacy of older adults.

OBJECTIVES: We developed an adaptive computer assistant for the supervision of diabetics' self-care, to support limiting illness and need for acute treatment, and improve health literacy. This assistant monitors self-care activities logged in the patient's electronic diary. Accordingly, it provides context-aware feedback. The objective was to evaluate whether older adults in general can make use of the computer assistant and to compare an adaptive computer assistant with a fixed one, concerning its usability and contribution to health literacy. METHODS: We conducted a laboratory experiment in the Georgia Tech Aware Home wherein 28 older adults participated in a usability evaluation of the computer assistant, while engaged in scenarios reflecting normal and health-critical situations. We evaluated the assistant on effectiveness, efficiency, satisfaction, and educational value. Finally, we studied the moderating effects of the subjects' personal characteristics. RESULTS: Logging self-care tasks and receiving feedback from the computer assistant enhanced the subjects' knowledge of diabetes. The adaptive assistant was more effective in dealing with normal and health-critical situations, and, generally, it led to more time efficiency. Subjects' personal characteristics had substantial effects on the effectiveness and efficiency of the two computer assistants. CONCLUSIONS: Older adults were able to use the adaptive computer assistant. In addition, it had a positive effect on the development of health literacy. The assistant has the potential to support older diabetics' self care while maintaining quality of life.

The effect of a single course of antenatal corticosteroids on maternal adrenal function at term.

OBJECTIVE: To determine if one course of antenatal corticosteroids at 32 weeks produces maternal adrenal suppression at term. METHODS: The adrenocorticotropic hormone (ACTH) stimulation test was administered at 38 weeks to 11 pregnant women who had received a single course of antenatal betamethasone prior to 33 weeks and to six control subjects. RESULTS: There was no difference in basal cortisol levels (mean+/-standard deviation) between the two groups: 41.6+/-6.9 microg/dl for controls versus 36.0+/-7.8 microg/dl for the steroid group, p=0.16. Peak cortisol levels at 45 min following ACTH stimulation were not different: 61.6+/-3.5 microg/dl for controls versus 55.0+/-2.6 microg/dl for the steroid group, p=0.16. The power of the study to detect a statistical difference in the observed peak cortisol levels was greater than 95%. None of the study subjects had laboratory criteria or clinical signs of adrenal suppression. CONCLUSIONS: A single course of betamethasone for women at risk for preterm delivery does not produce adrenal insufficiency at term and stress dose steroids are not recommended.

An evaluation of warning habits and beliefs across the adult life span.

Beliefs about warnings and habits associated with reading them were assessed for 863 individuals of various ages. Information gathered for various common household products included (a) how frequently people attend to warning information, (b) the degree of risk they believe is involved during product usage, and (c) how important they believe warnings are for different product types. Also assessed were perceived helpfulness and comprehension for symbols commonly found on product labels or on signs in the environment. Respondents 55 years and older reported reading product warnings more frequently than did younger adults, although they generally perceived warnings as less important. However, no overall age-related differences were found for perceived level of risk involved in using different product types. Although older adults generally perceived symbols to be very helpful when using a particular product, their comprehension levels were poorer than those of younger adults for half of the symbols. Overall, these data suggest that adults of all ages do read warnings on a variety of product types and that they believe warning information is important. This research illustrates the importance of including older adults in usability studies during the development of warning systems, given age-related effects may be associated with some aspects of the warning processing but not others.

Age-related differences in dual-task visual search: are performance gains retained?

Older and young adults practiced a verbal/spatial dual task and were tested for retention performance 1 month later. Participants first practiced each component task separately to individually determine component processing time. Thus, age-related differences in single-task detection sensitivity were minimized prior to performing the dual task. Participants practiced the dual task for two 1.5-hour sessions. Following the retention interval, they were retested on the single-task components and on the dual task. Correct detection as well as signal detection parameters were examined. Older adults demonstrated decreased sensitivity as well as a more conservative response bias during acquisition. Retention performance for the single tasks replicated previous retention studies, demonstrating age-related performance declines when stimulus-specific learning is assessed. Dual-task retention capability declined for both older and young adults equally when detection accuracy, but not perceptual sensitivity, was measured. Response bias changed differentially for older and young adults across the retention interval.

An individual differences analysis of ability and strategy influences: age-related differences in associative learning.

The relationships among abilities, strategies, and performance on an associative learning task were investigated for young (aged 17 to 34) and older adults (aged 60 to 82). Participants received extensive practice on a noun-pair task in which they could use a visual-scanning strategy or a memory-retrieval strategy. Older adults were more likely to use the scanning strategy. Age differences were reduced when comparisons were made only for participants using a retrieval strategy. Associative memory was predictive of learning on the task, and semantic memory access speed was predictive of practiced performance. Practiced performance on a memory-search task that also required associative learning was predictive of practiced noun-pair performance. Models of ability-performance relationships for skill acquisition are discussed.

Age-related performance in a multiple-task environment.

Younger and older adult participants performed a dynamic multiple-task requiring concurrent processing of 4 independent tasks. Component-task performance emphasis (i.e., task priorities) was biased by differential point allocations across task components. After training, the point structure was modified. Older adults exhibited larger multiple-task performance deficits compared with younger adults; however, the age-related gap in multiple-task performance decreased with practice. The age-related performance difference increased again when task emphasis was changed, but not when demands were changed. Consistent with the training data, the age-related differences diminished again with additional experience on this new task-component emphasis. The data suggest that higher-order, strategic processing may be an important source of age-related differences in complex multiple-task performance. Actual or potential applications of this research include the facilitation of techniques for age-related comprehensive usability testing for products of even moderate complexity.

Measuring skill acquisition and retention with an ATM simulator: the need for age-specific training.

The present study focused on the type of information presented during training and its effects on initial and retention performance of older and younger adults interacting with computerized, new technology. The effects of emphasizing concepts versus actions during training on performance immediately after training and after a 1-month retention interval were examined. Younger and older adults completed either action or concept training for operating a virtual automatic teller machine (ATM). Overall, action training was associated with faster and more accurate performance immediately after training and better retention performance for older adults. For older adults, value of type of training interacted with type of task component. These findings are applicable to the development of age-specific training materials for computerized tasks.

Functional limitations to daily living tasks in the aged: a focus group analysis.

We assessed constraints on daily living of 59 healthy, active adults 65-88 years of age in focus group interviews. Individual comments about specific problems were coded along the dimensions of (a) the locus of the problem (motor, visual, auditory, cognitive, external, or health limitations); (b) the activity involved (e.g., transportation, leisure, housekeeping); (c) whether the problem was attributable to task difficulty or the perception of risk; and (d) response to limitations (perseverance, cessation, compensation, or self-improvement). The data provide information about the types of difficulties encountered in everyday activities as well as the way in which individuals respond to such difficulties. Each comment was also coded in terms of whether it was remediable via training, design changes, or some combination of the two. More than half of the problems that were reported had the potential to be improved in some way, thus providing direction for future research in human factors and aging. Actual or potential applications of this research include identifying problems and difficulties that older adults encounter in daily activities such as transportation and leisure; more specifically, determining the degree to which such problems are potentially remediable by human factors solutions. Applications of this research also include understanding the types of systems, products, and technologies that older adults interact with currently, or are interested in learning to use.

Aging and decision making: driving-related problem solving.

We examined age-related effects on decision making in a task environment familiar to most younger and older adults. Participants made route-selection decisions in real time. Participants received information about traffic density and expected speed limits of main and alternative routes, from which they determined the optimality of their present route versus alternative routes. The experiment evaluated the effects of information type, amount of congestion, alternative route speed limit, and age on speed and quality of decision making. Measures of optimal route selection revealed main effects of alternative route speed limit, congestion level, and message type, but there was not a main effect of age, and age did not interact with any variable. In terms of decision speed (but not quality of decision making), older participants were slower, and age interacted with alternative route speed and with message type. The data are interpreted in relation to previous data examining everyday problem solving and aging.

Automatic category search and its transfer: aging, type of search, and level of learning.

We examined the site of learning as a function of task type and age. Two experiments examined whether learning in semantic category search is exclusive to trained elements of categories or generalizable to other elements of the trained categories. Specifically, we examined how practice searching for small subsets of exemplars from taxonomic categories transferred to untrained elements of those categories. Young and old adults received extensive practice on memory search (Experiment 1) or visual search (Experiment 2) tasks. Participants then transferred to conditions assessing whether learning was exclusive to the trained words or generalizable to other elements of the trained categories. The site of learning in memory search appears to be at the category level for both young and old adults. Level of learning in visual search appears to differ as a function of age. Young adults' learning generalizes to the category level, whereas older adults' learning is specific to the trained words.

Age-related differences in movement control: adjusting submovement structure to optimize performance.

In this experiment older and younger adults were compared on their ability to position a cursor with an electromechanical mouse. Distance of the movement, size of the target, and relative emphasis on the speed or accuracy of the movement were manipulated. The study was designed to isolate and evaluate the effects of age-related differences in the noise-to-force ratio, perceptual feedback efficiency, strategy differences, and the ability to produce force as explanations for age-related differences in movement control. This was done by using two types of movement tasks and by analyzing movement performance according to stages of movement. The study showed that all four factors, when isolated, are significantly different for the two age groups. However, in the task component where all factors could simultaneously affect performance, the age-related difference in performance was less than the difference in either the measure of noise-to-force ratio or perceptual efficiency. Analysis of the submovement structure revealed how older adults compensated for the greater noise and less perceptual efficiency by adjusting the velocity and number of submovements. These findings are discussed in light of the optimized submovement model.

Training older adults to use automatic teller machines.

The present study assessed the success of several instructional programs in teaching the use of automatic teller machines (ATMs). Fifty-six older adults (aged 61 to 81) participated in the study, randomly assigned to each of four, 14-member groups. The description group received only a general overview of an ATM; the text guide group received written instructions for performing various transactions on an ATM; the pictorial guide group received written instructions accompanied by pictures of corresponding ATM screens; and the online tutorial group completed a step-by-step tutorial on a simulated ATM. Participants practiced on an ATM simulator. They were tested after a 24-h interval on their ability to perform familiar transactions on an unfamiliar ATM simulator and to perform completely novel transactions. Accuracy was best for the online tutorial group, intermediate for the text and pictorial guide groups, and worst for the description group. These data demonstrate both the importance of providing older adults with ATM training and the fact that the type of training influences the level of performance. The online tutorial, which provided specific practice on the task components, best facilitated acquisition and transfer performance.

Aging and individual differences in the development of skilled memory search performance.

We examined individual differences in measures of multiple intellectual abilities and performance on a pure memory search task over 5 experimental sessions. Old (n = 104) and young (n = 97) participants showed expected patterns of substantial improvement in memory search intercepts and slopes in consistently mapped (CM) conditions, relative to varied mapping (VM) conditions. Initial (unskilled) CM and VM memory search was highly correlated with a Semantic Memory Access Speed factor and moderately correlated with General Intelligence. Structural equation models showed that measures of Semantic Memory Access remained a strong predictor of skilled CM search performance in both age groups despite individual differences in CM memory search performance changes. These results indicate qualitative differences in the nature of automaticity between memory search and visual search and suggest age invariance in the mechanisms determining automaticity in memory search.

Supporting perception in the service of dynamic decision making.

Skilled performers in complex environments rely heavily on heuristic strategies to cope with the time pressure and complexity of dynamic tasks. We suggest that the use of task simplification strategies based largely on perception and pattern recognition is fundamental to the novice-expert shift in dynamic decision making. We therefore suggest that interface training interventions should support the development of highly effective and robust heuristic strategies, rather than the development of more abstract, cognitively intensive strategies. A pair of empirical studies are presented that investigated the benefits of training interventions aimed at supporting perceptual and pattern-recognitional activities in dynamic environments. Results suggest that the acquisition of skilled performance in dynamic environments can be accelerated by supporting perceptual activities in the service of dynamic decision making. Implications of these results for training, aiding, and interface design are discussed.

A survey of automatic teller machine usage across the adult life span.

The purpose of this study was to analyze automatic teller machine (ATM) usage across the adult life span. We conducted an extensive survey of 9000 people in the Memphis and Atlanta metropolitan areas. Approximately 17% of those people responded. The survey assessed detailed demographic information, experience with technology in general, experience specifically related to ATMs, problems and dislikes with ATMs, and reasons that people do not use ATMs. The survey provided a valuable set of data. First, we have detailed information about the demographics and individual characteristics of ATM users and nonusers; importantly, these data are stratified across the adult life span. In addition, we know the likes and dislikes of ATM users and the types of problems they typically have using ATMs. Moreover, we have detailed analysis of why adults of all ages may choose not to use ATMs. Training and design implications of these data are discussed.

Understanding performance and learning in consistent memory search: an age-related perspective.

This study was conducted to investigate learning in memory search tasks. Young and old participants were trained for 5,640 trials of consistent mapping (CM) and varied mapping (VM) memory search. After training, participants were transferred into New CM and CM Reversal conditions. During training, both young and old adults improved reaction time performance, with more rapid improvement for the young adults. In CM training, both age groups achieved zero comparison slopes, indicating automaticity in CM memory search. VM training maintained a large age-related difference in search times. Age did not moderate the transfer effects, suggesting similar learning mechanisms were responsible for the original CM training gains in both age groups; however, transfer effects were different for CM Reversal and New CM. The pattern of transfer data argues against several possible mechanisms for automaticity in memory search. The data are most compatible with a hypothesis of memory-set unitization as the locus of automaticity in memory search.

Age-related retention of skilled memory search: examination of associative learning, interference, and task-specific skills.

Young and older adults performed skilled memory search after either a 3- or 6-month retention interval. Participants were first trained in consistent-mapping (CM) memory search; then, one of the search conditions was subjected to interfering processing activity prior to the retention interval. Retention testing simultaneously examined situations where interfering processing activity either did or did not intervene between original learning and retention testing. In addition, general task-specific learning was assessed. Results indicate that (a) old and young adults equally retained general, task-specific skills; (b) old adults' performance declined more than young adults' performance for trained CM stimuli; (c) when an interfering processing activity was inserted prior to the retention interval, old adults' performance declined disproportionately more than young adults' performance, especially when compared to the task not subjected to such interference; and (d) for both old and young adults all initial retention deficits were quickly eliminated within retention retraining.

Long-term retention of skilled visual search: do young adults retain more than old adults?

Young and old Ss received extensive consistent-mapping visual search practice (3,000 trials). The Ss returned to the laboratory following a 16-month retention interval. Retention of skilled visual search was assessed using the trained stimuli (assessment of retention of stimulus-specific learning) and using new stimuli (assessment of retention of task-specific learning). All Ss, regardless of age group, demonstrated impressive retention. However, age-related retention differences favoring the young were observed when retention of stimulus-specific learning was assessed. No age-related retention differences were observed when task-specific learning was assessed. The data suggest that age-related retention capabilities depend on the type of learning assessed.

Do ability-performance relationships differentiate age and practice effects in visual search?

Relationships between abilities and performance in visual search were investigated for young and old adults. Ss received extensive practice on category search task. A consistent version allowed development of an automatic attention response; a varied version allowed general performance improvements. Transfer conditions assessed learning. General ability, induction, semantic knowledge, working memory, perceptual speed, semantic memory access, and psychomotor speed were assessed. LISREL models revealed that general ability and semantic memory access predicted initial performance for both ages. Improvements on both the consistent and varied tasks were predicted by perceptual speed. Ability-performance relationships indexed performance changes but were not predictive of learning (i.e., automatic process vs. general efficiency). Qualitative differences in the ability-transfer models suggest age differences in learning.

Brinley plots and theories of aging: the explicit, muddled, and implicit debates.

We evaluated the debates concerning Brinley plots and the associated theories of age-related slowing. We concluded that an explicit debate regarding a single-factor, general slowing model was no longer a debate as most, if not all, agree to the disconfirmation of that model. We address sources of confusion in the debates that have muddled the areas of genuine disagreement. When the confusion is lifted, the remaining debate centers, rightly, on evaluation of theories of aging. We show that Brinley plot analyses can lead to both falsely accepting and falsely rejecting theories of age-related slowing. Although plotting data most certainly can assist with the evaluation of cognitive theory, we argue that models of performance and learning must play a more central role in advancing theories of cognitive aging.