The Gerontologist Vol. 42, No. 1, 39–50 Copyright 2002 by The Gerontological Society of America A Social–Cognitive Model to Predict the Use of Assistive Devices for Mobility and Self-Care in Elderly People Marc Roelands, MA,1 Paulette Van Oost, PhD,1 AnneMarie Depoorter, MD,2 and Ann Buysse, PhD1 Purpose: To provide insight into the contribution of psychological variables in understanding use and nonuse of assistive devices (ADs) for self-care and mobility. Design and Methods: A survey on a representative sample of 491 community-dwelling elderly people in Flanders was conducted. A theoretical model of the factors predicting the use of ADs was tested with path analysis. Results: The possession and use of 32 ADs were found to be bivariately related to intention to use ADs, awareness of ADs, attitude toward AD use, subjective norm regarding AD use, self-efficacy concerning AD use, and socio-demographic characteristics. In regression analyses the intention to use ADs was found to be related to self-efficacy concerning AD use, attitude toward AD use, and subjective norm regarding AD use. Implications: This study suggests that a social–cognitive model of human behavior can be helpful in explaining the use of ADs. The model can suggest measures to be developed aiming to improve the introduction of ADs. The prevalence of AD use has been estimated in the general population in the United States (LaPlante, Hendershot, & Moss, 1992; Mann, Hurren, & Tomita, 1993). In the age group of 65 to 74 years old, 22% of the U.S. population use mobility ADs, which increases in the age group of 75 years old and older to 43%. The type of AD used by elderly persons has been investigated in 76-year-old Swedish people: Grab rails (16% of total population) and canes (15%) were used by the largest proportion of the population (Sonn & Grimby, 1994). To explain and predict the use of ADs, models that incorporate many factors involved in the use of ADs, including social and psychological factors, have been proposed on the basis of clinical evidence or literature review (Rogers & Holm, 1992; Scherer & McKee, 1989). We believe, however, that these important models could benefit from both empirical/methodological improvements, such as appropriate data-analytic techniques, and theoretical improvements. A more established theoretical background would allow an effective protocol to be developed to introduce ADs in a more structured way by taking crucial variables into consideration. Several social–cognitive models of human behavior have been described that predict human behavior quite adequately. They emphasize the role of attitudes, subjective norms, behavior intention, self-efficacy (Ajzen, 1991; Bandura, 1997), and awareness (McGuire, 1985). Key Words: Socioeconomic status, Self-efficacy, Attitude, Care, Planned behavior During the last few decades assistive devices (ADs) for self-care and mobility have been developed to improve the fit between the competency of elderly persons and the demands of the environment. Examples of ADs are wheelchairs, shower seats, and sock aids. They are intended to improve the autonomy and quality of life of elderly persons in a cost-effective way. Furthermore, they are prescribed with the aim of influencing the quality and cost of social care and health care in a positive way by lowering the need for intramural care (de Klerk & Huijsman, 1995). A Social-Cognitive Model of the Use of ADs A tentative model can be described, incorporating some of the factors that are involved in communitydwelling elderly persons’ use of ADs. The framework is the social–cognitive theory of planned behavior (Ajzen, 1991). The theory of planned behavior takes into consideration the influence of personal evaluations (attitudes), perceived social pressure (subjective norm), and perceived behavioral control in predicting the intention to perform a specific behavior. The predictors are understood as expectations weighted by personal values. It is therefore also referred to as an expectancy value model (Ajzen & Fishbein, 1980). It has been used suc- This study was funded by Flemish Government Grant 174BG998. We gratefully acknowledge the assistance of Hilde Merckx and Marianne DeGrave in the data collection. Address correspondence to Marc Roelands, Ghent University, Faculty of Psychology, Dunantlaan 2, 9000 Ghent, Belgium. E-mail: marc. [email protected] 1 Ghent University, Faculty of Psychology, Research Group Ageing Studies, Belgium. 2 Free University of Brussels, Public Health Care, Belgium. Vol. 42, No. 1, 2002 39 cessfully in understanding and predicting health-related behaviors such as alcohol consumption (Conner, Warren, Close, & Sparks, 1999), uptake of a cervical smear test (Bish, Sutton, & Golombok, 2000), condom use (Gagnon & Godin, 2000), dietary behavior (Povey, Conner, Sparks, James, & Sheperd, 2000), the adoption of hormone replacement therapy (Legare, Godin, Guilbert, Laperriere, & Dodin, 2000), and exercise adherence (Rhodes et al., 1999). An overview can be found in Armitage and Conner (2000). When adapted to the use of ADs, the following concepts can be organized at four levels. In the context of the presence of disabilities (disabilities are understood as activity limitations; see World Health Organization, 1980), the intention to use ADs (Level 2) is the most direct determinant of the actual use of ADs (Level 1). The intention to use ADs is determined by the following proximal variables (Level 3): the person’s attitude toward the use of ADs, the AD self-efficacy, and the subjective norm regarding ADs. Bandura (1997, p. 3) described self-efficacy as “beliefs in one’s capabilities to organize and execute the courses of action required to produce given attainments.” Ajzen and Madden (1986) stated that the self-efficacy concept is similar to the perceived-behavior-control concept in their theory of planned behavior. Perceived behavior control is defined as the person’s belief as to how easy or difficult the performance of a behavior is likely to be. Because self-efficacy beliefs are specific regarding activity domain, we prefer the term assistive devices self-efficacy, which is used in our model. The subjective norm refers to a person’s perception of the strength of significant others’ expectations to use ADs and the will to conform to these expectations, and has been proposed as a factor influencing the decision to use ADs or ask for personal help (Scherer & McKee, 1989). The theory of planned behavior does not include background variables, as it is assumed that individual characteristics affect the central concepts in the model indirectly by shaping beliefs and evaluations (Ajzen & Fishbein, 1980). Following McGuire (1985), we included awareness as distal variable (Level 4). 70 and 89 years old in 1999, (c) were communitydwelling, and (d) own a telephone. Exclusion criteria were severe hardness of hearing, severe cognitive dysfunctioning, and a general health level that hampers an interview. Severe impairment of sight was an exclusion criterion because this hampers the measurement of several variables. Only Dutch-speaking persons were included. The sample was limited to persons between 70 and 89 years old because in the age group of people 90 years and older refusal and exclusion were expected to be high because of high morbidity. Moreover, exclusion was expected to increase because of lack of a telephone. The group of persons under 70 years is too healthy for the purpose of this study, which would require a considerably larger sample. The required sample size was estimated to be about 500 persons. For statistical reasons the sample ought to be large enough to include at least 20 individuals of each gender with severe disabilities on the basis of the assumption that AD use is related to disability. The prevalence of most disabilities is lowest in men and is about 10% or higher (Demarest et al., 1998). Taking into consideration the distribution of gender in the population and assuming that AD use is related to disabilities, it could be calculated that the smallest groups in the study include 20 persons if the sample size is 500 (500 persons 0.4 men 0.1 disabilities). We opted for a representative sample to allow extrapolation of the descriptive data toward the population. An indirect or two-stage sample selection was used to limit the geographic dispersion. In the first stage, a limited number of municipalities were randomly sampled. Because a connection between health behavior and socioeconomic status (SES) has been established (Lynch, Kaplan, & Salonen, 1997; Stronegger, Freidl, & Rasky, 1997) and to avoid a loss of representativeness regarding this important variable—which could occur because of the relatively small number of municipalities selected—the sample of municipalities was stratified by SES. A typology of all municipalities in Flanders, on the basis of the equipment of the houses and supplemented with some financial indicators, was available (Thomas & Vanneste, 1997), and was chosen for two reasons. First, the equipment of the houses is considered to be a good indicator of the SES of a population (Thomas & Vanneste, 1997). Second, the quality of the house is a measure of the environment in which AD is used. Six types are distinguished, of which five can be found in Flanders. The municipalities were selected as follows (Deliens, 1998). A random sample of municipalities was selected within each stratum. To make the sample self-weighting, which avoids conversion of the results to the population, the population size had to be taken into consideration. A municipality could receive 0, 1, or more sample points. A sample point consists of 14 effective persons and 2 14 persons who were on the reserve list. The 40 sample points were distributed over the strata of municipalities in proportion to the size of their population. Within these strata, sample points were distrib- Objectives This study aimed to provide insight into the contribution of psychological variables in understanding AD use and nonuse. The objectives of this study were threefold: (a) to describe the possession and use of ADs in community-dwelling elderly adults in Flanders; (b) to describe community-dwelling elderly persons’ awareness of ADs, their attitudes toward the use of ADs, their subjective norm regarding the use of ADs, and their self-efficacy concerning the use of ADs; and (c) to test the link between these determinants and AD use in a social–cognitive model of AD use. Methods Sample Selection The population was defined by the following criteria: (a) people living in Flanders who (b) were between 40 The Gerontologist uted using a systematic drawing to warrant random selection. If a municipality received no sample points, this implied that it was not included in the sample of municipalities. In the second stage, in each of these municipalities a random sample of people born between January 1, 1910, and December 31, 1929, was selected, using the population register provided by the municipalities. Of the 35 municipalities that were originally contacted, 2 refused to cooperate for reasons of protection of privacy and were replaced at random within the same stratum. Persons who could not be contacted after three attempts or refused to cooperate were replaced by the next person on the randomly ordered list from the random sample in the municipality until this sample was exhausted. Five subscales were developed to measure the awareness of groups of ADs (those that help with dressing, washing, mobility, going to the toilet, bed transfer) and five subscales were developed to measure the use by groups of ADs. The items can be found in the table on awareness. In the scale on mobility the personal alarm system was not included. The Cronbach’s alphas of the awareness scales were acceptable, taking into account the dichotomous nature of the items: .57, .53, .74, .54, and .54, respectively. The Cronbach’s alphas of the scales relating to AD use were .05, .25, .38, .12, and .43, respectively. Attitudes Toward AD Use.—The Attitudes Toward Assistive Devices Scale (AADS) consists of 12 items that measure the opinions of community-dwelling elderly persons on different aspects regarding ADs: the substitution of care, the financial aspect of care, and the perceived effect on privacy. The items are listed in the table on attitudes. To increase the internal consistency of the scale to measure the multidimensional concept, while keeping it short, we first slightly modified the 21 items that were developed by de Klerk and Huijsman (1993) regarding wording and presented them to community-dwelling elderly persons. Twelve items could be retained using factor analysis on the 21 items (Roelands, Van Oost, & Depoorter, 1999). An example of an item is “When using ADs you are less dependent on others.” We chose a Likert scale with 5 points from 5 (totally agree) to 1 (totally disagree). A total score was calculated. The theoretical minimum score on this interval scale is 12, the maximum score is 60. A high score indicates a positive attitude. The scale has high face validity. Comparison with established measures of attitudes toward ADs to estimate the validity is hampered by a lack of these measures. The internal consistency was .61, which is sufficient. To further elaborate on the attitude toward the use of ADs, we developed a scale within the expectancy-value model of attitude, measuring attitude according to the guidelines of Ajzen and Fishbein (1980). Six modal salient beliefs about the consequences of AD use were obtained from the literature (de Klerk & Huijsman, 1995; Sonn & Grimby, 1994), regarding autonomy, stigmatization, efficiency, speed of performance, care substitution, and safety. An example of an item is “If I use ADs other people will think I am old and sick.” The strength of the belief could be indicated on a bipolar Likert scale with 5 points from 5 (totally agree) to 1 (totally disagree). All respondents also valued the personal importance of these beliefs. An example of an item is “To what extent is it important to you that other people do not think you are old and sick?” For each belief the person evaluated on a bipolar, 5-point Likert scale whether this consequence was 5 (very important) to 1 (very unimportant) to her or him. The product of the level of agreement regarding a consequence and its evaluation was calculated for each perceived consequence. A total score of the six products was calculated. A high score indicates that positive conse- Data Collection Procedure Once selected, the person was informed of his or her selection by letter. Thereafter, the researcher contacted the person by telephone to ask for his or her consent and to make an appointment. Trained interviewers, using a standardized questionnaire to interview elderly persons at home, did the data collection. The response categories were presented to the interviewee. Instruments Awareness, Possession, and Use of ADs.—The Assistive Devices Awareness Scale (ADAS) has been developed by the authors to measure awareness of ADs and possession and use of ADs (Roelands, Van Oost, Buysse & Depoorter, in press). Thirty-two ADs were chosen on the basis of their prevalence in patients of home nurses, as reported by a sample of nurses. The 32 ADs were presented visually by using photographs and mentioning their names. For every AD, the person was asked some interrelated questions. First, the person was asked whether he or she possessed it or not. If the person did possess the AD, he or she was asked whether he or she used it (code 2) or not (code 1). If the person did not possess the AD, he or she was asked whether he or she knew of its existence (code 3) or not (code 4). Because of the interdependency of the concepts of awareness, use, and possession and because of the attention that has been given to the smoothness of the interaction during the interview some calculations are necessary to arrive at the dichotomies of “aware or not,” “use or not,” and so forth. To calculate the scale for awareness of supply, we recoded the answer for every AD in two categories: aware (codes 1, 2, and 3) versus not aware (code 4). To calculate the scale for the possession of AD, we recoded the answer for every AD in two categories: possessed (codes 1 and 2) versus not possessed (codes 3 and 4). To calculate the scale for the use of the AD, we recoded the answer for every AD in two categories: use of AD (code 2) versus not used (codes 1, 3, and 4). The minimum score on the three interval scales is 0, the maximum score is 32. Cronbach’s alpha was .86, .68, and .50, respectively. Vol. 42, No. 1, 2002 41 quences are expected. The scale ranges from 6 to 150. The reliability of this interval scale was sufficient: Cronbach’s alpha was .58. intention to do this at all) to 5 (I certainly have the intention to do this) was used. A total score was calculated, with a high score indicating a high intention to use ADs. The interval scale ranges from 3 to 15. To optimize the distribution of the scores, we performed a square root transformation. This score was used in the analysis. The reliability of this scale was good: Cronbach’s alpha was .71. Subjective Norm on AD Use.—The Subjective Norm on Assistive Devices Scale (SNADS) was developed for the study to predict the subjective norm regarding the use of ADs. The scale measures with one item the significant others regarding the use of ADs (partner, children, friends, general practitioner, nurse, and other professional caregiver), with six items on the normative beliefs, and with six items on the motivation to comply with the beliefs. There is one item per significant other measuring the normative belief and one item measuring motivation to agree with that significant other. An example of a normative belief item is “Do you think that your partner thinks you should use assistive devices in case you cannot perform an activity on your own?” A bipolar, 5-point Likert scale ranging from 5 (totally agree) to 1 (totally disagree) was used. An example of the motivation to comply items is “To what extent would you allow yourself to be guided by the opinion of your partner?” The unipolar, 5-point Likert scale ranged from 5 (very strongly) to 1 (very little). The product of the strength of the normative belief and the corresponding motivation to comply was calculated for each normative belief, ranging from 1 to 25. A total score of the six products was calculated. A high score indicates a high subjective norm. The interval scale ranges from 6 to 150. Cronbach’s alpha was not computed as the assumption of internal consistency does not hold for this linear composite score. SES.—As a measure of the SES of the person, we used the objective income (“What is the net income of the household?”; Demarest et al., 1998). Functional Status.—Functional status was measured with two instruments. A subscale of the Dutch version of the Short Form-36 (SF-36) measured physical functioning (Aaronson et al., 1998). The scale on physical functioning consists of 10 items on a 3-point Likert scale, with 10 being totally unlimited, 5 being a bit limited, and 0 being seriously limited. With our scoring, a high score indicates a high level of physical functioning. Cronbach’s alpha was .93 in this study, a value that was also found by Gandek and colleagues (1998). An activities of daily living scale (ADL scale) was included, too. It consists of two scales. One is the Katz Scale (Katz & Akpom, 1976), which is a measure of functional status, more specifically a measure of the need for care. Our version of the Katz Scale is the official version that is used for financing Belgian home nursing care. It consists of six items: dressing, washing, indoor mobility, eating, continence, and going to the toilet. Each item has four particular response categories, the content of which depends on the specific item. Moreover, in this study, six items were added, giving valuable information regarding disabilities in instrumental ADL (IADL; shopping, cleaning, washing clothes, ironing, cooking, and managing the housekeeping). Four response categories are provided to measure IADL that differ from the Katz Scale: (a) “is completely independent regarding . . .” (b) “is independent regarding . . ., except for some aspects” (c) “is dependent regarding . . ., but can perform some aspects” and (d) “is totally dependent regarding . . ..” The minimum score on the total 12 items of the interval scale is 12; the maximum is 48. A high level of functional status results in a high score. The internal consistency as a measure of reliability was good: Cronbach’s alpha was .87 in this sample. Self-Efficacy Regarding AD Use.—The self-efficacy regarding the use of ADs was measured with three items we developed, each presenting increasing barriers. The introduction was “What do you think about yourself in these situations? Assume that you possess an assistive device.” The items were “You want to use the AD when you cannot perform the activity alone,” “You want to use the assistive devices every time if required by the situation,” and “You want to use the AD, even if it is tiring and difficult sometimes.” A 5-point Likert scale ranging from 1 (I know I cannot do it) to 5 (I know I can do it) was used. A total score was calculated, with a high score indicating high self-efficacy. The interval scale ranges from 3 to 15. The reliability of this scale was good: Cronbach’s alpha was .78. Statistical Analysis Intention to Use ADs.—The intention to use ADs was measured with three items we developed, presenting increasing levels of intention. The introduction was “What would you do in the following situations? Assume you possess an assistive device.” The three items were: “Would you use the assistive device when you cannot perform the activity alone?” “Would you use the assistive device every time if required by the situation?” and “Would you use the assistive device, even if it is tiring and difficult sometimes?” A 5-point Likert scale ranging from 1 (I do not have the To test for differences (regarding age, gender, and SES) between nonparticipants—those refusing participation—and participants, we used 2 analyses. In this analysis age was dichotomized in 10-year age groups. The differences between the different normative beliefs and between the perceived consequences were tested with paired-sample t tests to ascertain if the differences between normative beliefs or between perceived consequences could be accepted as real differences, indicating a hierarchy. The differences re42 The Gerontologist Table 1. Sociodemographic Characteristics of the Sample, Flanders, 1999 (N 491) garding awareness and use by specific groups of ADs between persons with and without a specific disability were tested with independent-sample t tests. Levene’s test was used to test for equality of variances. With path analysis, using correlations, partial correlations, and linear regression analysis, the relation was tested between the distal variable awareness of AD, the proximal variables of attitudes, self-efficacy, and subjective norm, and the intention to use ADs. Furthermore, regression analysis tested the relation between the intention to use ADs and functional status (measured with the ADL scale) as independent variables and the use of ADs as the dependent variable. The scale scores of the Intention to Use AD Scale were square root transformed because the data were skewed to the right. The distribution was ameliorated toward a more normal one. These results were the only ones that could benefit from a transformation. All analyses were performed with SPSS for Windows 7.5 (SPSS Inc., Chicago, IL). Statistic Characteristic Gender Women Living Arrangement Alone Only with partner With others Education Until 14 years old or less More than elementary school Age Number of Living Children Functional Status SF-36 subscale ADL scale M SD % 51.9 31.2 53.0 15.9 47.2 52.8 76.2 2.3 5.02 1.93 65.4 41.28 29.3 7.34 Note: SF-36 Short Form–36 (Dutch version); ADL activites of daily living. Results The Sample and only about 3% of variance was explained by SES. Therefore, we conclude that regarding the concepts in the study clustering is minimal or absent. One thousand four hundred nineteen people between 70 and 89 years old, who were living in Flanders, were contacted by letter and were asked by phone to participate in the study (see Methods). The selection criteria were met by 1,019 persons. The main causes of exclusion were having no telephone (38%), having a disease or impairment (21%), being unfindable or unattainable (17%), and being admitted to a care facility or hospital (14%); 528 of the 1,019 persons refused to be interviewed. The main reason for refusal was a lack of interest (34%); many gave no reason for refusing (51%). Four hundred ninety-one persons, or 48.2%, were willing to participate and were interviewed. Age is lacking for the inhabitants of one municipality, as this municipality did not communicate the date of birth. The respondents did not differ significantly from the persons who refused participation regarding SES of the municipality 2(4, N 1019) 6.16, p .05 and age 2(1, N 1004) 0.10, p .05. More women refused to cooperate than men 2(1, N 1019) 4.50, p .05. Of the women who satisfied the inclusion criteria, 54.1% refused; the refusal rate was 47.4% in men. The demographic characteristics of the sample are presented in Table 1. Sample selection in surveys can result in clustering of the respondents. We conducted a series of eight one-way analyses of variance (ANOVA) to check whether there was homogeneity within groups regarding eight key concepts that were due to SES of the municipality (the stratification variable). Four ANOVAs were not significant (possession, use, perceived consequences, and self-efficacy) and the intergroup coefficient omega was below .10, so less than 1% of variance in these key variables was explained by the SES of the municipality. Four ANOVAs were significant (intention, awareness, social norm, and attitude), but omega remained between .15 and .19, Vol. 42, No. 1, 2002 The Possession and Use of ADs The majority of the community-dwelling people aged between 70 and 89 possessed (89.9%) and used (80.6%) at least one of the 32 ADs. More specifically, the respondents possessed between 0 and 21 devices of the 32 ADs that were investigated; the mean number was 3.11 (SD 2.52). Table 2 presents the possession and use for each AD. The respondents used between 0 and 11 ADs of the 32 that were investigated; the mean number was 1.95 (SD 1.69). The absolute nonuse of specific ADs that community-dwelling people already possessed was rather low for most ADs (Table 2). It was highest regarding the cane (22.2% of elderly persons possessed one, but did not use it), the toilet seat (15.9%), and the long shoehorn (15.6%). The relative nonuse, or the proportion of people that possess a specific AD but do not use it out of the total number of persons that possess this AD, varied between 0% and 100% (Table 2). The mean nonuse of an AD people possessed was 51% over the 32 ADs. Regarding going to the toilet and bed transfer, there was a tendency toward the use of more ADs by people with the specific disability (any level of assistance needed in, respectively, going to the toilet and bed transfer, as measured with the respective items of the Katz Scale), compared with persons without this disability (no personal assistance or AD needed; Table 3). After a Bonferroni adjustment for multiple comparisons, the differences between persons with and without a disability were significant only for mobility (independent-sample t test: p .005). The difference regarding a disability in washing was in the opposite direction; persons without this disability tended to use more ADs compared with persons with 43 Table 2. Awareness, Possession, and Use of the Supply of 32 Assistive Devices in Community-Dwelling Elderly Persons, Flanders, 1999 (N 491) Aware Assistive Device Washing Bath seat Bath support rails Sponge on a handle Anti-slip mat Going to Toilet Handle in toilet Commode The raised toilet seat Incontinence material Feeding Special cutlery Dressing Long shoehorn Button hook Elastic shoe laces “Easy shoes” Sock aid Bed-Related Activities Hydraulic hospital bed Free standing lifting pole Bed raiser Adjustable back support in bed Removable bed rail Mobility Chair raiser Lifting seat Hoist Cane Crutch Walking frame Rollator Wheelchair Electric wheelchair Grab rails in the room Doorframe ramp Staircase lift Personal alarm system Possess Use Relative Nonuse % n % n % n 71.6 56.4 47.2 94.3 346 274 228 459 9.5 12.0 14.1 56.7 46 58 68 276 4.3 9.9 11.8 49.5 89.2 91.7 54.5 94.6 430 443 262 455 20.2 22.9 4.6 7.1 97 111 22 34 33.9 163 0.6 95.8 29.5 21.1 54.1 19.8 460 140 101 258 95 90.0 94.6 38.8 64.2 55.9 26.5 16.1 52.6 99.2 98.3 94.8 73.9 99.0 92.9 67.2 32.8 78.0 75.2 % n 21 48 57 241 54.7 17.5 16.3 12.7 25 10 11 35 15.4 7.0 1.5 3.1 74 34 7 15 23.8 69.4 67.4 56.3 23 77 15 19 3 0.0 0 52.0 2.1 1.7 9.0 1.0 250 10 8 43 8 36.4 0.4 1.3 7.3 0.4 175 2 6 35 2 30.0 81.0 41.2 18.9 60.0 75 8 2 8 3 429 452 185 307 267 3.8 3.5 3.1 12.0 2.3 17 17 15 57 11 2.3 0.8 2.5 5.9 0.8 10 4 12 28 4 39.5 77.1 19.4 50.8 65.2 7 13 3 29 7 126 78 252 478 469 454 354 473 444 322 157 373 341 0.2 0.2 0.6 41.7 13.5 4.4 2.8 7.1 0.6 5.9 0.4 0.6 3.1 1 1 3 201 64 21 13 34 3 28 2 3 15 0.0 0.2 0.0 19.5 5.5 0.6 1.5 3.1 0.0 4.6 0.2 0.2 3.1 0 1 0 94 26 3 7 15 0 22 1 1 15 100 0.0 100 53.2 59.3 86.4 46.4 56.3 100 22.0 50.0 66.7 0.0 1 0 3 107 38 18 6 19 3 6 1 2 0 this disability. The difference between persons with and without a disability regarding dressing was not significant. 100 3 People with a disability regarding dressing or washing (any level of assistance needed in, respectively, dressing and washing, as measured with the respective items of the Katz Scale) tended to have a slightly lower level of awareness of the specific groups of ADs compared with persons without this specific disability (no personal assistance or AD needed; independent-sample t test; Table 3). After a Bonferroni adjustment for multiple comparisons, these differences were not significant. An equal level of awareness was found in persons with and without a specific disability regarding mobility, bed transfer, and going to the toilet. Intention to Use ADs The mean intention score was 12.54 (SD 2.19, minimum 4, maximum 15). The intention to use an AD if they could not perform an activity on their own was high: They thought that they would use an AD every time if required by the situation, and they thought they would continue to use the AD, even if it was tiring and difficult (Table 4). Awareness of the Supply of ADs Attitudes Toward the Use of ADs The mean number of ADs people were aware of was 20.51 of the 32 ADs that were presented (mode 19, SD 5.49, minimum 1, maximum 32). Twelve of the 32 ADs were known by about 90% or more of elderly persons (Table 2). Nine ADs were known by less than 50% of the persons. On the basis of the 12-item scale, we found that the attitude toward the use of ADs of community-dwelling elderly adults was positive. The mean total score on the scale was 45.07 (minimum 28, maximum 58; SD 5.45). The item scores are presented in Ta44 The Gerontologist Table 3. Awareness and Use of Assistive Devices in Community-Dwelling Elderly Persons With and Without Specific Disabilities, Flanders, 1999 Persons Without Disability Disability Awareness Mobility Bed transfer Washing Dressing Going to toilet Use Mobility Bed transfer Washing Dressing Going to toilet Persons With Disability Independent Samples t test With Bonferroni Adjustment n M SD n M SD t df 423 467 429 452 7.98 3.23 2.73 2.17 2.26 1.27 1.07 1.24 68 21 61 37 7.88 3.43 2.18 1.76 2.18 1.29 1.35 1.16 0.343 0.712 3.048 1.983 489 486 71.0 487 9 3.56 0.73 1.031 488 481 3.23 0.93 423 467 429 452 0.20 0.04 0.78 0.45 0.48 0.21 0.77 0.57 481 0.25 0.50 68 21 61 37 1.44 0.57 0.48 0.46 Table 4. Intention to Use Assistive Devices (ADs), Subjective Norm Regarding the Use of Assistive Devices, and Self-Efficacy Regarding the Use of Assistive Devices, Flanders, 1999 Item 3.083 488 *p .005. ble 5. The agreement with the opinion that ADs are a good solution to certain problems was high. These community-dwelling elderly people expected increased autonomy when using ADs and believed in the potential of ADs to substitute care. They thought that the use of ADs would make someone less dependent on personal assistance, and they expected that by using ADs they would be able to remain longer at home when they become more in need for care. In general, the elderly persons subscribed to the opinion that care substitution has its limits and the use of ADs could not exclude personal help. Elderly persons valued the control they maintain over their lives when using ADs. The agreement with the belief that if they had an AD they could do things their own way was high. Negative effects were expected to a lesser degree. The belief that the respondents feel handicapped when using ADs was not greatly supported. The respondents rather disagreed with the statement that ADs are developed by people who know nothing about elderly persons. Opinions regarding the effects of the use of ADs on loneliness differed widely. Although the belief that ADs foster loneliness was supported by some, nearly as many disagreed. Willingness to pay for the AD was quite high. Willingness to invest in ADs and agreement with the statement that elderly persons can contribute to the costs of an AD when they need one were rather high. The opinion that they would like to have an AD, provided they do not have to pay for it themselves, did not receive much support. The attitudes were also measured with the scale of perceived consequences of the use of ADs (Table 6). The community-dwelling elderly adults particularly expected increased safety and efficiency. Secondly Vol. 42, No. 1, 2002 M SD Intention 1. I would use an AD if I could not perform an activity on my own 5 4.30 0.89 2. I would use an AD every time if required by the situation 5 4.43 0.79 3. I would continue to use the AD, even if it is tiring and difficult 4 3.82 1.06 Subjective Norm General practitioner 15 12.24 10.24 Children 5 9.44 10.12 Partner 0 8.75 10.26 Nurse 0 3.91 7.98 Friends 0 1.79 5.32 Self-Efficacy 1. I can use an AD when I cannot perform the activity on my own 5 4.05 0.80 2. I can use the AD every time if required by the situation 4 4.16 0.76 3. I can continue to use the AD, even if it is tiring and 4 3.81 0.91 difficult Notes: For Intention scale, 1 I do not have the intention to do this at all; 5 I certainly have the intention to do this. For Self-Efficacy scale, 1 I know I cannot do it; 5 I know I can do it. 0.95 10.550* 72.5 0.93 2.645 20.1 0.70 2.948 488 0.61 0.128 487 9 0.78 0.67 Mode Median they expected autonomy, followed by speed of performance. Care substitution and stigmatization were least expected. The differences between all pairs of means were significant (paired-sample t test: p .001), except for the difference between stigmatization and care substitution, and between efficiency and safety. Safety and autonomy were both valued most highly in the context of the performance of ADLs. Efficiency was the second most valued. The elderly persons valued to the same degree care substitution, stigmatization, and speed of performance. The differences between all pairs of means were significant (paired-samples t test: p .01), except for the difference between autonomy and safety and between stigmatization, care substitution, and speed of performance. The product of the degree to which one agrees with a perceived consequence on the one hand and the personal valuation of the importance of this consequence on the other is a measure of the importance of each attitude for this person (Table 6). The most important attitude was the perceived increase in safety, and the second most important attitudes were autonomy and efficiency. Less important attitudes are, in decreasing order, increased speed of performance, care substitution, and stigmatization. The differences between all pairs of means were significant (paired-samples t tests: p .01), except for the difference between autonomy and efficiency. The mean total score was 95.14 (minimum 29, maximum 150; SD 20.18). 45 Table 5. Attitudes Toward Assistive Device (AD) Use in Community-Dwelling Elderly Persons: Mode, Mean, and Standard Deviation of the Items of the Attitude Scale, Flanders, 1999 Item 1. By using ADs I would be able to remain longer at home when I become more in need for care 2. The use of ADs could not exclude personal help 3. I would dare to rely totally on ADs 4. The use of ADs would make someone less dependent on personal assistance 5. I would like to have an AD, provided I do not have to pay myself 6. ADs foster loneliness 7. Elderly persons can contribute to the costs of an AD when they need one 8. ADs are a good solution to certain problems 9. If I had an AD I could do things my own way 10. ADs are developed by people who know nothing about elderly persons 11. I feel handicapped when using certain ADs 12. I am willing to invest in ADs Mode M SD 5 4 4 5 2 4 4 5 5 2 4 4 4.57 3.52 3.51 4.46 2.52 2.87 3.62 4.72 4.35 2.32 3.06 4.07 0.74 1.11 1.27 0.80 1.24 1.27 1.01 0.51 0.82 1.07 1.31 1.05 Note: 1 totally disagree; 5 totally agree. Subjective Norm Regarding the Use of ADs reliable scale, the AADS. Moreover, significant relations were expected between the proximal variables and the more distal variable awareness of ADs. Finally, a relation between the intention to use ADs and functional status on the one hand and the use of ADs on the other was expected. After standardization of the raw scores in z scores of all variables that were included in the model to be tested, the relations were tested with two regression analyses and three correlations. The beta and p values are presented in Table 7. Awareness was found to be significantly correlated to attitudes (Pearson r .17, p .001) and subjective norms (Pearson r .16, p .001), but not to self-efficacy. Direct effect of awareness on the intention to use ADs was investigated with the partial correlation between awareness and intention, The general practitioner was mentioned most often (by 63.4% of the respondents) as the person whose opinion is valued regarding the use of ADs, closely followed by the children (52.8%) and the partner (47.2%). Nurses (21.3%) and friends (12.3%) were mentioned less. Only 4.1% of the elderly persons mentioned that they did not value anybody’s opinion regarding the use of ADs. The product of the perceived strength of a significant other’s expectation to use ADs on the one hand and the motivation to comply with this expectation on the other is a measure of the strength of a normative belief for this person. The most important normative belief concerned the general practitioner, and the second most important normative belief concerned the children and the partner (Table 4). In order of importance, the other important normative beliefs concerned the nurse and friends. The differences between all pairs of means were significant (paired-samples t test: p .0005), except for the difference between the normative beliefs regarding the partner and the children. The mean subjective norm was 36.13 (SD 23.41). Table 6. Perceived Consequence of Assistive Device (AD) Use, Valuation of the Perceived Consequence, and Attitude Toward AD Use in Community-Dwelling Elderly Persons, Flanders, 1999 Consequence Perceived Consequence Safety Efficiency Autonomy Speed of performance Care substitution Stigmatization Valuation of Perceived Consequence Safety Efficiency Autonomy Speed of performance Care substitution Stigmatization Attitude Safety Efficiency Autonomy Speed of performance Care substitution Stigmatization Self-Efficacy Regarding the Use of ADs The mean self-efficacy score, obtained with the three-item scale, was 12.02 (SD 2.06, minimum 3, maximum 15). Elderly persons believe they can use an AD when they cannot perform the activity on their own, they believe they can use the AD if this is required by the circumstances, and they believe they can continue to use the AD, even if it is tiring and difficult (Table 4). A Model of the Use of ADs The model was tested with path analysis, testing significant and nonsignificant relations. Specific, significant relations were expected between the proximal variables of attitude, subjective norm, and selfefficacy on the one hand and intention to use ADs on the other. The attitude was measured with the most Mode M SD 5 5 5 4 3 4 4.49 4.45 4.16 3.96 3.01 3.00 0.77 0.71 1.04 1.02 1.23 1.32 5 5 5 4 4 5 4.75 4.36 4.77 3.31 3.48 3.35 0.61 0.81 0.60 1.27 1.33 1.51 21.41 19.59 20.00 13.39 10.91 9.81 4.76 5.22 5.81 6.61 6.61 6.29 Notes: For perceived consequence: 1 totally disagree; 5 totally agree. For valuation of perceived consequence: 1 very unimportant; 5 very important. 46 The Gerontologist controlled for attitude, subjective norm, and self-efficacy. It was expected to be nonsignificant and proved to be so. All expected relations within the core of the model were confirmed, although the relationship was often small. It is important to note that the scores on the functional-status instrument and the measure of the use of ADs are positively related, which means that the concepts are negatively related. The significant relations are presented in Figure 1. helpful, not only to compensate for disabilities, but also to increase efficiency and safety in nondisabled persons. Even more problematic is the finding that the presence of specific disabilities, which is an obvious reason to use ADs, did not result in the use of more of the relevant ADs, except regarding mobility. Another topic of concern is the extent of the nonuse of ADs that elderly persons already possess. The proportion of the number of people that possess a specific AD but do not use it varies considerably from one AD to another. It can be hypothesized that the nonuse of canes, toilet seats, and long shoehorns can be attributed to the fact that these cheap ADs are passed on from one generation to the next and are stored until a person becomes disabled and needs them. But more expensive ADs are not used either, which is a waste of resources. Examples are the hydraulic hospital bed, the walking frame, the rollator, the manual wheelchair, the electric wheelchair, and the staircase lift. This finding suggests the need for improvement of follow-up. Although this finding is based on a small numbers of persons, we found nonuse for each of the 32 ADs, which supports the reliability of this result. The presence of disabilities is an obvious reason to use ADs: The data show that persons with disabilities regarding mobility, bed transfer, and going to the toilet used more ADs specific to these activities than did persons without these disabilities. Before people can develop an attitude toward ADs, they need to be aware of them. This study showed a good awareness in the community-dwelling elderly regarding the existence of ADs that have been available for decades. These relate to mobility and transfer: the wheelchair, the parrot, the cane, and crutches; for washing: the anti-slip mat; regarding dressing: the long shoehorn; regarding going to the toilet: the toilet chair and incontinence material. Many people also possess these ADs. But some ADs regarding mobility that have been developed more recently were also known to most elderly persons, for example, the hydraulic hospital bed, a grip in the toilet, and the walking frame. This may be due to their presence in hospitals and rehabilitation centers. They are also well known to home nurses. But a lower level of awareness of 21 other ADs was found. These ADs may nevertheless increase the comfort and safety of elderly persons, and consequently their quality of life. Moreover, it was expected that persons with a disability would be more aware of the specific AD that would be relevant to increase their functional level than would persons Discussion In this study we described the possession and use of ADs for self-care and mobility in a random sample of community-dwelling elderly persons between 70 and 89 years old in Flanders. We were especially interested in the ability of psychological variables to explain the use of ADs. Knowing them would suggest measures to be developed that could probably improve the introduction process of ADs. The concepts of the theory of planned behavior were operationalized in the context of AD use. We described the possession and use of ADs in community-dwelling elderly persons as they are related to the intention to use ADs, awareness of ADs, attitudes toward AD use, the subjective norm regarding AD use, and selfefficacy concerning AD use. We found that the use of ADs was indeed partially predicted by the intention to use ADs when controlling for functional status, and that the intention to use ADs was explained by the concepts of the model. Therefore, this study suggests that a social–cognitive model of human behavior can be helpful in explaining the use of ADs. The findings are discussed in the following section. Because the study focused on ADs for self-care and mobility in the general elderly population, almost all ADs were low tech, and many had already existed for quite a long time. The use of ADs was not exceptional. Almost all community-dwelling persons between 70 and 89 years possessed and used at least one of the 32 ADs for self-care and mobility. Elderly persons possess and use ADs especially regarding taking a bath and mobility. Notwithstanding these positive findings, the study revealed that many ADs were not possessed and used by a substantial proportion of elderly persons. ADs that have been developed more recently in particular (e.g., the personal alarm system, a device to pull on socks) were not possessed by the majority of elderly persons between 70 and 90 years of age, although these ADs can be Table 7. Path Analysis of Assistive Devices (ADs) Use (N 491) Model Dependent variable Independent variable 1 R2 .310 F(3,468) 71.455*** Intention to use ADs 2 R2 .159 F(2,478) 46.281*** The use of ADs Attitudes (AADS) Self-efficacy Subjective norm Functional status Intention to use ADs .170*** .469*** .091* .398*** .121** Regression Note: AADS Attitudes Toward Assistive Devices Scale. *p .05; **p .01; ***p .001. Vol. 42, No. 1, 2002 47 Figure 1. A social cognitive model of assistive device (AD) use in older persons. without this disability, as the acquisition of a disability would trigger a search for information. This was not confirmed by the data. The findings suggest an equal level of information regarding all groups of ADs in people with specific disabilities. In accordance with other studies, we found that the attitude of community-dwelling elderly people was rather positive toward the use of ADs. A clear order emerged. The finding that elderly persons do not use ADs solely with the aim of increasing autonomy (Sonn & Grimby, 1994) was confirmed. The elderly persons in particular expected safety and efficiency as consequences of AD use, even more than, in descending order, autonomy and speed of performance. They also valued safety in the context of ADL highest, together with autonomy. They valued efficiency as secondary. The elderly persons expected that AD use would substitute personal care at home and substitute intramural care by home care. But they thought that the substitution of care is limited and that AD use would not preclude personal help. For health care workers introducing ADs, the consequences that should be emphasized are not only the increase in autonomy but the increased efficiency and safety, too. Because people expect these consequences and value them highly, these aspects of ADs can facilitate their introduction. The effect on care substitution and on stigmatization should be discussed in greater depth with disabled persons, but with some caution as negative attitudes can be expected, and in general these consequences are valued less. Increased feelings of loneliness when using ADs because of the decreased help from other people were expected by some. This negative aspect of ADs deserves further consideration. The subjective norm regarding the use of ADs was found to be a small yet significant determinant of the intention to use ADs. General practitioners and relatives are the most important significant others. Their willingness and cooperation was found to be important for successful introduction of ADs. Data analysis was improved compared with preceding studies by using techniques that controlled for other variables. Social cognitive variables were orga- nized in a social cognitive model of AD use. The previously developed model to describe the use of ADs (Roelands et al., in press) was modified in accordance with the theory of planned behavior and proved to be of some significance. Functional status was confirmed as an important determinant of the use of ADs, which is evident because it is a prerequisite for using ADs. The model on the psychological variables increases the variance in the use of ADs that is explained. Intention to use ADs was determined especially by self-efficacy regarding the use of ADs, but the relative importance of the variables should be considered fully, as we discuss below. Intention to use ADs was also determined by the attitude toward the use of ADs and to a lesser extent by the subjective norm regarding the use of ADs. The role of awareness regarding ADs and attitudes toward the use of ADs were confirmed. But the partial correlations showed that because of the inclusion of more proximal factors in the model, awareness of ADs became a more distal factor and was positively correlated with the attitude toward the use of ADs and the subjective norm. The psychological variables in the model predicted a moderate amount of variance in intention to use ADs, but intention to use ADs and functional status predicted a rather small amount of variance in use of ADs. We can hypothesize that in use of ADs factors at other levels are involved, too. Some will act as barriers, others as facilitators. Some of these factors can be summarized as network factors, for example, aspects of AD supply by health care services (e.g., costs, sensitivity of needs screening, quality of AD introduction process, extent and quality of supply), and availability of personal help regarding ADLs and IADLs. The model presented in this article focuses only on psychological factors involved in use of ADs. The predictive power of other factors should be evaluated in intervention studies and studies comparing health care systems to allow further optimalization of AD use by older persons. A complex sample-selection procedure was used to obtain a representative distribution of age, sex, and SES (at municipal level). It is very likely that the sampling strategy resulted in a representative sample 48 The Gerontologist of elderly persons living in Flanders to be contacted. Nevertheless, some results suggest some selection bias in the final sample in favor of men. The gender ratio in Flanders in this age group is 57.7% women and 42.3% men, as it is 51.9% women and 48.1% men in the sample of respondents. Three mechanisms may be involved. First, as lack of a telephone is probably related to SES and a correlation has been established between being an older woman and low SES, perhaps more women were excluded. Second, it is well known that disabilities are more common in women. Third, more women refused to cooperate, which may be because of the procedure, because most of the telephone calls to obtain consent were done by a male researcher, although it was mentioned that the actual interviewer would be a woman. Nevertheless, the deviation from the distribution of the person’s sex in the population was rather small, and effects on the descriptive data are therefore minimal. The representativeness of the sample regarding physical functioning can be estimated by comparing the SF-36 subscale score of the people aged between 75 and 89 in this study (N 290) with a representative sample of community-dwelling and institutionalized people aged 75 and older in the Belgian National Health Interview Survey (BNHIS; N 195). When people are not disabled in any of the 10 types of activities, they have a score of 100, the minimum score is 0. The mean level of functioning is similar between the two samples (BNHIS: M 57.8; this study: M 59.6). A one-sample t test did not reveal any significant differences between this sample and the BNHIS, t(289) 1.027. The comparison of functional status with another representative sample, and the comparison of age and SES of the municipality between respondents and persons who refused, indicate that the respondents are a representative sample of the communitydwelling persons between 70 and 89 years old in Flanders, taking into account the exclusion criteria, most of which are common in surveys. Many instruments were developed for this study to measure the concepts that were introduced. The PCAD, the SEAD, and the IAD were sufficiently reliable, although there was some room for improvement. The reliability of the instruments and the relations in the model could increase if some adaptations were done. First, the beliefs were specified as related to devices for self-care and mobility. This was sufficient for this study as the model was confirmed. It is, however, likely that a person’s self-efficacy, perceived consequences and intention to use depend on the specific AD. A second reliability issue relates to the evaluation component of the attitude beliefs. As discussed by Young, Lierman, Powel-Cope, Kasprzyk, and Benoliel (1991, p. 142) regarding the theory of planned behavior, “in asking a respondent to evaluate an item . . . one assumes that the respondent holds an absolute opinion about that belief and that it is possible to make this evaluation acontextually.” It is, however, likely that this opinion depends on related factors such as the activity and the AD concerned. For example, “to perform activities indeVol. 42, No. 1, 2002 pendently with an AD” can be considered very unimportant if it regards self-care, but very important regarding mobility. A third reliability issue is especially important in lower educated elderly people. It has already been mentioned (Mullen, Hersey, & Iverson, 1987; Young et al., 1991) that the format of the attitude questionnaire can pose a conceptual challenge to participants. The evaluation items are similar to the belief items and the respondents may perceive them as the same, resulting in feelings of frustration. With some respondents, we had the impression that they fell back on response sets. In general, the intention of elderly persons to use ADs was high, but some caution is required. The high level of intention may also be due to a socially desirable response. Although showing a positive attitude regarding the use of ADs was always avoided, and this was strongly stressed during the training of the interviewers, it is not entirely impossible that the presentation of the study or the items was not totally neutral. The socially desirable response, if and to the degree it occurred, would also decrease the relationships in the model. Some caution is necessary regarding the interpretation of the correlations between variables and their determinants and consequently regarding the interpretation of the relative weights of the determinants to predict a variable (the s). In the theory of planned behavior the correspondence between the variables regarding action, target, context, and time elements is essential to be able to predict behavioral intentions and behavior (Ajzen & Fishbein, 1980, p. 56). It seems to us that the rather low betas in the model on the use of ADs can be increased by improving the correspondence between the measures. Where correspondence was high, so too was the self-efficacy (intention correlation) beta. Although it has been mentioned that this high correlation between two successive similar scales that follow one another in the questionnaire may be due to a response set, the different response categories slightly reduce the effect. An improvement in the correspondence would also challenge the relative importance of the variables in predicting the intention to use ADs. Moreover, an improved correspondence would probably increase the predictive validity of the intention to use ADs. The correspondence could be increased by being more specific regarding the context of using ADs (e.g., when one is alone, when it is free of charge), the target (e.g., to increase autonomy or safety or efficiency), the time elements (every time when it is required or sometimes), and also regarding the object of the behavior under consideration. We assume that the relative weights of the concepts differ depending on the specific AD. The low number of ADs in the subscales related to AD use and the dichotomous nature of the individual items limited the internal consistency of the scales. The finding that AD use is related to presence of specific disabilities requires confirmation with larger subscales. A social–cognitive model of the use of ADs was empirically tested in community-dwelling elderly persons. Finally, the validity and usefulness of this 49 model should be evaluated in intervention studies, which could develop and evaluate protocols to introduce ADs in a more structured way. We expect that interventions meant to increase the intention to use ADs and the use of ADs, when required by the functional status and wanted by the person with disabilities, to increase their effectiveness, should take these variables into consideration. These concepts and the model could also be evaluated in specific populations (e.g., persons with arthritis or dementia, persons living in institutions) and for specific ADs. Aronson (Eds.), The handbook of social psychology, (Vol. 2, pp. 233– 346). New York: Random House. Mullen, P., Hersey, J., & Iverson, D. (1987). Health behavior models compared. Social Science & Medicine, 24, 973–981. Povey, R., Conner, M., Sparks, P., James, R., & Sheperd, R. (2000). Application of the theory of planned behaviour to two dietary behaviours: Roles of perceived control and self-efficacy. British Journal of Health Psychology, 5, 121–139. Rhodes, R. E., Martin, A. D., Taunton, J. E., Rhodes, E. C., Donnelly, M., & Elliot, J. (1999). Factors associated with exercise adherence among older adults: An individual perspective. Sports Medicine, 28, 397–411. Roelands, M., Van Oost, P., Buysse, A., & Depoorter, A. M. (in press). Awareness of assistive devices for mobility and self-care in community-dwelling elderly and attitudes towards their use. Social Science and Medicine. Roelands, M., Van Oost, P., & Depoorter, A. M. (1999). Psychometric characteristics of the Attitudes toward Assisitive Devices Scale (AADS) (Internal document). Ghent, Belgium: Ghent University. Rogers, J., & Holm, M. (1992). Assistive technology device use in patients with rheumatic disease: A literature review. The American Journal of Occupational Therapy, 46, 120–127. Scherer, M., & McKee, B. (1989). But will the assisitive technology device be used? Proceedings of the 12th Annual Conference: Technology for the Next Decade (pp. 356–357). Washington, DC: RESNA. Sonn, U., & Grimby, G. (1994). Assistive devices in an elderly population studied at 70 and 76 years of age. Disability and Rehabilitation, 16, 85–92. Stronegger, W., Freidl, W., & Rasky, E. (1997). Health behaviour and risk behaviour: Socioeconomic differences in an Australian rural county. Social Science and Medicine, 44, 423–426. Thomas, I., & Vanneste, D. (1997). Ruimtelijke onevenwichten in de Belgische huisvesting op intraregionale en intrastedelijke schaal [Spatial imbalances in Belgian housing at an intraregional and intraurban scale]. Tijdschrift van de Belgische Vereniging voor Aardrijkskundige Studies, 1, 7–42. World Health Organization. (1980). International classification of impairments, disabilities and handicaps. Geneva, Switzerland: Author. Young, H., Lierman, L., Powel-Cope, G., Kasprzyk, D., & Benoliel, J. (1991). Operationalizing the theory of planned behavior. Research in Nursing and Health, 14, 137–144. References Aaronson, N., Muller, M., Cohen, P., Essink-Bot, M., Fekkes, M., Sanderman, R., Sprangers, M., te Velde, A., & Verrips, E. (1998). Translation, validation, and norming of the Dutch language version of the SF-36 Health Survey in community and chronic disease populations. Journal of Clinical Epidemiology, 51, 1055–1068. Ajzen, I. (1991). The theory of planned behaviour. Organizational Behaviour and Human Decision Processes, 50, 179–211. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting behavior. Englewood Cliffs, NJ: Prentice-Hall. Ajzen, I., & Madden, T. J. (1986). Prediction of goal-directed behaviorattitudes, intentions, and perceived behavioral control. Journal of Experimental Social Psychology, 22, 453–474. Armitage, C., & Conner, M. (2000). Social cognition models and health behaviour: A structured review. Psychology and Health, 15, 173–189. Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman. Bish, A., Sutton, S., & Golombok, S. (2000). Predicting uptake of a routine cervical smear test: A comparison of the health belief model and the theory of planned behaviour. Psychology and Health, 15, 35–50. Conner, M., Warren, R., Close, S., & Sparks, P. (1999). Alcohol consumption and the theory of planned behavior: An examination of the cognitive mediation of past behavior. Journal of Applied Social Psychology, 29, 1676–1704. de Klerk, M., & Huijsman, R. (1993). Ouderen en het gebruik van hulpmiddelen. Een marktbehoefte-onderzoek [Elderly persons and use of assistive devices. A study on market needs]. Rotterdam, The Netherlands: Erasmus University Rotterdam. de Klerk, M., & Huijsman, R. (1995). KITTZ-Randstadproject Thuiszorgtechnologie: Effectmeting [KITTZ-Randstadproject home care technology: Effects]. Rotterdam, The Netherlands: Ersasmus University Rotterdam. Deliens, L. (1998). Sociale determinanten van kennis van kankerpreventie [Social determinants of knowledge about cancer prevention]. Unpublished doctoral dissertation. Demarest, S., Van Oyen, H., Leurquin, P., Tafforeau, J., Tellier, V., & Van der Heyden, J. (1998). De gezondheid van de bevolking in België, Gezondheidsenquête, België, 1997 [Population health in Belgium. Health Survey, Belgium, 1997]. Brussels, Belgium: Centrum voor Operationeel Onderzoek in Volksgezondheid, Wetenschappelijk Instituut Volksgezondheid—Louis Pasteur. Gagnon, M., & Godin, G. (2000). The impact of new antiretroviral treatments on college students’ intention to use a condom with a new sexual partner. AIDS Education and Prevention, 12, 239–251. Gandek, B., Ware, J., Aaronson, N., Alonso, J., Apolone, G., Bjorner, J., Brazier, J., Bullinger, M., Fukuhara, S., Kaasa, S., Leplège, A., & Sullivan, M. (1998). Tests of data quality, scaling assumptions, and reliability of the SF-36 in eleven countries: Results from the IQOLA project. Journal of Clinical Epidemiology, 51, 1149–1158. Katz, S., & Akpom, C. (1976). A measure of primary sociobiological functions. International Journal of Health Services, 6, 493–507. LaPlante, M., Hendershot, G., & Moss, A. (1992). AT devices and home accessibility features: Prevalence, payment, need, and trends. Hyattsville, MD: National Center for Health Statistics. Legare, F., Godin, G., Guilbert, E., Laperriere, L., & Dodin, S. (2000). Determinants of the intention to adopt hormone replacement therapy among premenopausal women. Maturitas, 34, 211–218. Lynch, J., Kaplan, G., & Salonen, J. (1997). Why do poor people behave poorly? Variation in adult health behaviours and psychological characteristics by stages of the socioeconomic lifecourse. Social Science and Medicine, 44, 809–819. Mann, W., Hurren, D., & Tomita, M. (1993). Comparison of assistive device use and needs of home-based older persons with different impairments. The American Journal of Occupational Therapy, 47, 980–987. McGuire, W. (1985). Attitudes and attitude change. In G. Lindsay & E. Received October 19, 2000 Accepted September 10, 2001 Decision Editor: Laurence G. Branch, PhD Appendix Perceived Consequences of the Use of Assistive Devices (AD) Scale Using AD has some consequences. To which degree do you agree with the following statements? Totally agree 5 Rather agree 4 Don’t know 3 Rather disagree 2 Totally disagree 1 • By using AD I can do things independently and I don’t need help. • By using AD other people will think I am old and sick. • By using AD it goes faster. • By using AD it goes easier. • By using AD the nurse or home carer doesn’t visit me anymore. • Using AD gives me a sense of security. To what extent is it important to you that . . . Very important 5 Rather important 4 Not important, not unimportant 3 Rather unimportant 2 Very unimportant 1 • you can do activities independently and you don’t need help? • other people do not think you are old and sick? • an activity goes fast? • an activity goes easy? • the nurse or home carer doesn’t visit you anymore? • you have a sense of security? 50 The Gerontologist
© Copyright 2026 Paperzz