REVIEW OF FORENSIC ASSESSMENT INSTRUMENTS

Clinicians who are asked to evaluate an adult's capacity to care for self or financial resources will likely rely on four sources of standardized assessment data:

• independent activities of daily living (IADL) rating scales

• specific guardianship or conservatorship instruments

• neuropsychological or cognitive testing

• mental health diagnostic interviews and scales.

It is beyond the scope of this chapter to review procedures for neuropsy­chological testing and diagnostic assessment, although a few points on these instruments will be made.

As noted, neuropsychological assessment is likely to be a key compo­nent of assessments of capacities for independent functioning. They help to describe the extent of cognitive and decisional impairments, may assist in differential diagnosis and prognosis, and may help to clarify the link between disabling conditions and functional outcomes and between personal deficits and environmental resources. Furthermore, the bulk of referrals for assessments regarding guardianship will be for elderly adults and the majority of these will be for adults with dementing illnesses. Since deficits in memory and cognition are the hallmark of dementing illness, an assessment of these deficits is key. Neuropsychological assessment is discussed by Lezak (1995) and LaRue (1992) and others.

In most cases a brief screening of cognition (e.g., Mini-Mental Status Exam) would not be sufficient to be informative, but a full neuropsycholog­ical battery will not always be necessary. Evaluators may wish to select spe­cific tests that are likely to be informative about the specific capacity in question (e.g., a test of arithmetic when financial capacities are questioned) and about the disabling condition presented (e.g., tests of memory when the individual presents with dementia). Visual-spatial problem solving and visual memory appear to be especially predictive for some tasks of every­day functioning (Isaac & Tambly, 1993; Palmer & Dobson, 1994; Richardson, Nadler, & Malloy, 1995).

A large number of instruments have been developed to assess the mental or psychological capacities of elderly individuals, as well as symp­toms of psychopathology in the elderly (see Lichtenberg, 1999). These are not reviewed here merely because they do not focus on functional abilities of everyday life, but rather on such general constructs as depression, anx­iety, delusions, and behavioral self-control. Nevertheless, many of them may be of considerable benefit in examinations in guardianship cases, in order to obtain information with which to provide psychological explana­tions for functional deficits.

The remainder of this chapter reviews instruments developed to assess everyday functioning, including IADL rating scales and guardian­ship or conservatorship instruments. The latter are more rightly consid­ered FAI’s, but IADL instruments are close and may be especially useful in forensic evaluations.

Regarding IADL rating scales, a number of scales have been designed to assess independent activities of daily living (some including sections on activities of daily living and cognitive abilities) by means of patient or informant report. These scales are useful in organizing assessments of and reports about functioning and are appropriate when there is good evi­dence that an informant's report is reliable or when the examiner has observed the patient completing tasks in the home or residential setting.

IADL scales selected for inclusion in this chapter met certain specific criteria. They:

• employed a multidimensional approach;

• focused on instrumental or higher-order activities of daily living;

• had multiple studies or citations; and

• were appropriate for use with older adults.

Instruments that assess and describe only the most basic ADL functions probably will be of limited importance in forensic assessments for guardian­ship cases, for which reason they were not selected for review. Examiners who do have a need for them may choose from several instruments with adequate conceptual and empirical backgrounds, such as the Katz Index of ADL (Katz et al., 1963) or the Barthel Index (Mahoney & Barthel, 1965).

Several promising scales that were not reviewed (because they did not meet certain inclusion criteria) deserve special mention for completeness:

• Assessment of Living Skills and Resources (ALSAR; Williams et al., 1991) « Comprehensive Assessment and Referral Evaluation (CARE; Gurland

et al., 1977) (assessing situational and environmental resources)

• Performance Assessment of Self-Care Skills (PASS; McCue, Rogers, & Goldstein, 1990)

• Structured Assessment of Independent Living Skills (SAILS; Mahurin, DeBettignies, & Pirozzolo, 1991).

In addition, the chapter does not review occupational therapy instru­ments such as the Kohlman Evaluation of Living Skills (KELS; McGourty, 1979) and the Functional Independence Measure (FIM; Research Foundation, 1987), because they tend to be developed for and used by occupational therapists only.

In the remainder of the chapter, nine instruments are reviewed under three headings referring to their type:

A. 3 IADL instruments based on interview

B. 2IADL instruments based on performance, and

C. 4 instruments designed specifically for guardianship and conser­vatorship evaluations.

A. IADL INSTRUMENTS BASED ON INTERVIEW OR OBSERVATION

Adult Functional Adaptive Behavior Scale (AFABS)

Author

Philip S. Pierce

Author Affiliation

Department of Veterans Affairs, Togus, ME

Primary Reference

Pierce, P. S. (1989). Adult Functional Adaptive Behavior Scale (AFABS): Manual of Directions (1989 Edition). Togus, ME: Author.

Description

The Adult Functional Adaptive Behavior Scale (AFABS) was developed to assist in the assessment of ADL and IADL functions in the elderly to evaluate their capacity for personal responsibility and the matching of a client to a placement setting (c).

The AfABS consists of 14 items. Six items rate ADL's: eating, ambula­tion, toileting, dressing, grooming, and managing (keeping clean) personal area. Two items tap IADL's: managing money and managing health needs. Six items tap cognitive and social functioning: socialization, envi­ronmental orientation (ranging from able to locate room up through able to travel independently in the community), reality orientation (aware of person, place, time, and current events), receptive speech communication, expressive communication, and memory. Items are rated on four levels: 0.0 representing a lack of the capacity, 0.5 representing some capacity with assistance, 1.0 representing some capacity without assistance, and 1.5 rep­resenting independent functioning in that area. Individual scores are summed to receive a total score in adaptive functioning.

The AFABS assesses adaptive functioning through interviewing an informant well acquainted with the functioning of the individual in ques­tion. The informant data is combined with the examiner's observation of and interaction with the client to arrive at final ratings.

The AFABS is administered in a semi-structured format, after client observation and medical record review. During the administration the examiner asks the informant about each area of functioning, first asking a general question as written on the rating form (e.g., "does he eat totally by himself or does he require assistance with preparing and eating meals?"). Follow-up questions are then asked on the basis of clinical judgment until the exact level of performance can be rated (c).

Conceptual Basis

conceptual definition. The AFABS was based on Heber's (a) con­struct of adaptive behavior developed for the developmentally disabled. Heber proposed that adaptive behavior has two components: being inde­pendent in personal function/maintenance and meeting societal rules for personal responsibility. The AFABS focuses on the first of these as a prerequisite for being able to have personal responsibility (d).

operational definition. Specific items were developed through a review of journal articles, texts, and patient records as well as clinical experience in the fields of developmental disabilities, mental retardation, and psychogeriatrics. Items were piloted and revised in consideration of clarity, content coverage, ease of administration, and age-appropriateness

(c). Kerby, Wentworth, and Cotten (b) note that the AFABS is well suited for older clients in contrast to other adaptive behavior scales.

critique. A strength of the AFABS is its conceptual basis in a theory of adaptive behavior which is modified for use with the elderly. The AFABS may be well suited to answer questions for evaluations concerning level of placement and requisite supervision.

Psychometric Development

standardization. The AFABS manual provides description of each of the 14 content areas and examples for ratings on each of four levels. A rating form provides a place to rate each item and to note problem areas (e.g., hearing impairment) and daily (e.g., "sundowning") or monthly (e.g., delirium) mental status changes.

reliability. The Cronbach's alpha reliability coefficient was.95 and the Spearman-Brown and Guttman split-half reliability coefficients were both.96 (d) in a mixed sample (n = 432). Similar internal consistency reli­ability was found in a sample of psychiatric inpatients (n = 91), yielding an alpha coefficient of.93 and split half coefficient of.97.

norms. Mean score and standard deviation are available for 25 community-dwelling elderly adults who served as a control sample in one study (d).

cRiTiQuE. Internal consistency reliability is excellent. Additional normative data are needed. The availability of a clear manual and test rat­ing form is a plus. The scale is intended as a semi-structured interview, and as such, does not provide standardized items for asking specific ques­tions or rating clients' performance. This suggests that adequate training and experience may be necessary to obtain accurate estimations of rat­ings. Studies of inter-rater reliability would help to establish the extent to which it is possible for independent raters to agree given this format.

Construct Validation

In a factor analytic study of 432 adults from various settings, the AFABS yielded one factor with an eigenvalue greater than 1.00 (d).

Kerby, Wentworth, and Cotten (b) examined the relationship between the AFABS and two other adaptive behavior scales. They found signi­ficant correlation between the AFABS and other scores (r =. 72 and.85), providing support for the construct validity of the AFABS.

critique. These studies support the author's contention that the AFABS measures a unidimensional construct, adaptive behavior. Addi­tional studies comparing the AFABS with ADL and IADL scales devel­oped for use with elderly populations would further support its construct validity with older adults.

Predictive or Classificatory Utility

In a study of predictive utility, the relationship between the AFABS total score and a 1-6 "level of placement" was determined (1-5 = inpa­tient wards graded in their level of supervision and 6 = community senior housing) for 25 community dwelling elderly adults and 126 residents of a multi-level state mental health institution (d). The Spearman rho coeffi­cient between the AFABS total score and placement level was statistically significant (r =.86). An ANOVA comparing the mean score at each of six levels of placement was also statistically significant, and mean score dif­ferences on the AFABS between each level of placement were significant in post-hoc tests.

A second study of predictive validity utilized similar methods to compare AFABS total score and a 1-3 "level of placement" for 91 psychi­atric inpatients (e). The Spearman rho coefficient was statistically signifi­cant (r =.71) as were ANθVA and post-hoc tests comparing mean AFABS scores at each of three levels of placement.

critique. These studies suggest that the AFABS can be useful in pre­dicting the level of supervision required by adults.

Potential for Expressing Person-Situation Congruency

In that the AFABS begins with a review of history and ends with a review of special problem areas, it would allow for interpretation of adap­tive behaviors within a personal and situational context. However, there are not specific items built into the AFABS item ratings to assess person­situation congruency.

References

(a) Heber, R. F. (1961). A manual of terminology and classification in mental retardation (2nd ed.). American Journal of Mental Deficiency. Monograph Supplement.

(b) Kerby, D. S., Wentworth, R., & Cotten, P. D. (1989). Measuring adaptive behavior in eld­erly developmentally disabled clients. The Journal of Applied Gerontology, 8, 261-267.

(c) Pierce, P. S. (1989). Adult Functional Adaptive Behavior Scale (AFABS): Manual of Directions (1989Edition). Togus, ME: Author.

(d) Spirrison, C. L. & Pierce, P. S. (1992). Psychometric characteristics of the Adult Functional Adaptive Behavior Scale (AFABS). Gerontologist, 32, 234-239.

(e) Spirrison, C. L., & Sewell, S. M. (1996). The adult functional adaptive behavior scale (AFABS) and psychiatric inpatients: Indices of reliability and validity. Assessment, 3, 387-391.

Multidimensional Functional Assessment Questionnaire (MFAQ)

Authors

Older Americans Resources and Services (OARS) Project

Authors' Affiliation

Center for the Study of Aging and Human Development, Duke University Medical Center, Durham, NC

Primary Reference

Center for the Study of Aging and Human Development (1978). Multidimensionalfunctional assessment: The OARS methodology. Durham, NC: Duke University.

Description

The Multidimensional Functional Assessment Questionnaire (MFAQ) was developed to provide a reliable and valid method for characterizing elderly individuals and for describing elderly populations, in ways that would be "useful to clinicians... to program analysts, to resource alloca­tors, and to research scientists in a variety of disciplines" (i, p. 4). The MFAQ supersedes the nearly identical Community Survey Questionnaire (CSQ, a predecessor which also was developed by the Duke Center). Both instruments frequently have been called the OARS, in reference to the program that developed the instrument throughout the 1970s. The MFAQ or the CSQ was already in use by well over 50 service centers, researchers, or practitioners nationally when the MFAQ was published (1978).

The MFAQ is a structured interview with 72 items (105 questions) requiring about one hour to administer. Part A deals with the examinee's functioning, and Part B elicits information about the services recently used by the examinee.

Part A provides information in five areas of functioning (m, pp. 68-70):

• Social resources: extent, quality, and availability of social interac­tions (e.g., whom one lives with, in whom one confides, who is willing to provide care in case of illness).

• Economic resources: financial ability to meet needs and obtain serv­ices (e.g., employment status, sources of income, home ownership).

• Mental health: mental status and symptomatology (e.g., orientation, memory, symptoms such as anxiety and depression).

• Physical health: physical status (e.g., number of doctor visits, medical prescription drugs, physical handicaps and extent of disability).

• Activities of daily living: instrumental-activities necessary to main­tain household; and physical-capacity to take care of own bodily functions.

The Activities of Daily Living (ADL) dimension assesses 14 functions including both instrumental and physical ADL's. Instrumental ADLs include: use telephone, use transportation, shopping, prepare meals, do housework, take medicine, handle money. Physical ADLs include: eat, dress oneself, care for own appearance, walk, get in/out of bed, bath, get­ting to bathroom, continence.

Each of these ADL functions is represented by one item in which examinees are asked whether they can perform the function. For example, (b, p. 169):

Can you use the telephone...

2-without help, including looking up numbers and dialing

1-with some help

0-or are you completely unable to use the telephone?

Part B of the MFAQ assesses the individual's utilization of services, that is, whether and to what extent the examinee has received assistance from various community programs, agencies, relatives, or friends, espe­cially within the latest six months. Questioning also includes the exami­nee's perceived need for the various services. Services items focus on:

• transportation

• social/recreational services

• employment services

• sheltered employment

• educational services

• remedial training

• mental health services

• psychotropic drugs

• personal care services

• nursing care

• physical therapy

• continuous supervision

• checking services (i.e., services that "check on" the elderly individ­ual periodically)

• relocation and placement service

• homemaker-household services

• meal preparation

• administrative, legal, and protective services

• systematic multidimensional evaluation of status

• coordination, information and referral services

Answers to items in both Parts A and B are given numerical (ordinal) values, but they are not summed to produce scores. Part B answers are left at this level of recording, whereas Part A answers are considered by the examiner when arriving at a rating for each of the five domains of func­tioning. The examinee may be rated 1 to 6 on each of the five dimensions, with the ratings carrying the following labels: 1 = excellent, 2 = good, 3 = mildly impaired, 4 = moderately impaired, 5 = severely impaired, 6 = totally impaired. This is a subjective rating; the examiner is not instructed how to translate answers in a domain of functioning into a rating for the domain.

The MFAQ manual offers five ways to use these ratings to summarize the examinee's current level of functioning (m).

First, the examinee's SEMPA profile (from the first letters of the five dimensions of functioning in Part A) is simply a listing of the five ratings given to the examinee in the order specified by the acronym. For example, a 2-5-3-5-5 examinee has good Social Resources (2), is severely impaired in Economic Resources (5), is mildly impaired in the domain of Mental Health (3), and so forth.

Second, the Cumulative Impairment Score (CIS) is the sum of the exami­nee's five ratings. The above examinee would have a CIS of 20. CIS scores may range from 5 to 30, and these scores are sometimes used to form classes of individuals (e.g., 5-17 and 18-30) for research or descriptive purposes.

Third, examinees may be described in terms of their number of signifi­cant impairments, a significant impairment being a rating of 4 to 6. Thus the above examinee would be said to have three significant impairments.

Fourth, the above definition of significant impairment may be used to designate the examinee as belonging to one of 32 functionally equivalent classes. These classes cover all of the possible combinations of significant impairments on the five functional dimensions. Class 0 individuals have no significant functional impairments, and Class 31 individuals have sig­nificant functional impairment on all five dimensions. A chart in the man­ual (b, p. 67) allows one to identify the above examinee (2-5-3-5-5), who has significant functional impairment on the Economic, Physical, and Activities of Daily Living dimensions, as belonging to Class 21.

Fifth, a major MFAQ study in Cleveland (j) used a combination of the above criteria to classify individuals on eight levels of well-being (k, p. 91).

Examinees may be compared on many of these summary indexes to large validation samples of elderly individuals described in the manual.

Conceptual Basis

concept definition. The Duke University group responsible for the MFAQ included a wide range of disciplines: physicians, social workers, nurses, sociologists, psychologists, economists, and systems analysts. This group chose the five dimensions of the MFAQ, guided by a survey of literature and clinical experience to determine what information was nec­essary to characterize functioning of the elderly (d).

Operationaldefinition. Many items were drawn from other instru­ments and were used without change or were modified. Each item had to meet one of several criteria concerning known or potential reliability and validity, relevance for present theory, or satisfaction of certain profes­sional standards (d, p. 17). The Activities of Daily Living scale, which assesses instrumental and physical functioning in everyday life, was bor­rowed directly from an instrument with the same name by Lawton and Brody (1).

critique. The choice and definition of concepts for the MFAQ are a product of consensus among professionals of various disciplines who worked with the elderly and were aware of practical needs for describing their functioning, and literature review. This is an acceptable method for selecting concepts, and the widespread use of the instrument suggests that the group's consensus is supported by the perspectives of mental health professionals who were not party to the MFAQ's authorship. The utility of the concepts from a legal perspective is less certain, because judges' or lawyers' opinions apparently were not included in selection of MFAQ dimensions.

At face value, the items of the MFAQ in most of the five domains seem sufficiently comprehensive with regard to functioning of the elderly. On the other hand, one can question whether specific abilities in the Activities of Daily Living dimension are sampled adequately, because each ability (e.g., management of money, use of transportation) is repre­sented by only one item.

The test authors chose merely to ask examinees whether they could perform particular functions, rather than actually testing their ability to do so. This operational choice is not explained, and it raises the question of the relation between examinees' self reports of abilities and their actual abili­ties. Two studies have investigated the relationship between self reports on the MFAQ and direct assessment and found consistency for healthy older adults but potential inconsistency with increasing dementia. Rogers et al. (n) found good agreement between self reports and direct assessments at the initial time of evaluation for 58 elderly subjects. However, on subse­quent assessments six months later, depressed patients' self reports contin­ued to agree with direct assessments while demented patients' self reports showed widening discrepancy with direct assessments. Doble et al. (c) found a similar pattern for 64 elderly subjects with self reports and direct assessments of motor ability; the pattern, however, was found for the non­demented subsample and not the demented subsample.

Psychometric Development

standardization. The manual (b) provides very careful instructions for the MFAQ interview and helpful suggestions for interviewing the eld­erly. It encourages users to obtain training in the MFAQ interview through a training program offered at the Center for the Study of Aging and Human Development at Duke University. Instructions for scoring individual items are clear and objective, although criteria for assigning a rating to each of the five domains allow considerable room for subjective judgment and discretion.

reliability. Test-retest reliability for the MFAQ has not been reported. Results with the CSQ (an earlier instrument nearly identical to the MFAQ) suggest that correlations between administrations five weeks apart are best for the Economic and Activities of Daily Living domains (.79 and.81), and poorest for the Mental Health domain (.32-42) (e).

Interrater agreement for the MFAQ (using summary ratings on each domain) produced intraclass correlations of.80 or better for the Social, Mental Health, and Activities domains, and was lowest (.66) for the Physical Health domain (h). Intrarater reliability for the MFAQ has not been examined, but results with the CSQ indicated generally consistent ratings when examiners rerated their own protocols 12 to 15 months after initial ratings (e).

norms. The manual (b) provides means and standard deviations of ratings on each domain for large samples of elderly persons seen by the Duke University Center. Separate norms are provided for elderly persons sampled randomly from the community, elderly who were referred to a clinic for age-related problems, and elderly living in institutions. A report by Fillenbaum (f) provides additional normative data for individual IADL items for three large community samples of elderly adults.

critique. Standardization and basic information on reliability sug­gest that the MFAQ can be employed with consistency by examiners. The available normative data for summary ratings (CIS data) and items (IADL) are with large enough samples to lend confidence to their stability. Additional normative data for individual subscales on Part A and Part B would be useful.

Construct Validation

Fillenbaum (f) described a factor analysis of instrumental and physi­cal ADL items of the MFAQ for two samples, finding two factors with instrumental items on one factor and physical items on another, except for the 75+ age group which found three factors with instrumental items on one factor and physical items split between two factors. Using a factor score of.65 as the acceptance criterion, only five items could be consid­ered to load onto the IADL scale (travel, shop, meals, housework, finances) while two items had insufficient scores (telephone, medication).

Whitelaw and Liang (o) used structural equation modeling to con­firm a model in which MFAQ instrumental and physical ADL ratings were linked in a causal framework to functional limitations and physical health using two large elderly samples. They suggest that reports of func­tional limitations may be a meaningful component of self-rated health.

Blazer (a) reported intercorrelations between scales of the latter instrument. The Social and Economic domains correlated with the other three domains in the.30 to.40 range, whereas the Mental Health and Physical Health domains correlated with the Activities of Daily Living domain in the.50 to.70 range for several different research samples.

Fillenbaum and Smyer (h) examined the relations between various external criteria and summary ratings on four of the five MFAQ dimen­sions. The Economic domain correlated.68 with scores on another scale (e) based on total income and assets. The Mental Health domain ratings corre­lated.67 with geropsychiatrists' ratings based on their clinical interviews. Physical Health ratings correlated.82 with physicians' ratings based on their personal examinations, and Activities of Daily Living ratings corre­lated.89 with the ratings of physical therapists who conducted home visits to examine individuals' capacities to perform everyday functions. Research with the CSQ (e) suggests that the Activities of Daily Living scale produces significantly different estimates of everyday functioning com­pared to ratings based on clinical interviews, with the clinicians' estimates tending to indicate poorer functioning than the Activities scale.

As previously noted, two studies found good agreement, for non­demented subjects, between scores on the IADL of the MFAQ and scores on instruments that directly assess functional status (c, h).

The MFAQ manual (a, e) reports differences in CSQ scores between three samples: a random sample of 997 community residents over 65 years of age, 98 clients 50 years or older who were referred to a clinic for age-related problems, and 102 elderly individuals in institutions. Differences in expected directions are shown for summary ratings on the five domains, for Cumulative Impairment Scores, and for frequencies of subjects within impaired and unimpaired classes. Evidence for statistical significance of differences, however, is not offered.

Fillenbaum (f) reported the extent of ability to perform IADL's for three groups of community dwelling elderly adults from different states; percentage endorsement is provided separately for each group with no statistically significant differences between the three groups.

critique. The initial validity research with the MFAQ, together with past research on its predecessor instrument, provides some evidence that the MFAQ scales measure the aspects of functioning that they claim to measure. Many types of research still need to be accomplished, however, in order to provide clear support for interpretations of MFAQ scores (for example, statistical comparison of criterion groups of elderly individuals living in various degrees of autonomy and dependence).

Predictive or Classificatory Utility

There have been no studies of the ability of MFAQ scales to predict future functioning under specific conditions, or to correctly classify eld­erly individuals according to their levels of adaptation and functioning in everyday life.

Fillenbaum (f) studied the relationship between five items on the MFAQ IADL scale (travel, shop, meals, housework, finances) and survival rate. Initial IADL ratings were inversely related to the one-year death rate, with those unable to perform any lADL’s initially having the highest death rate one year later.

critique. Numerous additional studies (not cited here) have used the MFAQ as one measure among many in epidemiological surveys or in regards to resource utilization, but there is no research on its utility in pre­dicting future level of care or legal competency status for clinical groups. This is probably related to the MFAQ's history as a tool developed for use more in describing large samples in survey research than for making specific predictions about individual elderly adults in clinical settings. Mental health professionals are in need of this research in order to facili­tate predictive interpretation with MFAQ data for these purposes.

Potential for Expressing Person-Situation Congruency

The Social and Economic items in the MFAQ may be useful in describing the nature of the elderly individual's current environmental circumstances, or at least the individual's perceptions of the demands of current living arrangements.

The description of person-situation congruency might be enhanced by developing normative MFAQ data for samples of persons living suc­cessfully in settings with varying degrees of autonomy or structured assistance. For example, greater incongruence would be suggested if an individual's MFAQ ratings were significantly greater (more impaired) than the normative ratings for persons living successfully in a particular environment.

References

(a) Blazer, D. (1978). The OARS Durham surveys: Description and application. In Center for the Study of Aging and Human Development, Multidimensionalfunctional assessment: The OARS methodology (pp. 75-88). Durham, NC: Duke University.

(b) Center for the Study of Aging and Human Development (1978). Multidimensional func­tional assessment: The OARS methodology. Durham, NC: Duke University.

(c) Doble, S. E., Fisk, J. D., MacPherson, K. M., Fisher, A. G., & Rockwood, K. (1997). Measuring functional competency in older persons with Alzheimer's disease. International Psychogeriatrics, 9, 25-38.

(d) Fillenbaum, G. (1978a). Conceptualization and development of the Multidimensional Functional Assessment Questionnaire. In Center for the Study of Aging and Human Development, Multidimensionalfunctional assessment: The OARS methodology (pp. 16-24). Durham, NC: Duke University.

(e) Fillenbaum, G. (1978b). Validity and reliability of the Multidimensional Functional Assessment Questionnaire. In Center for the Study of Aging and Human Development, Multidimensional Functional assessment: The OARS methodology (pp. 25-35). Durham, NC: Duke University.

(f) Fillenbaum, G. G. (1985). Screening the elderly: A brief instrumental activities of daily living measure. Journal of the American Geriatrics Society, 33, 698-706.

(g) Fillenbaum, G., & Maddox, G. (1977). Assessing the functional status of LRHSparticipants: Technique, findings, implications (Technical report No. 2). Durham, NC: Duke University, Center for the Study of Aging and Human Development.

(h) Fillenbaum, G., & Smyer, M. (1981). The development, validity, and reliability of the OARS Multidimensional Functional Assessment Questionnaire. Journal of Gerontology, 36, 428-434.

(i) Fillenbaum, C., Dellinger, D., Maddox, G., & Pfeiffer, E. (1978). Assessment of individual functional status in a program evaluation and resource allocation model. In Center for the Study of Aging and Human Development, Multidimensional functional assessment: The OARS methodology (pp. 3-12). Durham, NC: Duke University.

(j) Laurie, W. (1978a). Population assessment for program evaluation. In G. Maddox (Ed.), Assessment and evaluation strategies in aging, (pp. 100-110). Durham, NC: Duke University.

(k) Laurie, W. (1978b). The Cleveland experience: Functional status and services use. In Center for the Study of Aging and Human Development, Multidimensional functional assessment: The OARS methodology (pp. 89-99). Durham, NC: Duke University.

(l) Lawton, M., & Brody, E. (1969). Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist, 9, 179-186.

(m) Pfeiffer, E. (1978). Ways of combining functional assessment data. In Center for the Study of Aging and Human Development, Multidimensionalfunctional assessment: The OARS methodology (pp. 65-71). Durham, NC: Duke University.

(n) Rogers, J. C., Holm, M. B., Goldstein, G., McCue, M., & Nussbaum, P. D. (1994). Stability and change in functional assessment of patients with geropsychiatric disorders. American Journal of Occupational Therapy, 48, 914-918.

(o) Whitelaw, N. A., & Liang, J. (1991). The structure of the OARS physical health measures. Medical Care, 29, 332-347.

Philadelphia Geriatric Center Multilevel Assessment Inventory (MAI)

Authors

Lawton, M. P., & Moss, M.

Author Affiliation

Philadelphia Geriatric Center, Philadelphia, PA

Primary Reference

Lawton, M. P., & Moss, M. (undated). Philadelphia Geriatric Center Multilevel Assessment Instrument: Manual for full-length MAI. Philadelphia, PA: Author.

Description

The Philadelphia Geriatric Center Multilevel Assessment Inventory (MAI) was designed to assess characteristics of the elderly relevant for determin­ing their needs for services and placement in residential settings. it was developed for both research and applied use. The authors were motivated "to deal with some deficiencies and gaps in existing assessment systems" for the elderly (k), most notably the OARS instruments (see MFAQ review). Nevertheless, the authors acknowledge having borrowed much from earlier instruments that seemed to be of value.

The MAi is a structured interview procedure that obtains descriptive information about an elderly respondent related to seven domains. Each of the domains (except one) is sampled by interview questions in two or more subclasses, which the authors call sub-indexes. The full-length MAi consists of 165 items; the middle length MAI has 38 items, and the short­form has 24 items.

The domains, their sub-indexes, and the number of items in the long form, are (j):

• Physical Health (33 items): Self-rated Health Index (3): examinee rates own health. Health Behavior Index (3): reports activities engaged in for health maintenance or care. Health Conditions Index (25): reports on existence of specific health problems.

• Cognitive (15 items): Intellectual Functioning Index (11): mental status exam. Cognitive Symptoms Index (4): reports on memory lapses or confusion.

• Activities of Daily Living (16 items): Self-Maintenance Index (7): items focus on basic, life sustaining functions, such as feeding, bathing, dressing.

• Instrumental Activities of Daily Living Index (9): items include use of telephone, transportation, doing shopping, preparing meals, doing housework, doing handyman work around the house, doing laun­dry, taking medicine, managing money.

• Time Use (19 items): Time Activities Index (19): reports on fre­quency with which person has engaged in a variety of leisure and social activities.

• Personal Adjustment (14 items): Morale Index (9): reports on feelings of loneliness, discouragement, frustration. Psychological Symptoms Index (5): questions regarding insomnia, depression, fears.

• Social Interaction (8 items): Interaction with Friends (5): reports on visits and other contacts. Interaction with Family (3): reports on visits and other contacts.

• Perceived Environment (24 items): Subjective Housing Index (9): respondent's description of quality of current housing. Subjective Neighborhood Index (12): respondent's description of quality of neighborhood, convenience for everyday activities, neighbors. Personal Security Index (3): respondent's perception of degree of safety in current environment.

Each item includes one or more interview questions, with specifica­tions for assignment of scores to various possible answers. For example, the Managing Money item is typical of the format of most of the items (j, p. 12):

38a. Do you manage your money:

— without help (Score 3, skip to Question 39),

— with some help, (e.g., manage day-to-day buying but have help with checkbook and paying bills) (Score 2) or

— don't you handle money at all (no day-to-day buying) (Score 1)?

38b. Why is it that you (have some help/don't handle money at all)? (No score)

38c. Can you manage your own money:

— without help (Score 3)

— with some help (Score 2)

— are you completely unable to handle money? (Score 1)

Scores may be summed for items within an index (giving an index score), and for indexes within a domain (providing a domain score).

In addition, the MAI manual (j, pp. D1-D3) provides a method for the examiner to give an examinee a rating of 1 to 5 on each domain, called the "Domain Rating"; the authors recommend these ratings as more conven­ient than raw index or domain scores for communicating an examinee's results.

Conceptual Basis

concept definition. The guiding conceptual framework for the MAI was Lawton's model for behavioral competence of the elderly (g, h). This model outlines several spheres or domains of functioning, ranging from biological to social, with functions within a domain ranging from simple to complex. It was the product of over a decade of conceptual and empir­ical work by Lawton in gerontology and in provision of services for the elderly.

Not all of the domains in Lawton's model are represented in the MAI. Further, the MAI combines certain domains that were separate concepts in Lawton's model (e.g., Physical Self Maintenance and Instrumental Self Maintenance became "Activities of Daily Living" in the MAI, with two sub-indexes corresponding to the two earlier concepts). In addition, the MAI includes certain concepts that are not in Lawton's earlier conceptual model, especially the domains of Personal Adjustment and Perceived Environment. Conceptual background for these latter domains was bor­rowed from other investigators in the field (d, f, and m for Personal Adjustment; b and e for Perceived Environment). Originally the MAI also included an "Objective Environment" domain (e.g., quality of the plumb­ing, number of steps in the dwelling), but this was abandoned because of difficulties in operationalization or lack of variability across cases (k).

Operational definition. Neither the manual nor the journal article on the MAI describes the process by which concepts were translated into items for the instrument. The Activities of Daily Living items, however, clearly were derived from Lawton's earlier work (i) on a scale for this domain.

critique. Lawton's earlier work in conceptualizing behavioral competence of the elderly is highly regarded by other workers in geron­tology and development of the elderly. This conceptual base is an asset for the MAI.

Many of the domains and sub-indexes of the MAI would seem to provide legally relevant data for characterizing an individual's functional abilities to care for oneself in everyday life. Noteworthy in this regard are the domains of Activities of Daily Living and Physical Health. Other, less functional domains (e.g., Personal Adjustment, Perceived Environment) may provide relevant background for making inferences about the degree to which individuals will be likely to mobilize their resources to make the most of existing physical and instrumental capacities.

Content coverage of the MAI probably is incomplete for assessing an individual's functional abilities to manage property or financial matters. This content area is sampled in only a limited way by one sub-index in the Activities of Daily Living domain.

It is important to note that the MAI does not require an individual to demonstrate the functional abilities associated with the various concep­tual domains. The MAI's operational definition of its domains and sub-indexes is an individual's self-report of the things that one can do (or could do, or typically does), the individual's report of personal feel­ings or perceptions, and so forth. This method for assessing the domains can be adequate for many purposes.

Psychometric Development

standardization. The MAI manual (j, pp. E17-E19) provides con­siderable structure for the process of the interview, sequence and content of questions, and scoring. It describes criteria for 1 to 5 rating of each of the domains, but these criteria are not tied specifically to item scores. The manual discusses general considerations for interviewing elderly individ­uals and dealing with special problems of test administration with this population (e.g., with limited hearing or vision).

reliability. Interrater agreement in assignment of summary ratings was examined by comparing ratings by interviewers to ratings by trained raters who had not interviewed the respondents (484 elderly individuals). Pairs of raters manifested more than a 1-point discrepancy between their summary ratings in only 5% of cases (ranging from r =.88 on Activities of Daily Living to r =.58 on Social Interaction). Alexopoulous et al. (a) found a high degree of inter-rater reliability using inter-class correlations (r =.91-.97) in a sample of depressed older adults.

Test-retest reliability at a 3-week interval was reported as "accept­able, except for the physical self-maintenance sub-index, where variabil­ity was very low, the majority receiving perfect scores" (k, p. 95). Internal consistency (alpha) was reported as acceptable for all but the Health Behavior and Personal Security sub-indexes (r =.39 and.57, respectively). Alexopoulos et al. (a) found an acceptable degree of test-retest reliability using inter-class correlations (r =.66-.87) in a sample of depressed older adults.

norms. The MAI manual (j, pp. E17-E19) provides raw score means (not summary rating scale data) for four types of elderly respondents on all domains and their sub-indexes (for the MAI long, middle-length, and short forms). The four samples (obtained in Philadelphia) are: independ­ently living community residents, independently living tenants of public housing, high intensity in-home service recipients, and institutional waiting-list clients living in the community.

critique. Users of the MAI will find it somewhat like the well- known Diagnostic Interview Schedule in terms of its degree of structure and clarity concerning scoring or coding of data. Its summary rating scales require greater degrees of subjective estimation, yet interrater agreement generally is good even on this less structured feature of the MAI. The nor­mative data are based on sufficiently large sample sizes to warrant their general application, although samples from other geographic areas are needed to verify the generalizability of these results.

Construct Validation

Current data on validity of the MAI are from a publication by Lawton and associates (k) using the samples noted previously. First, MAI scores correlated above.55 with summary domain ratings for most domains and sub-indexes (exceptions: Interactions with Friends, and all sub-indexes in the Perceived Environment domain). Second, multiple correlations between domain items and summary ratings generally were above.40 (exceptions similar to above). Third, 14 of the 21 domain and sub-index scores correlated.20 or better with respondents' independent/dependent living status. Finally, many MAI domain and sub-index scores correlated substantially with ratings of the elderly respondents by clinicians or hous­ing administrators who had no knowledge of the respondents' MAI scores.

Most important for the present review, the sub-index that performed best on each of the above examinations of validity was the Instrumental Activities of Daily Living sub-index, that is, the items evaluating an indi­vidual's abilities to perform functions required to manage one's everyday affairs. Raw scores on this sub-index correlated.91 with summary ratings of the Activities of Daily Living domain. In comparison to other sub­indexes, it manifested the best internal validity (.86), correlated greatest with current independent or dependent living arrangement of respon­dents (.56), and correlated the highest of any sub-index with administra­tors' independent ratings of the elderly respondents (.59).

critique. This study shows that several of the MAI domain or sub­index scales may have satisfactory internal validity, may concur with inde­pendent experts' ratings of the functional characteristics of elderly person, and are related as expected to elderly individuals' degree of independent living that they apparently manage to maintain. These results are most satisfactory for the scale that assesses functional abilities in everyday life, a sub-index that would be of special interest in assessments of the elderly in guardianship cases.

Especially helpful to further demonstrate construct validity would be studies examining the relation of each of the domain or sub-index scales to other indexes. For example, one might examine the relation of Physical Health to the results of actual physical/medical diagnoses, the relation of IADL to actual demonstrations of individual's abilities to perform the functions at the level reported to examiners on the MAI, or the relation of Personal Adjustment reports to psychological measures of adjustment.

Predictive or Classificatory Utility

MAI scores predicted classification of cognitive impairment (impaired/ not impaired on MMSE cut score of 24) among 2,713 community dwelling elders (c). MAI scores have been significantly associated with severity of depression in 75 elderly adults with major depressive illness (a) and signifi­cantly associated with level of hope in 86 older patients with cancer (l).

critique. Examiners are in need of research demonstrating the rela­tion of MAI scores to examinee's success in managing various levels of independence in living when they are placed residentially after MAI assessment. These results are critical for the use of MAI scores when form­ing predictive opinions.

Potential for Expressing Person-Situation Congruency

The MAI would seem to have considerable potential for comparing an individual's functional abilities to environmental demands of various living arrangements and settings. Administration of the MAI to elderly individuals who are managing successfully in a particular type of setting could produce normative data on each MAI sale for successful adaptation to that setting. Comparison of an examinee's MAI scores to these setting­specific norms might suggest the degree to which the examinee's charac­teristics and abilities match, exceed, or fall short of the characteristics of individuals who currently meet that environment's demands. This could be described as the congruence or incongruence between the individual's abilities and situational demands, which addresses the interactive ques­tion in legal competency determinations.

Comparison of this type would be meaningful, however, only if it is clear that MAI scales (e.g., the activities scale measuring everyday func­tional abilities) actually measure what they purport to measure. Currently the greatest need is evidence that examinees' reports of their functional abilities on the MAI correspond to their actual abilities.

References

(a) Alexopoulos, G. S., Vrontou, C., Kakuma, T., Meyers, B. S., Young, R. C., Klausner, E., & Clarkin, J. (1996). Disability in geriatric depression. American Journal of Psychiatry, 153, 877-885.

(b) Andrews, R., & Withey, S. (1976). Social indicators ofwell-being. New York: Plenum Press.

(c) Barberger-Gateau, P., Commenges, D., Gagnon, M., Letenneur, L., Sauvel, C., Dartigues, J. (1992). Instrumental activities of daily living as a screening tool for cognitive impair­ment and dementia in elderly community dwellers. Journal of the American Geriatrics Society, 40,1129-1134.

(d) Brandburn, N. (1969). The structure of psychological well-being. Chicago: Aldine.

(e) Campbell, A., Converse, P., & Rodgers, W. (1976). The quality of American life: Perceptions, evaluations, and satisfactions. New York: Russell Sage.

(f) George, L., & Bearon, L. (1980). Quality of life in older persons. New York: Human Sciences Press.

(g) Lawton, M. (1972). Assessing the competence of older people. In D. Kent, R. Kastenbaum, & S. Sherwood (Eds.), Research, planning and action for the elderly (pp. 122-143). NewYork: Behavioral Publications.

(h) Lawton, M. (1982). Competence, environmental press, and adaptation of older people. In M. Lawton, P. Windley, & T. Byerts (Eds.), Aging and the environment: Theoretical approaches (pp. 33-59). New York: Springer.

(i) Lawton, M., & Brody, E. (1969). Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist, 9,179-186.

(j) Lawton, M., & Moss, M. (undated). Philadelphia Geriatric Center Multilevel Assessment. Instrument: Manual for full length MAI. Philadelphia, PA: author.

(k) Lawton, M., Moss, M., Fulcomer, M., & Kleban, M. (1982). A research and service ori­ented multilevel assessment instrument. Journal of Gerontology, 37, 91-99.

(l) McGill, J. S., & Paul, P. B. (1993). Functional status and hope in elderly people with and withoutcancer. OncologyNursing Forum, 20,1207-1213.

(m) Nydegger, C. (Ed.). (1977). Measuring morale: A guide to effective assessment. Washington, DC: Gerontological Society.

B. PERFORMANCE BASED INSTRUMENTS TO ASSESS IADLS

Direct Assessment of Functional Status (DAFS)

Author

David A. Loewenstein

Author Affiliation

Wien Center for Alzheimer's Disease and Memory Disorders, Mount Sinai Medical Center, Miami Beach and University of Miami School of Medicine

Primary Reference

Loewenstein, D. A., Amigo, E., Duara, R., Guterman, A., Hurwitz, D., Berkowitz, N., Wilkie, F., Weinberg, G., Black, B., Gittlenman, B., & Eisdorfer, C. (1989). A new scale for the assessment of functional status in Alzheimer's disease and related disorders. Journal of Gerontology, Psychological Sciences, 44, 114-121.

Description

The Direct Assessment of Functional Status (DAFS) was designed to assess functional abilities in individuals with dementing illnesses. The scale assesses seven areas:

• time orientation (16 points)

• communication abilities (including telephone and mail; 17 points)

• transportation (requiring reading of road signs; 13 points)

• financial skills (identifying/counting currency, writing a check, balancing a checkbook; 21 points)

• shopping skills (involving grocery shopping; 16 points)

• eating skills (10 points)

• dressing and grooming skills (13 points).

The composite functional score has a maximum of 93 points, exclusive of the Driving subscale which is considered optional.

The DAFS requires that the patient attempt to actually perform each item (e.g., is given a telephone and asked to dial the operator). The entire assessment is estimated to require 30-35 minutes to complete. The scale can be administered by any psychometrically trained administrator. The DAFS has been used for staging functional impairment in dementia, from one to three, in a group of 205 individuals with probable Alzheimer's disease (b).

Conceptual Basis

conceptual definition. The conceptual intent of the DAFS scale was to provide an assessment of higher order function to better answer ques­tions regarding independent function and legal competency than neu­ropsychological testing alone (a). A goal of the scale development was to utilize direct assessment to avoid the potential bias of informant report. Another goal was to assess higher order functional abilities (not just phys­ical ADL's) in some depth and with some degree of discrete measurement. The scale is conceptually based in the IADL literature.

operational definition. Scale items were developed by identifying seven domains through review of the IADL literature. Once these domains were selected, experienced geriatricians identified the functional deficits their patients experienced in each domain, the types of behaviors that could be assessed in vivo in a laboratory setting, and the types of behav­iors that were clinically significant in terms of clinical treatment. Specific functional behaviors were then developed from these suggestions and refined through piloting with cognitively impaired older adults. When possible, a hierarchical format was used for items within domains (a).

critique. The DAFS is conceptually related to other IADL scales, although it seems to be more related to clinicians' reports of functional abilities of practical clinical significance. The scale has limited coverage of issues of care for and safety in the home and care of and management regarding one's health and medications. It samples some traditional phys­ical ADL's but not toileting, transferring, or ambulation. More discussion of the conceptual relevance of these choices would be useful.

Psychometric Development

standardization. The scale, as provided in the appendix of the pri­mary reference, utilizes a detailed rating form. Specific items are listed, although standardized prompts are not provided.

reliability. Interrater reliability for each functional subscale and for the composite measure was computed for 15 memory disordered patients and 12 control subjects. Inter-rater agreement was 85%. Kappa coefficients for subscale and composite scores ranged form.911 to 1.000. Test retest reliability was computed for a 3-7 week retest for 14 memory impaired patients and 12 controls. Test-retest reliabilities ranged from.71 to.91 for the patient group and from.91 to 1.0 for the control group (a).

norms. Although formal normative data are not presented, a num­ber of elderly control groups are cited. Mean score and standard deviation for 18 elderly controls are presented in the primary reference (a). A second control sample of 50 elderly adults is also referenced (b) although data for this group are not provided.

critique. Although standardized instructions are not provided, a detailed rating form and high interrater reliability suggest good standard­ization across examiners. Other reliability estimates confirm the scale has good reliability. It appears there is a good start on normative data with two potential samples. More specific information on these samples for clinicians interested in normative comparisons would be useful.

Construct Validation

Performance on the DAFS was significantly correlated with caregiver report of functional status in the home (as measured by the Blessed Dementia Rating Scale) for a sample of 30 general memory disordered patients and a sample of 11 Alzheimer's patients (a). In another study of 72 patients with Alzheimer's disease, caregivers tended to overestimate the functional capacities of patients with objective functional impairment on the DAFS (c), especially those with higher MMSE scores.

A test of global cognition was significantly correlated with five DAFS functional tasks: telling time, using the telephone, counting currency, writing a check, and balancing a checkbook, among 33 individuals with a diagnosis of probable Alzheimer's disease. Additional cognitive tests pre­dicted functional performance in addition to global cognition for some tests. Using the telephone was also predicted by DAFS scores; letter preparation was also predicted by figural memory scores, writing a check was also predicted by WAIS Object Assembly, Fuld retrieval, and DAFS scores; balancing a checkbook was also predicted by WMS figural mem­ory and DAFS scores; shopping was also predicted by WAIS Similarities scores (d). The authors suggest the scale shows modest correlation with cognitive measures, suggesting relationship with cognitive status and also supporting the importance of administering functional scales in addi­tion to cognitive tests (since correlations were moderate not high).

The DAFS appears useful for measuring functional decline over time in a report of 52 patients with Alzheimer's disease who were assessed ini­tially and again one year later (e).

Critique. Relationships between the DAFS and caregiver reports of function in the home suggest the DAFS is useful in measuring everyday functioning. Relationships between the DAFS and cognitive measures suggest the DAFS is useful in tapping functional difficulties related to cognitive impairments and cognitive decline across time. Additional studies of the internal structure of the DAFS and its relationship to other IADL performance measures would provide additional data on construct validity.

Predictive or Classificatory Utility

DAFS subscale scores were compared between patients judged as impaired or not impaired based on chart review in areas of driving, telling time, remembering a list, and both higher and simple financial skills. Mean subscale scores were significantly different within each group (n = 10-15 per group; a). In addition, DAFS total and subscale mean scores were compared for groups of elderly controls (n = 18), Alzheimer patients (n = 12), and depressed patients (n = 11). On most subscales mean scores were significantly different between normals and Alzheimer patients, and between depressed patients and Alzheimer patients, but not between normals and depressed patients (a). In another study, patients with mild Alzheimer's dementia and mild multi infarct dementia were relatively equivalent in functional abilities and less impaired than patients with moderate Alzheimer's disease (f).

critique. Initial studies of the predictive validity of the DAFS are promising. Further studies are recommended to compare the DAFS' abil­ity to predict level of supervision or placement needed and real clinical or legal outcomes relevant to legal competence.

Potential for Expressing Person-Situation Congruency

The DAFS appears to be an ecologically valid measure, tapping per­formance on functional abilities relative to independent living. The scale itself does not provide instructions or scoring to address situational and environmental differences. These would need to be evaluated clinically given the results of the DAFS and other evaluations.

References

(a) Loewenstein, D. A., Amigo, E., Duara, R., Guterman, A., Hurwitz, D., Berkowitz, N., Wilkie, F., Weinberg, G., Black, B., Gittlenman, B., & Eisdorfer, C. (1989). A new scale for the assessment of functional status in Alzheimer's disease and related disorders. Journal of Gerontology: Psychological Sciences, 44, 114—121.

(b) Loewenstein, D. A., & Rupert, M. P. (1995). Staging functional impairment in dementia using performance-based measures: A preliminary analyses. Journal of Mental Health and Aging, 1, 47-56.

(c) Loewenstein, D. A., Arguelles, S., Bravo, M., Freeman, R. Q., Arguelles, T., Aceredo, A., & Eisdorfer, C. (2001). Caregivers' judgments of functional abilities of the Alzheimer's disease patient: A comparison of proxy reports and objective measures. Journal of Gerontology, 56, P78-P84.

(d) Loewenstein, D. A., Rubert, M. P., Berkowitz-Zimmer, N., Guterman, A., Morgan, R., & Hayden, S. (1992). Neuropsychological test performance and prediction of functional capacities in dementia. Behavior, Health, and Aging, 2, 149-158.

(e) Loewenstein, D. A., Duara, R., Rubert, M. P., Arguelles, T., Lapinski, K. J., & Eisdorfer, C. (1995). Deterioration of functional capacities in Alzheimer's disease after a 1-year period. International Psychogeriatrics, 7, 495-503.

(f) Zimmer, N. A., Hayden, S., Deidan, C., & Loewenstein, D. A. (1994). Comparative per­formance of mildly impaired patients with Alzheimer's disease and multiple cerebral infarctions on tests of memory and functional capacity. International Psychogeriatrics, 6, 143-154.

Everyday Problems Test (EPT)

Author

Sherry L. Willis and colleagues

Author Affiliation

Pennsylvania State University

Primary Reference

Willis, S. L. (1996). Everyday cognitive competence in elderly persons: Conceptual issues and empirical findings. Gerontologist, 36, 595-601.

Description

The Everyday Problems Test (EPT) was developed to focus on the cognitive demands of IADL tasks, while acknowledging that functional capacity involves multiple dimensions, including physical health and social relationships. The tests consists of six stimuli for each of seven IADL areas: managing medications, shopping for necessities, managing finances, using transportation, using telephone, maintaining a household, and meal preparation and nutrition. For each stimulus presented the examinee is asked to solve 2 problems, making 42 stimuli and 84 items in total. Willis and colleagues have also developed the Everyday Problemsfor the Cognitively Challenged Elderly (EPCCE), a similar but shorter test with 16 stimulus materials for low educated normal adults and early stage Alzheimer's patients (d, e). This review focuses on the EPT.

Conceptual Basis

conceptual definition. The EPT derives from life span developmental theories emphasizing the importance of adults' knowledge of the pragmatics of everyday life, also referred to as practical problem solving, everyday prob­lem solving, everyday cognition, and practical intelligence (b). The EPT builds on concepts of IADL's described by Lawton and Brody (1969), empha­sizing performance of IADL's and the cognitive aspects of that performance.

Operational definition. In operationalizing everyday problem solv­ing in the EPT the authors were concerned with maximizing the external or ecological validity of the test, to capture real functioning in the home environment better than previously captured in laboratory measures. It is noted that there is no consensus on what tasks best measure everyday problem solving (b). Items were developed from surveys of IADL tasks, involving actual materials which may be encountered in daily life in solv­ing everyday problems (such as a medicine label) (c).

critique. The test is nicely grounded in a theoretical framework derived from models of intelligence combined with previous work on IADL's. The operationalization of items to be ecologically valid, yet still relatively easy to administer in a laboratory setting, is a strength. The test is geared towards assessing functional competence from a psychological per­spective, some of which may be relevant to questions of legal competency to care for self and property, although additional testing may be required.

Psychometric Development

standardization. The EPT is a standardized, structured perform­ance test, with standard stimuli and instructions. For example, the exami­nee is given a phone bill and asked to identify all the days on which calls to a certain city were made. The examiner records whether the examinee finds the right page of the bill and identifies each of four correct dates.

reliability. Internal consistency reliability measured by Cronbach's alpha in a sample of nondemented elderly (n = 417) was.94 (c). Two month test-retest reliability in this sample was.94 (c).

norms. Normative data are available via the standardization sam­ple, which consisted of 417 non-demented persons (M age = 74.6, M edu­cation = 12.0) (c).

critique. The availability of standardized stimuli and prompts is a plus. Information on inter-rater reliability would be useful in establishing the extent of reliability of scores between examiners. Internal consistency reliability and test-retest reliability are excellent. The normative sample is large, although restricted to the older population.

Construct Validation

EPT scores were significantly correlated with direct observation of task performance in homes (r =.67), with elderly person's self-ratings of IADL limitations (r= -.23), spousal ratings of the elderly person's IADL limitations (r= -.24), and a measure of "functional literacy" (r =.87) (c). Subscales of the EPT and an observational measure of in-home task per­formance loaded onto the same factors in confirmatory factor analysis of data from 62 older adults.

In lisrel analyses of the fit of a seven-domain subscale structure, a seven factor solution was obtained although the factors were highly inter­related (b). In lisrel analyses comparing the EPT to two other everyday problem solving tests, the best solution was a model involving factors for each separate instrument rather than one including a general common factor upon which two or more instruments loaded (b).

critique. The studies of the factor structure and inter-test relation­ships of the EPT suggest the test measures seven domains of IADL's consistent with everyday functioning. While comparisons of the EPT and other IADL tests suggest significant method variance; additional studies which compare the EPT to more similar measures of IADL's for older populations (e.g., DAFS, ILS) would be helpful.

Predictive or Classificatory Utility

There were no studies of the EPT s relationship with level of place­ment or other clinical or legal status.

critique. The EPT is an exceptionally well-designed test, developed for research purposes concerning studies of everyday cognition. Studies that compare the EPT with various clinical or legal outcomes might extend its validity for clinical and legal applications.

Potential for Expressing Person-Situation Congruency

There are no specific items or scoring built into the EPT to consider person-situation congruency.

References

(a) Diehl, M., Willis, S. L., & Schaie, K. W. (1995). Everyday problem solving in older adults: Observational assessment and cognitive correlates. Psychology and Aging, 10, 478-491.

(b) Marsiske, M., & Willis, S. L. (1995). Dimensionality of everyday problem solving in older adults. Psychology and Aging, 10, 269-283.

(c) Willis, S. L. (1996). Everyday cognitive competence in elderly persons: Conceptual issues and empirical findings. Gerontologist, 36, 595-601.

(d) Willis, S. L., Allen-Burge, R., Dolan, M. M., Bertrand, R. M., Yesavage, J., & Taylor, J. L. (1998) Everyday problem solving among individuals with Alzheimer's Disease. The Gerontologist, 38, 569-577.

(e) Willis, S. L., Dolan, M. M., & Bertrand, R. M. Problem solving on health-related tasks of daily living. Unpublished manuscript.

c. performance based instruments to assess need FOR GUARDIAN OR CONSERVATOR

Decisionmaking Instrument for Guardianship (DIG)

Author

Stephen Anderer

Author Affiliation

Schnader, Harrison, Segal and Lewis LLP, Attorneys at Law, Philadelphia, PA

Primary Reference

Anderer, S. J. (1997). Developing an instrument to evaluate the capacity of elderly persons to make personal care and financial decisions. Unpublished doctoral dissertation, Allegheny University of Health Sciences.

Description

The Decisionmaking Instrumentfor Guardianship (DIG) was developed to evaluate the abilities of individuals to make decisions in everyday situ­ations often raised in guardianship proceedings. The instrument consists of eight vignettes describing situations involving problems in eight areas: hygiene, nutrition, health care, residence, property acquisition, routine money management in property acquisition, major expenses in property acquisition, and property disposition. Examinees are read a brief vignette describing these situations in the second person. For example, the first vignette is "David is an 80 year old man. His wife died two years ago and he lives alone in his home. He has severe arthritis. For two months, David has had difficulty washing himself because of his arthritis. He has not been able to do his cleaning. Food scraps have been left out in the kitchen. The rugs have become dirty." Gender is varied to match the examinee's gender. To facilitate disclosure, individuals are provided a list of major points in each vignette in large print. For example, in the vignette above the list reads: "lives alone, severe arthritis, difficulty washing self, not able to do cleaning, food scraps left out, rugs dirty."

Each disclosure is followed by a series of questions to evaluate problem solving for that situation, such as:

• What is the problem?

• What could David do to solve the problem?

• Which solution to David's problem do you think is the best one?

• How would you describe to David the reasons that he should pre­fer this solution?

Scoring criteria are used to assign points for aspects of problem solving including defining the problem, generating alternatives, consequential thinking, and complex/comparative thinking.

Conceptual Basis

conceptual definition. The DIG was based on social problem solving theory, using a social problem solving model developed by D'Zurilla and Nezu (b). The model has two components: a general motivational compo­nent known as "problem orientation," and a set of four problem solving skills. Problem solving skills include the ability to define the problem, generate alternatives, evaluate alternatives, and implement and verify the solution.

operational definition. These concepts were operationalized in the context of guardianship through a multi-step process to insure legal rele­vance (a). First, interviews were conducted with legal and social service professionals experienced with issues of guardianship. Interviewees were asked to identify typical problematic situations confronted by elderly individuals for whom guardianship is sought. These interviews were used to generate a list of problematic situations which was distributed to a second sample of legal and social service professionals. Respondents were asked to rate each situation in terms of frequency of occurrence in their experience and relevance to legal competency. Vignettes were devel­oped to address situations rated as most frequent and most relevant and to cover a range of different decision making situations.

critique. The DIG appears to be nicely and appropriately grounded in problem solving theory. The efforts to operationalize this theory in items that are legally relevant to questions of guardianship through the surveys of legal and social service professionals is a major strength. The DIG appears to be an innovative and promising instrument to assess problem solving for vignettes of everyday situations. Additional perform­ance based testing would supplement the DIG in providing a comprehen­sive assessment in questions of guardianship. For example, some individuals may have adequate working memory and problem solving, but impaired recent memory. Such individuals could conceivable do well on the DIG yet still be impaired in their everyday function when left at home alone, i.e., memory problems could cause them to mis-pay bills or mistake medications.

Because the DIG has one vignette for eight areas it may inadequately sample all aspects related to functional abilities for legal compe­tencies, such as abilities to care for home via cleaning, maintenance, and personal safety. A strength of the DIG is the inclusion of four vignettes concerning financial management, and in this area the coverage appears to be better.

Psychometric Development

standardization. The DIG is carefully standardized. Standard instructions, vignettes, questions, and prompts are provided in the manual. In addition, detailed scoring criteria are provided. Sheets with simplified lists of salient points of each vignette are provided (in large type), helping to standardize vignette administration and emphasize the assessment of problem solving and not reading comprehension or memory. Vignettes are kept simple, easy to understand, and brief.

reliability. Inter-scorer reliability between three raters ranged from.93 to.97 for the total scale, and.77 to.98 for subscale scores. Internal con­sistency reliability estimates on the basis of Cronbach's alpha was.92 and Spearman-Brown split half reliability was.93 (a).

norms. Mean and standard deviation scores for 61 individuals ranging in age from 61 to 90 years old are provided (a). Specific informa­tion on the neurological and physical health of these individuals is not provided.

critique. This instrument has been extensively standardized in both administration and scoring, and this appears to be reflected in the excellent inter-rater reliability. The standardization sample provides a good start for normative data, but since subjects were drawn from residential facilities, adult day health care programs, and senior centers, the sample is likely to be mixed, reflecting some normal and some potentially neurologically compromised adults. Additional normative data would be useful.

Construct Validation

The DIG total score was significantly correlated with four measures of cognition: the CERAD summary score, WAIS Block Design, WAIS Vocabulary, and a measure of capacity to reason when giving informed consent (Thinking Rationally About Treatment, or TRAT: c). Correlations ranged between.54 and.73, with highest correlation between the DIG and the TRAT (a). The DIG total score was significantly correlated with total scores on CERAD tests, ranging from.35 to.59, with highest correlation between the DIG and a test of verbal fluency.

critique. Initial comparisons of the DIG with other cognitive tests find moderate positive correlations. This is a good start to examining con­struct validity issues. Studies are needed that look further at the internal structure of the DIG and relationship to other functional and guardianship measures.

Predictive or Classificatory Utility

There are no studies of the predictive or classificatory utility of the DIG at this point.

critique. Studies that examine the ability of the DIG to predict level of supervision needed or type of guardianship procured would be useful.

Potential for Expressing Person-Situation Congruency

The DIG is highly standardized and does not contain items within the scale to assess the fit of the functioning on the issue in question with the individual's environmental and situational resources. Such adjustments could be made in interpreting the DIG. Note however that the vignettes on the DIG are very relevant to everyday situations encoun­tered by adults for whom guardianship is in question and in this way appears to be quite ecologically valid.

References

(a) Anderer, S. J. (1997). Developing an instrument to evaluate the capacity of elderly persons to make personal care and financial decisions. Unpublished doctoral dissertation, Allegheny University of Health Sciences.

(b) D'Zurilla, T. J., & Nezu, A. M. (1982). Social problem solving in adults. In P. C. Kendall (Ed.), Advances in cognitive-behavioral research and theory (Vol. 1). New York: Academic.

(c) Grisso, T., & Appelbaum, P. S. (1993). Manual for Thinking Rationally about Treatment. Worcester, MA: University of Massachusetts Medical School.

Financial Capacity Instrument (FCI)

Author

Daniel Marson and colleagues

Author Affiliation

Department of Neurology, Alzheimer's Disease Center, and Center for Aging at University of Alabama at Birmingham, Birmingham, AL

Primary Reference

Marson, D. C., Sawrie, S. M., Snyder, S., McInturff, B., Stalvey, T., Boothe, A., Aldrige, T., Chatterjee, A., & Harrell, L. E. (2000). Assessing financial capacity in patients with Alzheimer's disease: A conceptual model and prototype instrument. Archives of Neurology, 57, 877-884.

Description

The Financial Capacity Instrument (FCI) (d) was designed to assess abilities associated with management of everyday financial activities. The instrument assesses six domains of financial activity: basic monetary skills, financial conceptual knowledge, cash transactions, checkbook management, bank statement management, and financial judgment. The FCI is reported to require between 30-50 minutes to administer, depending on the cognitive level of the examinee. Of note, Marson and his group are also validating a semi-structured interview to assess financial capacity, the Clinical Assessment Interview for Financial Capacity (CAI) (e).

Conceptual Basis

conceptual definition. Financial capacity was identified as an advanced ADL, said to be conceptually and statistically distinct from other ADL's such as those related to household management, and is asso­ciated with cognitive functioning. The authors define financial capacity as a multidimensional construct involving declarative knowledge, proce­dural knowledge, and judgment, and requiring simple and complex pro­cessing (b).

Operational definition. An interactive process of test development was used to operationalize financial capacity into domains, with tasks for each domain (c). Inclusion criteria for domains were: theoretical relevance to independent functioning; clinical relevance to health care profession­als; and general relevance to Alabama statutes for financial competency. Tasks within each of these domains were identified as being: theoretically relevant to the domain; practical to implement in the laboratory; repre­sentative of procedural knowledge, declarative knowledge, or judgment; and of varying difficulty levels sensitive to dementia at different stages.

critique. The FCI is nicely grounded in theories of declarative and procedural knowledge. One of its strengths is the consideration of the multiple types of cognitive skills hypothesized to be important for differ­ent types of financial capacities. Although many IADL tests have included items for money management, the FCI represents a significant improve­ment over these through its use of six domains, providing for comprehen­sive assessment.

Psychometric Development

standardization. The FCI uses an explicit protocol for administra­tion and scoring (a).

reliability. Test-retest reliability was investigated on a subset of 17 Alzheimer's patients and controls and was found to range from.85 to.98 for the six domains of financial capacity (d). Percent exact inter-rater agree­ment was examined on another subset of 11 patients and controls using two independent raters and was found to range from 86.4 to 99.7 for the six domains (d). Internal consistency reliability was computed using Cronbach's alpha and was found to range from.85 to.93 over the six domains (d).

norms. Normative data are available for 23 normal older controls, mean age 70.3 years, recruited for initial reliability and validity studies who were assessed to have sufficient pre-existing experience in all of the financial domains (to exclude individuals who may not have been involved in specific financial tasks and whose performance on these would reflect lack of experience rather than capacity impairment) (d).

critique. Initial reliability estimates are excellent and suggest the FCI has acceptable test-retest, interrater, and internal consistency reliabil­ity. The normative sample is appropriate for use in early test develop­ment, and the sample appears to be appropriately selected with an ingenious consideration of previous financial experience. Additional normative data are needed to enable more broad scale clinical use.

Construct Validation

Performance on the FCI was compared with performance on a neu­ropsychological battery for 35 Alzheimer's patients (b). In these analyses, Basic Monetary Skills were predicted by performance on Trails A and Tokens; Conceptual Knowledge was predicted by performance on Boston Naming and Dementia Rating Scale Attention; Cash Transactions were predicted by performance on Trails A; Checkbook Management was predicted by performance on Dementia Rating Scale Attention; Bank Statement Management was predicted by performance on Tokens, WAIS Similarities and Dementia Rating Scale Construction; Financial Judgment was predicted by performance on the Boston Naming Test (b). Working memory was found to be key in financial capacity (a).

critique. Early construct validation work is interesting and promis­ing. Additional construct validation that considers performance on the FCI across multiple groups is needed. Factor analyses that explore the six domains of financial capacity developed on the FCI would be most inter­esting, possibly proving useful in relating performance on the FCI to dif­ferent legal capacities regarding estate management that are in question in a specific guardianship situation.

Predictive or ClassiIicalorY Utility

The performance of 23 normal controls was compared to the per­formance of 30 individuals with mild Alzheimer's disease and 20 indi­viduals with moderate Alzheimer's disease (d). Mean performance of controls differed from patients with moderate Alzheimer's disease for all six domains, and from patients with mild Alzheimer's disease for five of six domains (the exception being basic monetary skills involving naming, valuing, and counting coins and currency). Patients with mild versus moderate Alzheimer's disease were classified as "capable," "moderately capable," and "incapable" according to FCI scores, based on control- referenced methods. Distributions across categories were statistically different for the two groups (d).

critique. These studies suggest the FCI can be useful in predicting the level of financial incapacity and subsequent financial management needed. The authors nicely include the consideration of a "marginally incapable" group to allow for finer discriminations appropriate to mod­ern limited guardianships.

Potential for Expressing Person-Situation Congruency

There are no specific items within the FCI to address person-situation congruency. However, performance on each of the six domains could be clinically compared to the financial demands and resources in individual cases.

References

(a) Earnst, K. S., Wadley, V. G., Aldridge, T. M., Steenwyk, A. B., Hammond, A. E., Harrell, L. E., & Marson, D. (2000). Loss of financial capacity in Alzheimer's disease: The role of working memory. Aging, Neuropsychology, and Cognition, xx, 1-11.

(b) Marson, D. C. (2001). Loss of financial competency in dementia: Conceptual and empir­ical approaches. Aging, Neuropsychology, and Cognition, xx, 1-17.

(c) Marson, D. C., Sawrie, S., Stalvey, T., McInturff, B., & Harrell, L. (1998, February). Neuropsychological correlates of declining financial capacity in patients with Alzheimer's dis­ease. Paper presented at the meeting of the International Neuropsychological Society, Honolulu, Hawaii.

(d) Marson, D. C., Sawrie, S. M., Snyder, S., McInturff, B., Stalvey, T., Boothe, A., Aldridge, T., Chatterjee, A., & Harrell, L. E. (2000). Assessing financial capacity in patients with Alzheimer's disease: A conceptual model and prototype instrument. Archives of Neurology, 57, 877-884.

(e) Marson, D. (2000, November). Assessing financial capacity in Alzheimer's disease: A clinical interview approach. Paper presented at the meeting of the Gerontological Society of America, Washington DC.

Hopemont Capacity Assessment Interview (HCAI)

Author

Barry Edelstein

Author Affiliation

West Virginia University

Primary Reference

Edelstein, B. (1999). Hopemont Capacity Assessment Interview manual and scoring guide. West Virginia University: author. (Available from Barry Edelstein, Department of Psychology, P.O. Box 6040, West Virginia University, Morgantown, WV 26506.)

Description

The Hopemont Capacity Assessment Interview (HCAI) (b) is a semi­structured interview in two sections. The first section is for assessing capacity to make medical decisions. The second section, discussed here, assesses capacity to make financial decisions.

In the interview the examinee is first presented with concepts of choice, cost, and benefits and these concepts are reviewed with the exami­nee through questions and answers. The examinee is then presented three financial scenarios. In the first, the examinee has a limited amount of money (40 cents) and only he or his friend can use it to make a purchase. In the second scenario, the examinee wants to make a purchase and is offered that item or is offered twice the money. In the third scenario, a friend has a limited amount of money (5,000 dollars) and is trying to choose between using the money for a child's college tuition versus saving the money for his nursing home care needs. For each scenario the individual is asked basic questions about what he or she has heard, and then asked to explain costs and benefits, to make a choice, and to explain the reasoning behind that choice. There are 30 items in all. Detailed scoring procedures are not described, but the examiner is referred to score the results with respect to legal standards articulated by Appelbaum and Grisso (1988).

Conceptual Basis

conceptual definition. The HCAI was developed in reference to four legal standards described by Appelbaum and Grisso (1988) relevant for assessing civil competencies. Questions in response to disclosed infor­mation are intended to assess the capacities to understand relevant

information, demonstrate appreciation of the significance of the informa­tion for the circumstance, rationally consider the risks and benefits of different choices, and express a choice.

operational definition. The interview was structured and made simple to be suitable for cognitively impaired elderly adults, including those residing in nursing homes. Two scenarios were chosen with differ­ent levels of risk to increase generalizability. Items were developed based on a standard sixth grade reading level.

critique. The HCAI is appropriate for older cognitively impaired adults, but may be too easy to provide fine discriminations for adults with mild or more subtle impairments. The conceptual relation of the HCAI to legal standards identified in the literature is a strength.

Psychometric Development

standardization. The HCAI uses a semi-structured format. General instructions are provided. Specific standardized introductions, scenarios, and follow-up questions are on the rating form. A three point rating sys­tem was developed for research purposes (2 = adequate, 1 = partially ade­quate, 0 = inadequate) (a,b).

reliability. Inter-rater reliability was calculated by comparing the scoring of two raters for 17 protocols (c). Inter-rater reliability, assessed through the exact agreement formula, was.93. Two week test-retest stabil­ity was.50 (d).

norms. Normative data are not available.

critique. Standardization of administration is excellent. Inter-scorer reliability is good. Test-retest reliability is moderate. However, few instru­ments even provide such estimates, so the investigators are commended for their efforts in this area. As the authors note, reasons for lower test- retest reliability need to be determined (d). Normative data are needed.

Construct Validation

The correlation between the HCAI-financial section and the MMSE for 93 residents of a long term care facility was.60 (a). Correlations between the HCAI-financial section and nine cognitive measures for 50 residents of several long term care facilities were calculated in a second study (b). All correlations were positive and statistically significant. The highest correla­tion was with WAIS Vocabulary, and this test also emerged as the best predictor of HCAI-financial scores in multiple regression analyses.

critique. These initial studies of relationships between the HCAI- financial and other cognitive measures are useful and may suggest what abil­ities are important for good performance on the HCAI instrument. Additional studies comparing the HCAI-financial to other measures of finan­cial decision making would help to extend the construct validity of the test.

Predictive or Classificatory Utility

Mean MMSE score for 74 individuals judged to have poor financial capacity on the basis of the HCAI was 12.8, while mean MMSE score for 19 individuals judged to have adequate financial capacity was 24.4 (c). Similar mean differences were found in a second study (d).

critique. Additional studies are needed to determine whether the HCAI accurately predicts capacity to manage finances as determined by other methods (e.g., other financial measures or consensus opinion).

Potential for Expressing Person-Situation Congruency

The HCAI is highly standardized and does not include items to assess the fit of financial capacities as assessed on the instrument with environmental or situational circumstances (e.g., the extent of the individ­ual's estate needing management).

References

(a) Edelstein, B. (2000). Challenges in the assessment of decision making capacity. Journal of Aging Studies, 14, 423-437.

(b) Edelstein, B. (1999). Hopemont Capacity Assessment Intervieiv manual and scoring guide. West Virginia University: author. (Available from Barry Edelstein, Department of Psychology, P.O Box 6040, West Virginia University, Morgantown, WV 26506.)

(c) Edelstein, B., Nygren, M., Northrop, L., Staats, N., & Pool, D. (1993, August). Assessment of capacity to make financial and medical decisions. Poster presentation at the APA Annual Convention, Toronto, Canada.

(d) Staats, N. (1995). Psychometric evaluation of the Hopemont Capacity Assessment Interview. Unpublished Master's Thesis, West Virginia University.

Independent Living Scales (ILS)

Author

Loeb, P. (Anderten)

Author Affiliation

Private Practice

Primary Reference

Loeb, P. A. (1996). Independent Living Scales. San Antonio: Psychological Corporation.

Description

The Independent Living Scales (ILS) is an individually administered instrument developed to assess abilities of the elderly associated with caring for oneself and/or for one's property. The ILS is a significantly mod­ified version of the earlier Community Competence Scale (CCS) (reviewed in the first edition of this book). The CCS was constructed specifically to be consistent with legal definitions, objectives, and uses, in order to enhance its value for expert testimony about capacities of the elderly in legal guardianship cases.

The ILS consists of 70 items in five subscales:

• Memory/Orientation (awareness of surroundings and short-term memory)

• Managing Money (ability to count money, do monetary calcula­tions, pay bills, and take precautions with money)

• Managing Home and Transportation (ability to use telephone, public transportation, and maintain a safe home)

• Health and Safety (awareness of personal health, ability to evaluate health problems, handle medical emergencies, take safety precau­tions), and

• Social Adjustment (mood and attitude towards social relations)

The five subscales may be summed to obtain an overall score which is meant to reflect the individual's capacity to function independently over­all. Two factors may be derived from items across the five subscales: Problem Solving and Performance/Information.

ILS items use a direct performance format (e.g., examinee is asked to demonstrate how to dial the operator) and Wechsler subscale type format (e.g., examinee is asked to explain a situation or solve a problem). Direct performance items require examinees to engage in some activity that will show the degree to which they know or can do a specific thing. Some of these items require special testing materials: for example, blank checks, schematic maps, money, an envelope (for the examinee to address), a checking account balance record, a key. The ILS kit includes many of these items, although examiners supplement these with common items in their possession (e.g., telephone, coins).

Scoring requires the assignment of 2 to 1 to 0 or 2 and 0 point credits based on scoring criteria provided in the manual for each item. Scoring criteria formats are similar to Wechsler Information, Comprehension, or Vocabulary subscale formats. For example, Item 6 regarding use of a bus schedule in "Managing Home and Transportation" is scored as follows:

• 2 points: Mentions two or more facts (route bus follows, when bus arrives, where the stops at a destination, how much it costs to ride)

• 1 point: Mentions one fact

• 0 point: Does not know

Testing begins with seven screening items to assess ability to see, read, hear, speak, write and walk to assess the examinee's test taking abil­ity in light of these functions. All items may be presented in a written for­mat for individuals with poor hearing but intact vision and reading. The manual encourages examiners to use nonleading questions to allow examinees to clarify their responses ("explain what you mean," "tell me more about that") or to give more than one answer (where relevant), in light of the instrument's intent to assess maximum functional ability of the examinee. Administration typically requires 45 minutes.

Conceptual Basis

concept definition. Loeb used an empirical method to select and define the component concepts for the ILS (a, Loeb and Anderten are the same author). "Caring for self and/or property" was identified as the pri­mary legal construct, and "caring for sell" and "caring for property" were considered to be two subconcepts. Loeb reviewed relevant statutes of all states, major relevant case law, and legal scholarly writings to collect an initial list of components, that is, broad categories of human abilities referred to in these sources. In addition, Loeb used an exploratory inter­view procedure to elicit additional components, as well as more specific abilities (referred to as "attributes" contributing conceptually to a com­ponent). Interviews were conducted with probate judges, attorneys, physicians, psychiatrists, psychologists, nurses, and social workers who worked with or made decisions about the elderly, as well as with many elderly individuals themselves. Major interview questions focused on the respondent's perceptions of the things that elderly individuals need to be able to do in order to care for themselves or their property. A composite set of tentative components and their specific ability attributes was constructed from the literature review and interview results.

Loeb (a) then employed a two-stage survey process in which respon­dents rated the components and their attributes for their importance in relation to the care-for-self subconcept and separately for importance regarding the care-for-property subconcepts. The 288 respondents in the final survey included Missouri probate judges, a random national selection of members of the Gerontological Society and of Division 20 (Adult Development and Aging) of the American Psychological Association, mental health professionals who regularly performed guardianship assessments in St. Louis, and elderly individuals who advised a local agency for the aging. (Significant differences between respondent groups in importance ratings were found on only five of the components.)

Nineteen components receiving the highest importance ratings for either the self or property concepts became the subscales: judgment, emer­gencies, acquire money, compensate incapacities, manage money, commu­nication, care medical, adequate memory, satisfactory living arrangement, proper diet, mobility, sensation, motivation, personal hygiene, maintain household, utilize transportation, verbal/math, social adjustment, and dangerousness. Attributes (specific abilities) with the highest importance ratings within a component provided the conceptual definition for the given component.

The above procedures were employed for developing the CCS from which the ILS was developed (by the author in collaboration with the Psychological Corporation: g). This revision resulted in 118 prototype items, including 94 from the CCS and 24 new items added with the inten­tion of improving reliability (g). These were administered to a nonclinical sample of 590 adults and a clinical sample of 248 adults. This process resulted in a final selection of 70 items for the ILS; 48 items were dropped because of lack of discrimination between adults at different levels of dependence, low inter-item correlation, and/or sex bias.

operational definition. Items were constructed to reflect the spe­cific abilities (attributes) that were determined in the above procedure as defining a given component. Whenever possible, Loeb (f) constructed items that required the examinee to demonstrate the specific ability in question, to demonstrate knowledge or relevant factual information, or to demonstrate the use of information in solving hypothetical, everyday problems in living. The number of items for a component was dictated primarily by the number of constituent attributes arising from the earlier empirical analysis, as well as the need to have adequate samples of behav­ior related to each attribute. Wechsler's item and scoring formats offered a general model for the ILS.

critique. Loeb's empirical approach to conceptual definition pro­vides the ILS with a firm base in terms of content validity. The compo­nents and specific abilities represent the combined perspectives of legal professionals who address the legal construct of competency in guardian­ship cases, the psychological and behavioral perspectives of mental health and social service personnel who assess and serve the elderly, and the practical perspectives of the elderly themselves. (Interestingly, the per­spectives of these various groups did not differ markedly (a).) Thus, the process allows one to postulate that the components have conceptual ties to both the legal notion of competency to care for self/property and clini­cal notions of capacities relevant for independent functioning. The extent to which the final scale is a comprehensive reflection of these efforts is less clear, since so many items were eliminated.

Loeb also seems to have been successful in constructing items that require an examinee's demonstration of the abilities (attributes) associ­ated with the various components. The MAI and the MFAQ (reviewed earlier) ask examinees to report whether or not they can perform a certain function; in contrast, the ILS asks them to perform it. This approach has considerable benefits for the examiner who wishes to use assessment data to make inferences about everyday functioning, because ILS responses may be much closer conceptually to everyday functioning than are self­reports. On the other hand, one must remember that performance in test situations—even performing everyday functions-may not always reflect an examinee's behavior in everyday life.

A variety of short forms of the original scale (CCS) have been sug­gested involving 42 items or 50 items (h, i), but currently only the 70 item standardized ILS is available (g).

Psychometric Development

standardization. Wording and sequence of items were carefully developed (f) and refined (b) in the initial scale, and further specified in the standardization efforts. Examiner comments about the clarity of rules for administration and scoring were compiled. Ambiguities were resolved and in some cases items that were difficult to score or administer were eliminated from the final scale (g). The final test kit contains a detailed administration form with each item appearing as read to the subject in its entirety and a scoring manual with examples of 2, 1, and 0 point responses.

reliability. Internal consistency in the standardization sample con­sisting of 590 nonclinical cases was.88 for the full scale. Subscale alphas were:.77 for Memory/Orientation,.87 for Managing Money,.85 for Managing Home and Transportation,.86 for Health and Safety,.72 for Social Adjustment. The developers note that the Memory/Orientation scales and the Social Adjustment scales have the fewest items, likely atten­uating alpha. Factor score alphas were.86 for Problem Solving and.92 for PerformanceZInformation.

Test-retest reliability estimates in 80 adults from the standardization sample who were administered the test approximately 2 weeks post ini­tial test (range 7-24 days) were.91 for the full scale,.84 for Memory/ Orientation,.92 for Managing Money,.83 for Managing Home and Transportation,.88 for Health and Safety,.81 for Social Adjustment,.90 for the Problem Solving factor, and.94 for the PerformanceZInformation factor. Test-retest decision consistency was compared by looking at the distribu­tions of individuals classified as high, moderate, or low functioning (determined on the basis of ILS cut-off scores) developed by examining current living situation and self-reports of areas of need for assistance, ini­tially and at retest. In general test-retest decision consistency was good except for the Social Adjustment scale. Individuals tended to improve over time, which may reflect clinical improvement and/or experience with the test. Test security is emphasized (g).

Interrater reliability for the ILS is excellent at.99 for the total scale and.95-.99 over subscales and factor scores (g).

norms. Normative data are available for 590 adults age 65+, approx­imately evenly split by gender and ranging across educational levels (g).

critique. The ILS has evidence for excellent reliability in all domains. The normative sample is adequate for assessing older adults, although additional normative data would be required if the test were to be used with younger, for example psychiatrically ill, adults. ILS develop­ment included extensive attention to standardization of procedures and scoring, resulting in an instrument that is relatively easy to administer and score. Overall, the test has exceptional psychometric properties.

Construct Validation

The ILS aims to have good validity for the legal construct of capacity to care for self and property through the identification of content domain through extensive survey methods described earlier. The final subscale development was guided by item q-sorts by four experts in the field. As noted earlier, factor analysis of the ILS revealed two factors: Problem solv­ing, comprised of questions that demand complex reasoning and problem solving, and PerformanceZInformation, comprised of questions that require knowledge of factual information.

Early studies of the ILS in its original form, the CCS, examined corre­lations between the CCS and other measures. The CCS correlated poorly (.21) with scores on the Geriatric Profile (a measure of symptomatic behav­ior in the elderly, d, f) and only moderately (.40) with scores on Zigler and Phillips' (k) Social Competence Scale (h). In both of these studies, however, the CCS was more closely related to the degree of independence that subjects maintained in living arrangements than were the measures to which the CCS was being compared (see Predictive and Classificatory Utility below). Dunn (c) found that the seven core measures in the Halstead-Reitan Battery forneuropsychologicalassessmentwerecapableof accounting for 41% of the variance (multiple regression) in total CCS scores, in an elderly sample being assessed for suspected dementia related to brain pathology. The highest Pearson correlations between Halstead measures and total CCS scores were for the Halstead's Speech Sounds Perception Test (-.56), Rhythm Test (-.47), and Tactual Performance (Memory) Test (.38).

Subsequent work with the ILS itself also examined correlations with cognitive measures (including WAIS-R subscale and IQ scores) among 90 adults in a nonclinical sample, and Microcog index scores among 47 adults in a nonclinical sample. In general, correlations were moderate except for Social Adjustment, suggesting the constructs are related but not identical.

In another study of the CCS (the predecessor to the ILS) Dunn (c) compared CCS scores with relatives' ratings of the examinee's function­ing in each area. Correlations between CCS subscale scores and their cor­responding SCIL ratings by relatives were above.40 for Communications and Money Management; between.30 to.39 for Care for Medical Needs, Emergencies, Personal Hygiene, and Utilize Transportation; and below.30 for Acquire Money, Compensate for Incapacities, Diet, and Manage Household. Total CCS and total SCIL scores correlated.41. Subsequent work with the ILS itself found that among 90 adults in a nonclinical sam­ple, ILS subscales correlated between.53 to.67 with MAI total scores, and ILS total scores correlated.71 with MAI total scores (g).

A series of studies examined the use of the ILS with clinical popula­tions including adults with mental retardation, traumatic brain injury, dementia, chronic psychiatric disturbance, schizophrenia, and major depression. In general, patients performed worse on ILS scales than did controls although this varied across groups and was not statistically significant for every subscale for every group, especially not for Memory/ Orientation and Social Adjustment (g).

critique. It is interesting that the ILS factor analysis suggests two factors that are consistent with two of the three item types that were iden­tified on the CCS. It is unclear whether this represents two important ele­ments in the construct of capacity to care for self and property or merely "method variance" associated with the items. It would have been interest­ing if factor analyses found more support for the ILS subscales and partic­ularly the notion of financial management as being somewhat different from Health and Safety or Home and Transportation skills. Patterns of correlation with the WAIS-R suggest these subscales may rely on slightly differentabilities.

There is some evidence that the ILS and its predecessor, the CCS, are related to the degree of independent living that one can maintain in everyday life, and that they relate more closely to this criterion than do indexes of general cognition or social competence. It would be interesting to compare scores on the ILS with scores on other performance-based IADL instruments like the DAFS or EPT.

The ILS exceeds other IADL scales in the quantity of data available comparing it with other tests and between patient groups. Yet the ILS is in need of research showing that the scores on various component subscales correspond to external indicators of the same specific functional abilities that the subscales claim to measure. For example, do Managing Money scores correlate highly with the degree of responsibility for personal financial dealings actually assumed by individuals in everyday life? Furthermore, do Managing Money scores correlate with judicial assess­ments of need for a conservator? The ILS was carefully developed to cor­respond with the legal construct of capacity to care for self and property, although some of the content coverage and specificity may have been diminished in the standardization efforts which reduced the number of items to less than half developed in the original clinical and legal surveys. Studies comparing the current version of the ILS to expert assessments of relevant legal constructs would be interesting.

Predictive or Classificatory Utility

In early studies, total CCS scores were related significantly to the degree of independence in examinees' current living arrangements. Loeb (f) found highly significant differences (for total CCS scores and for 14 of the 19 CCS subscales) between elderly individuals in three living arrange­ments: living independently, living in a home for the aged with little daily assistance, and living in a home for the aged with much daily assistance and structure. Post hoc analyses identified significant differences prima­rily on Communication, Emergencies, Personal Hygiene, Manage Money, Utilize Transportation, and Verbal/Math Skills. In a different population, Searight (i) found significant total CCS score differences (using a 16 sub­scale version of the CCS) between boarding home and independent apart­ment dwellers (not geriatric), all of whom were recently discharged from a psychiatric hospital. Searight (h) found the same result in another deinsti­tutionalized psychiatric sample involving both urban and rural settings.

Caul (b) examined the ability of the CCS to predict the future com­munity adjustment of psychiatric patients examined with the CCS during hospital treatment. Community adjustment was assessed with the Brockton Social Adjustment Scale (j, a rating of living and employment independence) and the Katz Adjustment Scale (e, professionals' ratings of behavioral symptomatology) approximately three months after CCS examination in the hospital. The CCS correlated.49 with later Brockton adjustment, but only.10 with Katz adjustment.

Subsequent work with the standardized ILS established cut scores on the ILS to correspond with level of independent functioning in the community (g). The cut scores are useful in interpretation of the scale for clinical purposes.

critique. When added to the results described above under Construct Validation, these findings lend further support to the inference that the ILS measures functional abilities of some importance in manag­ing one's everyday life and financial resources. With the development of cut-scores during the standardization of the ILS, the classificatory utility of the instrument is increased, although still must be provided in light of other data. The score may suggest a level of supervision needed and would need to be combined with other data to make recommendations regarding legal adjudication for guardianship.

Potential for Expressing Person-Situation Congruency

It is possible to interpret the ILS on several levels. The examiner can consider the full scale score as an indicator of level of functioning, high for independent functioning, moderate for semi-independent functioning, and low for dependent functioning. However, there is the possibility for more refined interpretation, considering person-situation congruency at other levels of interpretation. Subscale scores and individual item infor­mation may be interpreted in light of information about the individual's situation and resources. The extent to which family or professional assis­tance is available and the extent to which the individual will use and can benefit from that assistance all need to be considered.

The most direct approach to expressing person-situation congruency with the ILS in individual cases would be to establish normative data for ILS subscales in a variety of environmental living arrangements. This approach would determine the mean subscale scores for individuals who are manifesting adequate functional adaptation to completely independ­ent residential and financial situations, to settings that provide varying degrees of assistance, and finally for individuals who have adapted only under situations of extreme assistance and dependency. An examinee's ILS subscale scores then could be compared to the norms for these various settings. For example, a setting in which residents have ILS scores notably greater than those of the examinee might be seen as making demands that are incongruent with the abilities of the examinee. Examinees with scores more congruent with individuals in highly protective settings may be perceived as more likely to be in need of some form of guardianship, conservatorship, or other assistance in the relevant areas of functioning suggested by their poorer scores on certain ILS subscales.

Another method for expressing person-situation congruency with the ILS would be to develop methods for assessing and describing environments using dimensions that are parallel to various components of the ILS. Thus an elderly individual's particular circumstances might be described as demanding greater or less ability to manage money, to respond to emergencies, and so forth. The environment's demands could then be compared to the degrees of ability that the elderly person would bring to that situation. The development of this type of assessment system is discussed further in the third section of this chapter.

References

(a) Anderten, P. (1979). The elderly, incompetency, and guardianship. Unpublished master's the­sis, St. Louis University.

(b) Caul, J. (1984). The predictive utility of the Community Competence Scale. Unpublished doc­

toral dissertation, St. Louis University.

(c) Dunn, T. (1984). Halstead-Reitan Neuropsychological Battery and its prediction of functional daily living skills among geriatricpatients with suspected dementia. Unpublished doctoral dis­sertation, St. Louis University.

(d) Evenson, R. (1976). Geriatric Profile: Manual. Unpublished, Missouri Institute of Psychiatry, St. Louis, MO.

(e) Katz, M., & Lyerly, S. (1963). Methods for measuring adjustment and social behavior in the community: 1. Rationale, description, discriminative validity, and scale develop­ment. Psychological Reports, 13, 503-535.

(f) Loeb, P. A. (1983). Validity of the Community Competence Scale with the elderly. Unpublished doctoral dissertation, St. Louis University.

(g) Loeb, P. A. (1996). Independent Living Scales. San Antonio: Psychological Corp.

(h) Searight, H. R. (1983). The utility of the Community Competence Scale for determining place­ment site among the deinstitutionalized mentally ill. Unpublished doctoral dissertation, St. Louis University.

(i) Searight, H. R., Oliver, J., & Grisso, T. (1983). The Community Competence Scale: Preliminary reliability and validity. American Journal of Community Psychology, 11, 609-613.

(j) Walker, R. (1972). The Brockton Social Adjustment Scale. Diseases of the Nervous System, 33, 542-545.

(k) Zigler, E., & Phillips, L. (1961). Social competence and outcome in psychiatric disorders. Journal of Abnormal and Social Psychology, 63,264-271.