< - - Go Back

Methods and Measurement abstracts and tables:

Journal of Mental Health and Aging, Vol. 7, No. 1, 2001, pp. 91-118

Cognitive Assessment Among Minority Elderly: Possible Test Bias

 Mildred Ramirez, PhD

Jeanne A. Teresi, EdD, PhD

Stephanie Silver, MPH

Douglas Holmes, PhD

Barry Gurland, MD

Rafael Lantigua, MD  

Eighteen studies examining the performance of cognitive screening measures across samples representing ethnically diverse populations in terms of sensitivity and specificity with respect to a clinical diagnosis are summarized.  Studies were identified through a search of four databases (MEDLINE, CINAHL, HealthGate, PsycINFO) and presented in terms of: a) method of data collection and clinical criteria, b) sampling methods, response rate and sample characteristics, c) methods of analysis, and d) recommendations of the investigators.  Methodological caveats relevant to the interpretation of sensitivity and specificity across studies are discussed; scales showing promise for culture-fair assessment are highlighted. [PUBLICATION ABSTRACT] 

Back to Top - - No associated table.

Journal of Mental Health and Aging, Vol. 2, No. 3, 1996

 Defining And Measuring Outcomes In Research On Alzheimer’s Disease And Related Chronic Conditions: An Interdisciplinary Communication And Collaborative Challenge

Peter J. Whitehouse, MD, PhD

Abstract:  

This two-part special issue of the Journal of Mental Health and Aging provides an excellent review of measures used to assess individuals in chronic care settings.  This work is timely as we seek to enrich our understanding of the how such individuals live their lives and whether our environments and programs satisfactorily address their needs and enhance their quality of life.  In the later stages of dementia, such individuals, often are neglected by the broader research community.  This special issue represents not just a summary of current approaches but a step in the evolution of our understanding of assessment including that of important noncognitive symptoms; such symptoms are perhaps more disabling than the purely cognitive symptoms, and frequently contribute to impairment in quality of life for both patient and caregivers.

Back to Top - - No associated table.

Journal of Clinical Epidemiology, 1995, Vol. 48, No. 4, pp. 473-483

 Item Bias In Cognitive Screening Measures: Comparisons Of Elderly White, Afro-American, Hispanic And High And Low Education Subgroups

 Jeanne A. Teresi

Robert R. Golden

Peter Cross

Barry Gurland

Marjorie Kleinman

David Wilder

 A study of item bias in standard cognitive screening measures was conducted in a sample of Afro-American, Hispanic and non-Hispanic white elderly respondents who were part of a dementia case registry study. The methods of item-response theory were applied to identify biased items.  Both cross-cultural and high and low education groups were examined to determine which items were biased.  Out of 50 cognitive items examined from six widely used cognitive screening measures, 16 were identified as biased for either high and low education groups or ethnic/racial group membership. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Alzheimer=s Disease and Associated Disorders, Vol. 8, Suppl. 1, pp. S291-S298, 1994

Latent Structure Methods For Estimating Item Bias, Item Validity And Prevalence Using Cognitive And Other Geriatric Screening Measures

 Jeanne A. Teresi

Robert R. Golden

Screening measures often are used to assess chronic care population, frequently in large multisite or multicenter studies.  They are used to estimate prevalence in epidemiological surveys, identify risk factors for disease, identify cases and noncases for research purposes, and target clinical, programmatic, and pharmacological interventions.  For example, screening measures can be useful in identifying nursing home residents who might benefit from different types of dementia-specific special care.  The cost of misclassification error varies according to use.  A typical problem in screening is the high cost of false positives, persons who subsequently receive elaborate laboratory diagnostic measures based on erroneous screening results, or those allocated to additional and costly health services who can equally benefit from regular care.  On the other hand, missing cases of certain types of illness can result in death or significant morbidity or excess disability.  Those not identified might miss out on care that can be therapeutic.  Thus, methods of minimizing misclassification errors are critical.  This article presents some examples of the effects of item bias and discusses some models that can be used to maximize unbiased classifications.  An unbiased indicator is one that behaves in the same manner (defined later) and is inferred to be measuring the same underlying attribute across comparison subgroups, and has the same valid-positive and false-positive rates (the same sensitivities and specificities) for each subgroup.   [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Alzheimer Disease and Associated Disorders. An International Journal, Vol. 11, Suppl 6, Pp. 30-44, 1997      

 Measurement Choices In Multi-Site Studies Of Outcomes In Dementia

Kathleen A. Smyth

Steven H. Ferris

Patrick Fox

Albert Heyman

Douglas Holmes

John N. Morris

Charles D. Phillips

Richard Schulz

Jeanne Teresi

Peter J. Whitehouse

 This paper summarizes the measurement choices made by selected current or recently completed multi-site projects with a common emphasis on measuring outcomes in dementia.  Information on number of items and scoring, reason(s) for selecting the measure, and reliability and validity of the measure (either citations providing this information or a report of pertinent unpublished findings) is presented for eight domains: cognition, behavioral symptoms, physical health status, physical functioning and self-care abilities, quality of life, family/staff caregiver outcomes, service use, and cost.  We found considerable reliance on the published literature as a guide to measurement choice, motivated largely by measures= superior psychometric properties, their ubiquity in the literature, and/or their brevity or ease of use.  There is still evidence of Astarting from scratch@ in some domains, however.  To the extent that these projects reflect the state of the art in dementia-relevant outcomes research, we conclude that comparison of outcomes across studies will continue to be problematic.  However, as long as dissemination of methodological as well as substantive findings continues to characterize outcome studies in dementia, there is hope that a more congruent view of how to assess key outcomes in dementia will emerge. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Alzheimer=s Disease and Associated Disorders, Vol. 8, Suppl. 1, pp. S144-S183, 1994

 Measurement Issues In Chronic Care Populations: Dementia Special Care

 Jeanne Teresi

M. Powell Lawton

Marcia Ory

Douglas Holmes

 This article reviews possible measurement problems in conducting research involving individuals with dementia and presents conceptual issues related to measurement of different domains and constructs as they relate to dementia care.  Evaluation of dementia special care units (SCUs) poses special measurement problems that can affect the ability to determine the effectiveness of different types of care.  Poor measurement will reduce the chances of detecting an intervention effect, if one exists.  Paramount among these measurement problems are difficulties in assessing persons with severe cognitive deficits and problems in defining and measuring parameters of specialized care.  Because little measurement research has been conducted with persons with dementia or in dementia SCUs, this article will draw heavily on what is known about measurement issues in other chronic care populations. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Health Services Research, 2005, Vol. 40, No. 5, 1640-1657

 Measurement Issues In Health Disparities Research

 M. Ramirez, Ph.D.

M. Ford, Ph.D.

A. Stewart, Ph.D.

J. Teresi, Ed.D., Ph.D.

Background:  Racial and ethnic disparities in health and health care have been documented; the elimination of such disparities is currently part of a national agenda.  In order to meet this national objective, it is necessary that measures identify accurately the true prevalence of the construct of interest across diverse groups.  Measurement error might lead to biased results, e.g., estimates of prevalence, magnitude of risks, and differences in mean scores.   Addressing measurement issues in the assessment of health status may contribute to a better understanding of health issues in cross-cultural research.

Objective:   To provide a brief overview of issues regarding measurement in diverse populations.

Findings: Approaches used to assess the magnitude and nature of bias in measures when applied to diverse groups include qualitative analyses, classic psychometric studies, as well as more modern psychometric methods.  These approaches should be applied sequentially, and/or iteratively during the development of measures.

Conclusions: Investigators performing comparative studies face the challenge of addressing measurement equivalence, crucial for obtaining accurate results in cross-cultural comparisons. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Journal of Mental Health and Aging, 1996,  Vol. 2, No. 3, pp. 147-148

Measurement Of Older Chronic Care Populations

Jeanne A. Teresi, EdD, PhD

M. Powell Lawton, PhD

Douglas Holmes, PhD

Marcia Ory, PhD

 This two-part special issue of the Journal of Mental and Aging contains reviews of measures that can be used in chronic care settings among individuals whose assessment is complicated because of factors, such as frailty, perceptual and communication problems, and advanced cognitive impairment. [PUBLICATION ABSTRACT]  

Back to Top - - No associated table.

Alzheimer Disease and Associated Disorders, 1997,  Vol. 11, Suppl. 6, pp. 146-155

Methodological Issues In Cognitive Assessment And Their Impact On Outcome Measurement

Jeanne A. Teresi

 Douglas Holmes

 Assessment of cognitive impairment in chronic care populations is complicated by several factors that may interfere with the assessment process: physical frailty and disability; comorbid conditions such as depression; and decrements in vision, hearing, speech, and general communication.  Moreover, cognitive impairment itself affects assessment of outcome domains such as depression, behavior, and function, thus contributing to several sources of measurement bias.  Sources of bias are discussed in the context of findings from the literature relating individual and cognitive factors to outcome measurement.  Recommendations for further methodological research are provided. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Statistics in Medicine, 2000, 19, pp. 1593-1606

Modeling The Decline Pattern In Functional Measures From A Prevalent Cohort Study

Xinhua Liu

Jeanne Teresi

Christine Waternaux

In studying decline among cognitively impaired people, a prevalent cohort study design is commonly used to account for entry into the study at different levels of impairment.  The data set typically consists of many short series of repeated measurements collected over time.  However, the time origin, such as time of disease/impairment onset, is often uncertain.  In order to model non-linear decline patterns in functional test scores and associated risk factors with such data, we propose two approaches as alternatives to Liu et al. One approach models change over adjacent visits with varying time intervals.  The second models the change since baseline using a random effect for heterogeneity of change.  We used these two approaches to examine the decline in cognitive test scores among special care unit (SCU) and non-SCU residents at the New York sites of the National Institute on Aging (NIA) collaborative studies of special dementia care.  The analyses suggest that, controlling for several covariates, SCU residents experienced more rapid cognitive decline than did non-SCU residents.  The relative advantages and disadvantages of the two models are discussed.  [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

The American Journal of Geriatric Psychiatry, 1995,  Vol. 3, No. 2, pp. 96-107

 Operating Characteristics Of Brief Screens For Dementia In A Multicultural Population

 David Wilder, Ph.D.

Peter Cross, M. Phil.

Jiming Chen, Ph.D.

Barry Gurland, M.D.

Rafael A. Lantigua, M.D.

Jeanne Teresi, Ph.D.

Mabel Bolivar, B.A.

Priscilla Encarnacion, B.A.

 Operating characteristics of seven screens for dementia were compared across various groups for 795 persons who had received a criterion diagnostic evaluation.  Area under the curve (AUC), based on receiver operating characteristics, was compared between and within scales as an indication of their efficiency.  Differences in AUC were only 5% across all the scales for the entire sample, but increased to 11% across sociocultural groups and scales and to 20% across education groups and scales.  Two scales (the Mini-Mental State Exam and the Short Portable Mental Status Questionnaire) misclassified most nondementias for the entire sample, and all scales misclassified most nondementias among persons with less than 5 years of education. Findings could support a recommendation that certain shorter scales be used because they perform as well as longer ones, are more consistent across cultural and educational groups, and can be more easily modified to improve performance in culturally diverse populations. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Annual Review of Gerontology and Geriatric: Focus on Assessment Techniques, 1994, Springer, 14, pp. 1-22

Overview Of Methodological Issues In Gerontological And Geriatric Measurement

Jeanne A. Teresi

Douglas Holmes

This chapter provides an overview of methodological issues raised by different authors with respect to constructs and measures presented in this volume. While some measurement problems (e.g., response bias) may apply to specific age cohorts, we argue that many measurement problems are not necessarily a function of the age cohort assessed, but of the varying distributions, rates, or ratios of the underlying attribute when measured across different samples of elderly, ranging from the normal well to the chronically impaired.  Thus, although the focus of this review is on assessment of older persons, the concepts presented and the statistical problems and methods discussed are relevant to the entire age spectrum. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

International 

Journal of Mental Health and Aging, 2001, Vol. 7, No. 1, pp. 13-19

    Some Methodological Guidelines For Cross-Cultural Comparisons

 Jeanne A. Teresi, EdD, PhD

Douglas Holmes, PhD

 The purpose of this article is to review and revisit some methodological issues of relevance to cross-cultural research. Companion articles by Liang and Teresi in this volume discuss statistical invariance issues and the role of Confirmatory Factor Analyses and Item Response Theory in such research. Several guidelines related to interpretation of reliability and validity coefficients that are well addressed in the measurement and biostatistics literature, but that are frequently ignored in common practice are also discussed. Some of these guidelines are based on conclusions presented in the psychometric literature over 70 years ago, but have since been forgotten (or were never learned). [PUBLICATION ABSTRACT]    

Back to Top - - No associated table.

Journal of Clinical Epidemiology, 1997, Vol. 50, No. 2, pp.175-184

Reporting Source Bias In Estimating Prevalence Of Cognitive Impairment

Jeanne A. Teresi

Douglas Holmes

Reporting source bias was examined using cognitive data from a probability sample of adult day health care (ADHC) clients.  Data were obtained from four sources: client, family, staff, and chart. These data suggest that prevalence estimates are influenced by the reporting source and method of assessment, with chart diagnoses yielding the least convergent estimates.  Examination of agreement summary statistics and bias indices show that both staff and family underreport prevalence of cognitive impairment, but that more bias is associated with staff than is with family sources.  Such bias should be considered in studies of prevalence estimation of cognitive impairment. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.

Neuroepidemiology, 1999, 18, pp. 309-321

Use Of Latent Class Analyses For The Estimation Of Prevalence Of Cognitive Impairment, And Signs Of Stroke And Parkinson=S Disease Among African-American Elderly Of Central Harlem: Results Of The Harlem Aging Project

 Jeanne A. Teresi

Steven M. Albert

Douglas Holmes

Richard Mayeux

A probability sample (response rate of 79.4%) of 164 African-American elderly residing in central Harlem was assessed using several standard cognitive screening measures and scales measuring stroke and Parkinson=s symptoms, respectively.  Both traditional cutting scores and the methods of latent class analysis were used to estimate point prevalence.  The latent class prevalence ratio estimates range from 10 to 14%.  The confidence intervals for cognitive impairment ranged from 5 to 18% with an average estimate of 12%.  The latent class estimate for stroke effects was 11% and for two measures of neurological signs associated with Parkinson=s disease 7 and 12%.  About 2-3% of the sample were communication disordered, and 5% had significant ambulation disorder. These results indicate that a relatively large proportion of elderly community residents of central Harlem evidenced or reported health problems, any or all of which may constitute a public health challenge. [PUBLICATION ABSTRACT]

Back to Top - - No associated table.