Journal of survey statistics and methodology.
Material type:
- 2325-0984
Item type | Current library | Call number | Status | Barcode | |
---|---|---|---|---|---|
Continuing Resources | PSAU OLM Periodicals | JO JSSM AP2021 (Browse shelf(Opens below)) | Available | JO131 |
Collecting Objective Measures of Visual and Auditory Function in a National in-Home Survey of Older Adults Mengyao Hu and others Journal of Survey Statistics and Methodology, Volume 9, Issue 2, April 2021, Pages 309-334, https://doi.org/10.1093/jssam/smaa044 Abstract Maintenance of visual and auditory function is important for preventing the onset of activity limitations and preserving quality of life in later life. To date, national panel studies focused on health and aging have mostly collected subjective (self-reported) measures of visual and auditory function. The National Health and Aging Trends Study (NHATS), a study of Medicare beneficiaries ages sixty-five and older, recently developed a protocol for measuring objective visual and auditory function for its annual, in-home data collection conducted by trained interviewers. The protocol includes three vision tests-distance and near acuity and contrast sensitivity-and one hearing test-pure-tone audiometry-conducted using a tablet platform with results recorded in a scannable booklet. To identify operational issues and evaluate data quality for the proposed set of vision and hearing tests, NHATS incorporated a pilot study into its 2019 round (N = 417 participants and N = 9 interviewers). Using these pilot study data, the objectives of this paper are to: (1) describe the NHATS protocols to collect objective measures of visual and auditory function; (2) evaluate the quality of the data collected; and (3) assess whether results are influenced by interviewers. We found that respondents were highly likely to participate, with cooperation rates for each test about 90 percent. Data were high quality, with low rates of missingness, test results significantly associated with age and self-reported items, and percentages with poor vision or hearing consistent with prior population-based studies. Objective measures were more likely than self-reports to classify participants as having visual and auditory impairments and had stronger relationships with demographic correlates. Interviewer effects were small and not statistically significant in this small sample. Results of this study have demonstrated that objective visual and auditory functioning can be successfully incorporated into an interviewer-administered home-based protocol.
Differences in Proxy-Reported and Self-Reported Disability in the Demographic and Health Surveys Mahmoud Elkasabi Journal of Survey Statistics and Methodology, Volume 9, Issue 2, April 2021, Pages 335-351, https://doi.org/10.1093/jssam/smaa041 Abstract This article examines the presence, direction, and magnitude of differences between proxy reports and self-reports of disability among adults aged fifteen to fifty-nine years and elderly individuals aged sixty years and older in Demographic and Health Surveys conducted in Uganda, South Africa, and Mali. We use the propensity score weighted multivariate logistic regression to balance the weighted distributions of the covariates between self-reports and proxy reports. Disabilities that have an immediate effect on the communication with others or that require one-to-one help are likely to be over-reported by proxies or under-reported if proxies are not used, especially among the elderly aged sixty years and older. Disabilities that are not observable might be under-reported by proxies.
JSSAM Special Issue on Disability Measurement and Analysis: Preface Kirk Wolter and others Journal of Survey Statistics and Methodology, Volume 9, Issue 2, April 2021, Pages 205-208, https://doi.org/10.1093/jssam/smab016 Health, like disability, is a multidimensional concept having many different meanings. The physical dimension of health is most often assessed in terms of pathology or problems with body structure and function. But health is not merely the absence of pathology or impairment, as ill health is seen as creating obstacles in undertaking desired activities. Functional status, such as in seeing, hearing, walking, communication, and cognition, is the mechanism through which pathology, or the absence of pathology, can impact all aspects of participation in society such as in education, employment, financial well-being, and civic engagement. The critical role of functioning in attaining full participation is highlighted in the United Nations Convention on the Rights of Persons with Disability (UNCRPD) (United Nations General Assembly 2006), the 2030 Agenda for Sustainable Development Goals (SDGs) (United Nations General Assembly 2015), and the Americans With Disability Act (Americans With Disabilities Act of 1990 1990). Each of these initiatives requires that limitations in functioning or disability be accommodated, often through removing environmental barriers and instituting facilitators, so that those with disability can fully participate in all aspects of society. Reliable, valid, and comparable data are needed to formulate policy to improve the lives of those with disability and to monitor whether full inclusion has been attained. The UNCRPD and the SDGs also specifically include requirements for such data collection and active monitoring. While the need for information on disability is well established, obtaining that information has been, and continues to be, challenging. One reason for this is the complex nature of the concept itself (Altman 2001). As defined in the International Classification of Functioning, Disability and Health, disability is an umbrella term that involves the interaction of the individual's functional abilities and the environment in which he or she lives (World Health Organization 2001). The nature of the interaction determines whether full participation in society is achieved. Thus, to fully understand disability, data are needed on the multiple component dimensions and the relationships among them. Functional difficulties can exist, for example, across a range of basic activity domains such as sensory, movement, cognition, and communication. Moreover, each of the domains can be characterized by a range of abilities. The environment is similarly broad, including both the physical or built environment as well as the social environment. In addition, while there is interest in the identification of those with disability and the characteristics of this group, disability and functioning are not inherently dichotomies, but exist along continuums with no external gold standard to help define appropriate cut points on the continuum. Adding to the data collection challenge is the fact that the very word disability is understood very differently by different groups. For some, the term carries great stigma as the term has been used to perpetuate negative stereotypes. For others, it is limited to a few visible physical characteristics or defined by a small set of conditions. Some see it as an expected characteristic of the "natural" aging process. For many, disability has become increasingly understood as a human rights issue. These multiple, and sometimes inconsistent, definitions are embedded in legislation that provides support and services for those with disability and there is wide variation in the definitions that determine eligibility. Despite these complexities, data collections often focus on only one aspect of disability, and yet ascribe the label "disability" without fully addressing how the choice of definition affects the interpretation of the findings. Even when looking at the same component of disability, data collections use different operational definitions depending on the need for the data and the sources used. These data collection and reporting practices have led to a lack of consistency across data sets resulting in the inability to produce a coherent picture of disability in the United States and worldwide. Much methodological work has been undertaken to address disability data collection issues. The improvement of disability statistics has progressed along with the progression of the larger field of survey methodology and includes advances in the evaluation of data collection tools. Well-tested, standard data collection tools have been developed and are being adopted that will lead to more comparable information moving forward (Washington Group on Disability Statistics 2020). There is a greater understanding of the complexities involved in collecting information on disability. The papers in this special edition address selected data collection challenges and add to the body of knowledge that is needed to obtain high-quality information on disability to inform policies and programs. As noted above, disability is not inherently a dichotomy. Rather, individuals experience functioning along a continuum from no difficulty to complete inability. The identification of the group "with disability" and the resulting prevalence of disability depend on the selection of cut points along the functioning continuums. For some objectives, such as to support universal design where the objective is to make environments more broadly accessible and used to the fullest by all people, the cut point could be more inclusive identifying those with less severe levels of functional limitation as having disability. For many policy needs, however, the population with more severe limitations is of interest. This population is small enough that large samples are needed to accurately describe its characteristics, especially if there is interest in cross-classifying disability status with other characteristics such as age, sex, race and ethnicity, urbanicity, socioeconomic status, or geography. Using internet surveys to obtain information on disability is an attractive option given the lower cost of this mode of data collection. The paper by Houtenville, Phillips, and Sundar entitled "Usefulness of Internet Surveys to Identify People with Disabilities: A Cautionary Tale" in this issue investigates the use of internet surveys and finds that such methods can introduce bias if the target population lacks access to or the ability to use the necessary technology. Several of the papers in this issue deal with variations in how disability is conceptualized and measured. The paper by Pettinicchio and Maroto looks at the impact of using different data collection instruments by using IPUMS International Census microdata since 2000 to examine disability measurement across sixty-five countries. Analyzing these data with the Total Survey Error framework in mind, they find that definitions, translation, measurement, and instructions to both respondents and enumerators matter for understanding disability prevalence cross-nationally. The authors recommend that researchers take great care when using compiled cross-national census data to study disability and always consider how disability is defined and measured within surveys. The paper "Collecting objective measures of visual and auditory function in a national in-home survey of older adults" by Hu, Freedman, Ehrlich, Reed, Billington, and Kasper presents the results of a pilot study to address the feasibility of incorporating objective data collection methods and compared the results obtained from objective and subjective measures of vision and hearing. The authors found that respondents were highly likely to participate with low rates of missingness and that test results were significantly associated with age and self-reported measures of hearing and vision limitations. They conclude that objective visual and auditory functioning can be successfully incorporated into an interviewer-administered home-based protocol.
Differences resulting from obtaining information directly from the subject as opposed to from a proxy are investigated in the paper by Elkasabi entitled "Differences in proxy-reported and self-reported disability in the Demographic and Health Surveys." Based on data from Uganda, South Africa, and Mali, propensity score-weighted multivariate logistic regression models are used to balance the weighted distributions of the covariates between self and proxy reports. Disabilities that have an immediate effect on the interaction with others or that require one-to-one help are likely to be under-reported by self-reports, especially among the elderly age 60 and above, whereas disabilities that are not observable might be under-reported by proxies. The paper by Flaherty and Shono, "Parsimonious restricted latent class models for improved measurements of activities of daily living" presents a restricted latent class approach to summarize the multidimensional aspects of disability as measured by ten binary questions on activities of daily living. The authors suggest a restricted fourteen-class model to better capture heterogeneous manifestations of disability. Despite the larger number of classes, this model contains fewer parameters and has smaller measurement error than the unrestricted four-class latent class model. The suggested constrained latent class specification may be useful for practitioners interested to conduct subgroup analyses aimed at, for example, studying treatment responses. Finally, the relationship between the functional status and the environment is addressed in the paper, "Who is at risk of workforce exit due to disability? State differences in 2003-2016" by Ben-Shalom,
Many Classes, Restricted Measurement (MACREM) Models for Improved Measurement of Activities of Daily Living Brian P Flaherty and Yusuke Shono Journal of Survey Statistics and Methodology, Volume 9, Issue 2, April 2021, Pages 231-256, https://doi.org/10.1093/jssam/smaa047 Abstract Scientists use latent class (LC) models to identify subgroups in heterogeneous data. LC models reduce an item set to a latent variable and estimate measurement error. Researchers typically use unrestricted LC models, which have many measurement estimates, yet scientific interest primarily concerns the classes. We present highly restricted LC measurement models as an alternate method of operationalization. MACREM (Many Classes, Restricted Measurement) models have a larger number of LCs than a typical unrestricted model, but many fewer measurement estimates. Goals of this approach include producing more interpretable classes and better measurement error estimates. Parameter constraints accomplish this structuring. We present unrestricted and MACREM model results using data on activities of daily living (ADLs) from a national survey (N = 3,485). We compare a four-class unrestricted model with a fourteen-class MACREM model. The four-class unrestricted model approximates a dimension of functional limitation. The fourteen-class model includes unordered classes at lower levels of limitation, but ordered classes at higher levels of limitation. In contrast to the four-class model, all measurement error rates are reasonably small in the fourteen-class model. The four-class model fits the data better, but the fourteen-class model is more parsimonious (forty-three versus twenty-five parameters). Three covariates reveal specific associations with MACREM classes. In multinomial logistic regression models with a no limitation class as the reference class, past 12-month diabetes only distinguishes low limitation classes that include cutting one's own toenails as a limitation. It does not distinguish low limitation classes characterized by other common limitations. Past 12-month asthma and current disability status perform similarly, but for heavy housework and walking limitation classes, respectively. These limitation-specific covariate associations are not apparent in the unrestricted model analyses. Identifying such connections could provide useful information to advance theory and intervention efforts.
Risk of Workforce Exit due to Disability: State Differences in 2003-2016 Yonatan Ben-Shalom and others Journal of Survey Statistics and Methodology, Volume 9, Issue 2, April 2021, Pages 209-230, https://doi.org/10.1093/jssam/smab005 Abstract A better understanding of trends in workforce exit due to disability and how these trends vary across states and subgroups can help federal and state policymakers identify both individual-level and state-level factors associated with increased risk of workforce exit due to disability. Using national survey data and Bayesian multilevel modeling techniques, we produce yearly estimates of trends in the risk of workforce exit due to disability for states and demographic subgroups. These estimates are more stable and have narrower uncertainty intervals than estimates produced using classical statistical methods. We identify Current Population Survey respondents as being "newly at-risk" of exiting the workforce due to disability if they reported being employed in one month and out of the labor force because of a disability in the next month, and we refer to their share of the working-age population as the "at-risk rate." We find that age, education, race, and gender are important factors for the at-risk rate, in decreasing order. Holding demographics constant across states and time reduces the cross-state variation in the at-risk rate but does little to reduce variability over time. State at-risk rates are typically higher than application rates for the Social Security Administration's disability programs, but the relationship between these rates varies considerably by state. Our preliminary exploration of the reasons for cross-state variation in this relationship suggests that differences across states may be due to differences in their industrial composition, job opportunities, and safety net structure.
Usefulness of Internet Surveys to Identify People with Disabilities: A Cautionary Tale Andrew J Houtenville and others Journal of Survey Statistics and Methodology, Volume 9, Issue 2, April 2021, Pages 285-308, https://doi.org/10.1093/jssam/smaa045 Abstract Disability is an important characteristic to consider in survey research. However, people with disabilities are a hard-to-reach population. Internet survey methods offer tremendous potential to expand researchers' ability to reach and learn about people with disabilities. The goal of this study is to examine potential bias when using nonprobability Internet samples to investigate demographics and socioeconomic outcomes of people with disabilities. We compare the findings based on a national employment and disability survey instrument fielded to four samples: (1) a random-digit dial (RDD) sample, (2) a prescreened sample from a nonprobability Internet access panel, for which screening was based on the presence of 139 previously reported health conditions, (3) an unscreened sample from another nonprobability Internet access panel (without previously prescreened health conditions), and (4) a mixed nonprobability self-recruited (river and snowball) sample. Each sample was weighted on four demographic variables (gender, age, race/ethnicity, and region) using benchmarks from the American Community Survey (ACS). Three dichotomous outcome variables of interest (level of education, household income, and current employment status) were contrasted with weighted population estimates from the ACS. Results showed that the sample resulting from the RDD and all three nonprobability Internet samples differed significantly from ACS population estimates on all three outcome variables. Reweighting to include type of functional disability did not significantly reduce dissimilarities with ACS for any of the four samples. Nonprobability Internet survey methods offer relatively low-cost, easy-to-use avenues for disability-related research. Yet, researchers must proceed with caution to reduce or avoid known sources of bias in both the methodology and the interpretation of results.
Who Counts? Measuring Disability Cross-Nationally in Census Data David Pettinicchio and Michelle Maroto Journal of Survey Statistics and Methodology, Volume 9, Issue 2, April 2021, Pages 257-284, https://doi.org/10.1093/jssam/smaa046 Abstract Despite established recommended standard definitions, measures, and methods by the UN Washington Group on Disability Statistics and the International Classification of Functioning, Disability and Health (ICF) to assess dimensions of disability, national censuses vary widely in the questions used to identify people with disabilities. Although many seek to conform ex-ante to ICF definitions, they also deviate from this basic framework in different ways. This complicates ex-post harmonization and standardization for cross-national comparisons of disability prevalence and outcomes influenced by disability status, such as labor market participation. Addressing these issues, this study uses IPUMS International Census microdata since 2,000 to examine disability measurement across 65 countries. We find that definitions, terminology, measurement, and instructions to both respondents and enumerators matter for understanding disability prevalence cross-nationally. For instance, questions that included potentially stigmatizing language were associated with lower rates of disability reporting, but questions that listed specific limitations were associated with higher rates. Beyond disability, our findings also speak more broadly to ongoing challenges in survey harmonization for cross-national comparison.
There are no comments on this title.