Comparison of measuring devices

Although it is common to determine if two devices give the same results, we apply an affine transformation to make the two devices the same on the average. The device with a lower variability in its measurement results is more accurate. Repeated observations are required. The design is balanced if each object is measured the same number of times by each device. The properties this design yields justify this approach.     

Psychometric properties of the Transitions from Foster Care Key Leader Survey

This study summarizes the development and piloting of the Transitions from Foster Care Key Leader Survey (TFC-KLS), an instrument designed to measure change in systems serving young people transitioning from foster care to adulthood. The Jim Casey Youth Opportunity Initiative’s logic model was used as a basis for instrument development. The instrument was piloted with 119 key leaders in six communities. Seven of eight latent scales performed well in psychometric testing. The relationships among the 24 measures of system change were explored. A CFA testing overall model fit was satisfactory following slight modifications. Finally, a test of inter-rater reliability between two raters did not find reliable reporting of service availability in a supplemental portion of the survey. The findings were generally positive and supported the validity and utility of the instrument for measuring system change, following some adaptations. Implications for the field are discussed.     

Assessing participants’ perceptions on group-based principles for action in community-based health enhancing physical activity programmes: The APEF tool

In community-based health enhancing physical activity (CBHEPA) programmes, group-based principles for action such as active participation, enjoyment, and fostering group processes are widely advocated. However, not much is known about participants’ perceptions of these principles as there are no assessment tools available. Therefore, this article describes the development of the APEF (Active Participation, Enjoyment, and Fostering group processes) tool and reports on its implementation in a Dutch CBHEPA programme. Indicators for the principles have been identified from literature research, interviews with professionals, and secondary analysis of three group interviews with 11 practitioners. To address the identified indicators, the APEF tool was developed, pretested, and used in 10 focus groups with 76 participants.The APEF tool consists of eight statements about group-based principles for action, on which CBHEPA participants vote, followed by in-depth discussion. The voting procedure engages participants. Spider diagrams visualise participants’ perceptions of group-based principles.The APEF tool addresses the challenge of relating group level outcomes to individual outcomes such as physical activity behaviour. The tool facilitates as well as evaluates group-based principles for action, it stimulates dialogue and is culturally sensitive, but it needs strong facilitating skills to manage group dynamics.     

Justice involvement among homeless individuals with mental illnesses: Are self-report and administrative measures comparable?

Individuals with mental illnesses who experience homelessness have frequent interactions with the criminal justice system. Correctly measuring this involvement is essential to develop and evaluate the efficacy of intervention programs. Criminal justice involvement is typically assessed through administrative records or self-reported accounts. The aims of this study are to: 1) assess agreement between self-report and administrative data related to court appearances, and 2) identify individual characteristics that affect discrepancies between sources. Participants were 468 homeless persons with mental illness from the Montreal site of the At Home/Chez Soi randomized controlled trial, in Canada. Self-reported data was collected through an interviewer-administered questionnaire. Administrative data was collected through provincial and municipal court databases. Overall, agreement was good. Discrepancies were more common among those with a diagnosis of mood disorder with psychotic features, and those with a criminal history. Increased age and interviewer’s perception of sincerity and interest increased likelihood of concordance. Generally, high agreement between self-report and administrative data suggests that either source can provide reliable information. Further work to understand predictors of discrepancies could further enhance the quality of data collected through these different sources.     

Estimation in multiple linear regression Berkson model for processes with uncorrelated increments

This paper presents a method of estimating the regression and variance parameters in the multiple linear regression Berkson model for a continuous-time stochastic process with uncorrelated increments. Under minimal conditions, we establish (i) the Gauss–Markov theorem and the quadratic mean—as well as the strong consistency of the proposed estimate of the regression parameter and (ii) the weak consistency of the proposed estimate of the variance parameter.     

On the usefulness of knowledge of error variances in the consistent estimation of an unreplicated ultrastructural model

This article considers an unreplicated ultrastructural model and discusses the asymptotic properties of three consistent estimators of slope parameter arising from the knowledge of measurement error variances. Conditions are deduced when knowing the error variances associated with both the study and the explanatory variables is more/less beneficial than using a single error variance in the formulation of slope estimators.     

The Method to Identify a Biomarker and to Evaluate Its Efficiency for Survival by Using the Joint Model of the Accelerate Failure Time and Longitudinal Data

In this article, we discuss how to identify longitudinal biomarkers in survival analysis under the accelerated failure time model and also discuss the effectiveness of biomarkers under the accelerated failure time model. Two methods proposed by Shcemper et al. are deployed to measure the efficacy of biomarkers. We use simulations to explore how the factors can influence the power of a score test to detect the association of a longitudinal biomarker and the survival time. These factors include the functional form of the random effects from the longitudinal biomarkers, in the different number of individuals, and time points per individual. The simulations are used to explore how the number of individuals, the number of time points per individual influence the effectiveness of the biomarker to predict survival at the given endpoint under the accelerated failure time model. We illustrate our methods using a prothrombin index as a predictor of survival in liver cirrhosis patients.     

Two Pitfalls of Using Standard Regression Diagnostics When Both X and Y Have Measurement Error

Modern exploratory data analysis produces models that are not based on physical theory but that are consistent with pictures of the data. When both X and Y have error this can be risky, because important features are hidden. Two examples are given that show that systematic model departures and heteroscedasticity may not be detectable with standard regression diagnostics.     

Partial Identification of Economic Mobility: With an Application to the United States

Abstract

The economic mobility of individuals and households is of fundamental interest. While many measures of economic mobility exist, reliance on transition matrices remains pervasive due to simplicity and ease of interpretation. However, estimation of transition matrices is complicated by the well-acknowledged problem of measurement error in self-reported and even administrative data. Existing methods of addressing measurement error are complex, rely on numerous strong assumptions, and often require data from more than two periods. In this article, we investigate what can be learned about economic mobility as measured via transition matrices while formally accounting for measurement error in a reasonably transparent manner. To do so, we develop a nonparametric partial identification approach to bound transition probabilities under various assumptions on the measurement error and mobility processes. This approach is applied to panel data from the United States to explore short-run mobility before and after the Great Recession.