H. Jane Rogers

Validation of a new measure of concept of a good death

BACKGROUNDnThe concept of a good death is central to end-of-life care research. Despite its importance and the high interest in the topic, there are few measures currently available for use in clinical research.nnnPURPOSEnThe present work describes the development and testing of a set of items intended to measure the importance of several components posited to be critical to the concept of a good death. It is intended for use with health care providers and lay people in the context of end-of-life care research and education.nnnPOPULATIONnFour cohorts (n = 596) were recruited to participate, representing two helping profession disciplines, nonhelping professionals, and a range of ages, specifically: (1) undergraduate medical students; (2) masters degree students in nursing; (3) graduate students from the life sciences; and (4) practicing hospice nurses.nnnMETHODSnParticipants completed self-report questionnaires at baseline and retest. Psychometric analyses included item frequency distributions, factor analysis, alpha reliability, intraclass correlation, and measures of association.nnnRESULTSnThe new Concept of a Good Death measure demonstrated good item frequency distributions, acceptable internal consistency reliability, and test-retest stability. Its factor structure revealed that three distinct domains are measured, reflecting the psychosocial/spiritual, physical, and clinical aspects of a good death. An examination of patterns of correlations showed differential associations with death anxiety, spiritual beliefs and practices, anxious mood, and sociodemographic characteristics.nnnCONCLUSIONSnThe new Concept of a Good Death instrument appears to measure three distinct factors which people consider important to a Good Death. Ratings of the importance of these factors are reliable and valid. The instrument has the advantage of being a brief, self-report index for use in end-of-life care research.

An effect size measure and Bayesian analysis of single-case designs

This article describes a linear modeling approach for the analysis of single-case designs (SCDs). Effect size measures in SCDs have been defined and studied for the situation where there is a level change without a time trend. However, when there are level and trend changes, effect size measures are either defined in terms of changes in R(2) or defined separately for changes in slopes and intercept coefficients. We propose an alternate effect size measure that takes into account changes in slopes and intercepts in the presence of serial dependence and provides an integrated procedure for the analysis of SCDs through estimation and inference based directly on the effect size measure. A Bayesian procedure is described to analyze the data and draw inferences in SCDs. A multilevel model that is appropriate when several subjects are available is integrated into the Bayesian procedure to provide a standardized effect size measure comparable to effect size measures in a between-subjects design. The applicability of the Bayesian approach for the analysis of SCDs is demonstrated through an example.

Advances in criterion-referenced measurement.

One of the major changes in the testing field over the last 20 years has been the increased interest in and use of criterion-referenced tests (CRT). Criterion-referenced tests provide a basis for assessing the performance of examinees in relation to well-defined domains of content rather than in relation to other examinees, as with norm-referenced tests. Criterionreferenced tests are now widely used (1) in the armed services, to assess the competencies of servicemen; (2) in industry, to assess the job skills of employees and to evaluate the results of training programs; (3) in the licensing and certification fields, to distinguish “masters” from “nonmasters” in over 900 professions in the United States alone; and (4) in educational settings such as schools, colleges, and universities, to assess the performance levels of students on competencies of interest.

21 Assessing the Fit of Item Response Theory Models

Publisher Summary Item response theory (IRT) provides a framework for modeling and analyzing item response data. Assessing IRT model fit to item response data is one of the crucial steps before an IRT model can be applied with confidence to estimate proficiency or ability levels of examinees, to link tests across administrations, and to assess adequate yearly progress. Assessing model fit is an important part of the test validation process. It is multifaceted and, as in the verification of any scientific theory, it is an ongoing process where only through the accumulation of empirical evidence can one be confident of the appropriateness of IRT for the solution of a particular measurement problem. The procedures for assessing model–data fit described in this chapter have the potential for addressing the vexing problem of determining if the measurement procedures used are appropriate for addressing the practical measurement problems faced by practitioners.

Regression models and effect size measures for single case designs

A regression modelling approach for the analysis of single case designs (SCDs) is described in this paper. The approach presented addresses two key issues in the analysis of SCDs. The first issue is that of serial dependence among the observations in SCDs. The second issue is that of an effect size measure appropriate for SCDs. As with traditional between-subjects experimental designs, effect size measures are critical in assessing the impact of interventions in SCDs. Although effect size measures when there is level change without trend are straightforward to obtain and have been well studied, the situation is different when there are changes in both level and trend. An effect size measure that combines changes in levels and slopes and that is comparable to the d-type effect size measure obtained in between-subjects designs is presented. Finally, an inferential procedure for assessing the effect of the intervention based on the effect size measure is provided and illustrated.

Development and testing of a new instrument for measuring concerns about dying in health care providers.

A new measure of concerns about dying was investigated in this psychometric study. The Concerns About Dying instrument (CAD) was administered to medical students, nursing students, hospice nurses, and life sciences graduate students ( N = 207) on two occasions; on one occasion they also completed three related measures. Analyses included descriptive statistics, factor analysis, Cronbach’s alpha, test-retest correlations, t tests, and correlations with other measures. Results suggest the CADmeasures three distinct but related areas: general concern about death, spirituality, and patient-related concern about death. Reliabilityestimates were good, and correlations with related measures were strong. Between-group differences suggest scores are related to actual differences in level of concern and beliefs about death and dying. The CAD has the advantage of being very brief, and of explicitly assessing concerns about working with patients who are dying.

The Cancer Message Literacy Tests: Psychometric analyses and validity studies

OBJECTIVEnTo examine the psychometric properties of two new health literacy tests, and to evaluate score validity.nnnMETHODSnAdults aged 40-71 completed the Cancer Message Literacy Test-Listening (CMLT-Listening), the Cancer Message Literacy Test-Reading (CMLT-Reading), the REALM, the Lipkus numeracy test, a brief knowledge test (developed for this study) and five brief cognitive tests. Participants also self-reported educational achievement, current health, reading ability, ability to understand spoken information, and language spoken at home.nnnRESULTSnScore reliabilities were good (CMLT-Listening: alpha=.84) to adequate (CMLT-Reading: alpha=.75). Scores on both CMLT tests were positively and significantly correlated with scores on the REALM, numeracy, cancer knowledge and the cognitive tests. Mean CMLT scores varied as predicted according to educational level, language spoken at home, self-rated health, self-reported reading, and self-rated ability to comprehend spoken information.nnnCONCLUSIONnThe psychometric findings for both tests are promising. Scores appear to be valid indicators of comprehension of spoken and written health messages about cancer prevention and screening.nnnPRACTICE IMPLICATIONSnThe CMLT-Listening will facilitate research into comprehension of spoken health messages, and together with the CMLT-Reading will allow researchers to examine the unique contributions of listening and reading comprehension to health-related decisions and behaviors.

Factors influencing preceptors' responses to medical errors: a factorial survey.

Background Preceptors must respond to trainees’ medical errors, but little is known about what factors influence their responses. Method A total of 115 primary care preceptors from 16 medical schools responded to two medical error vignettes involving a trainee. Nine trainee-related factors were randomly varied. Preceptors indicated whether they would discuss what led to the error, provide reassurance, share responsibility, express disappointment, and adjust their written evaluation of the trainee. Results Almost all preceptors would discuss what led to the error; relatively few would express disappointment. The trainee’s prior history of errors, knowledge level relative to peers, receptivity to feedback, training level, emotional reaction, offering to apologize, and offering an excuse were predictive of preceptors’ responses; gender and time-in-office were not. Conclusion This study identified seven trainee-related factors as predictive of preceptors’ responses to medical errors. More research is needed to identify other influential factors, and to improve teaching from medical errors.

Semantic and Phonological Ability to Adjust Recoding: A Unique Correlate of Word Reading Skill?

ABSTRACT This study addresses whether reading involves a process termed semantic and phonological ability to adjust recoding (SPAAR). It was hypothesized that SPAAR helps readers link inaccurate pronunciations to lexical entries (e.g., spynitch to spinach). Psychometric properties of the Mispronunciation Correction Task (MCT), a measure of SPAAR, were investigated using a sample of 206 elementary-age children. The 25-item MCT appeared to measure a single construct, and the items discriminated well across the range of ability. Based on a sample of 94 children, construct validity analyses indicated that phonological awareness and decoding skill predicted MCT performance. Hierarchical regressions were used to test the relationship between SPAAR and word reading. After controlling for other skills, SPAAR showed a relationship with both timed and untimed word reading. Phonological awareness did not. These data suggest that SPAAR may be a reading-related ability involving phonological processing and some form of semantic processing.

The Role of Meta-analysis in Educational Research

Meta-analysis is now a well-established form of synthesizing literature, and syntheses of meta-analyses are now being produced in educational research. The chapter provides an overview of the method, discusses syntheses of meta-analyses, and argues that these methods can be of particular assistance for establishing benchmarks of comparison about educational outcomes. The literature relating to influences on achievement is used to explore the issues.