Asymptotically Correct Person Fit z-Statistics For the Rasch Testlet Model.

A well-known person fit statistic in the item response theory (IRT) literature is the l z statistic (Drasgow et al. in Br J Math Stat Psychol 38(1):67-86, 1985). Snijders (Psychometrika 66(3):331-342, 2001) derived l z ∗ , which is the asymptotically correct version of l z when the ability parameter is estimated. However, both statistics and other extensions later developed concern either only the unidimensional IRT models or multidimensional models that require a joint estimate of latent traits across all the dimensions. Considering a marginalized maximum likelihood ability estimator, this paper proposes l zt and l zt ∗ , which are extensions of l z and l z ∗ , respectively, for the Rasch testlet model. The computation of l zt ∗ relies on several extensions of the Lord-Wingersky algorithm (1984) that are additional contributions of this paper. Simulation results show that l zt ∗ has close-to-nominal Type I error rates and satisfactory power for detecting aberrant responses. For unidimensional models, l zt and l zt ∗ reduce to l z and l z ∗ , respectively, and therefore allows for the evaluation of person fit with a wider range of IRT models. A real data application is presented to show the utility of the proposed statistics for a test with an underlying structure that consists of both the traditional unidimensional component and the Rasch testlet component.

CFA Models with a General Factor and Multiple Sets of Secondary Factors.

We propose a class of confirmatory factor analysis models that include multiple sets of secondary or specific factors and a general factor. The general factor accounts for the common variance among manifest variables, whereas multiple sets of secondary factors account for the remaining source-specific dependency among subsets of manifest variables. A special case of the model is further proposed which constrains the specific factor loadings to be proportional to the general factor loadings. This proportional model substantially reduces the number of model parameters while preserving the essential structure of the general model. Furthermore, the proportional model allows for the interpretation of latent variables as the expected values of the observed manifest variables, decomposition of the variances, and the inclusion of interactions, similar to generalizability theory. We provide two applications to illustrate the utility of the proposed class of models.

A Variational Maximization-Maximization Algorithm for Generalized Linear Mixed Models with Crossed Random Effects.

We present a variational maximization-maximization algorithm for approximate maximum likelihood estimation of generalized linear mixed models with crossed random effects (e.g., item response models with random items, random raters, or random occasion-specific effects). The method is based on a factorized variational approximation of the latent variable distribution given observed variables, which creates a lower bound of the log marginal likelihood. The lower bound is maximized with respect to the factorized distributions as well as model parameters. With the proposed algorithm, a high-dimensional intractable integration is translated into a two-dimensional integration problem. We incorporate an adaptive Gauss-Hermite quadrature method in conjunction with the variational method in order to increase computational efficiency. Numerical studies show that under the small sample size conditions that are considered the proposed algorithm outperforms the Laplace approximation.

A modular approach for item response theory modeling with the R package flirt.

The new R package flirt is introduced for flexible item response theory (IRT) modeling of psychological, educational, and behavior assessment data. flirt integrates a generalized linear and nonlinear mixed modeling framework with graphical model theory. The graphical model framework allows for efficient maximum likelihood estimation. The key feature of flirt is its modular approach to facilitate convenient and flexible model specifications. Researchers can construct customized IRT models by simply selecting various modeling modules, such as parametric forms, number of dimensions, item and person covariates, person groups, link functions, etc. In this paper, we describe major features of flirt and provide examples to illustrate how flirt works in practice.

On the explaining-away phenomenon in multivariate latent variable models.

Many probabilistic models for psychological and educational measurements contain latent variables. Well-known examples are factor analysis, item response theory, and latent class model families. We discuss what is referred to as the 'explaining-away' phenomenon in the context of such latent variable models. This phenomenon can occur when multiple latent variables are related to the same observed variable, and can elicit seemingly counterintuitive conditional dependencies between latent variables given observed variables. We illustrate the implications of explaining away for a number of well-known latent variable models by using both theoretical and real data examples.

Recent developments in maximum likelihood estimation of MTMM models for categorical data.

Maximum likelihood (ML) estimation of categorical multitrait-multimethod (MTMM) data is challenging because the likelihood involves high-dimensional integrals over the crossed method and trait factors, with no known closed-form solution. The purpose of the study is to introduce three newly developed ML methods that are eligible for estimating MTMM models with categorical responses: Variational maximization-maximization (e.g., Rijmen and Jeon, 2013), alternating imputation posterior (e.g., Cho and Rabe-Hesketh, 2011), and Monte Carlo local likelihood (e.g., Jeon et al., under revision). Each method is briefly described and its applicability for MTMM models with categorical data are discussed.

Effectiveness of case management among older adults with early symptoms of dementia and their primary informal caregivers: a randomized clinical trial.

BACKGROUND: It is believed that timely recognition and diagnosis of dementia is a pre-condition for improving care for both older adults with dementia and their informal caregivers. However, diagnosing dementia often occurs late in the disease. This means that a significant number of patients with early symptoms of dementia and their informal caregivers may lack appropriate care. OBJECTIVES: To compare the effects of case management and usual care among community-dwelling older adults with early symptoms of dementia and their primary informal caregivers. DESIGN: Randomized controlled trial with measurements at baseline and after 6 and 12 months. SETTING: Primary care in West-Friesland, the Netherlands. PARTICIPANTS: 99 pairs of community-dwelling older adults with dementia symptoms (defined as abnormal screening for symptoms of dementia) and their primary informal caregivers. INTERVENTION: 12 months of case management by district nurses for both older adults and informal caregivers versus usual care. PRIMARY OUTCOME: informal caregiver's sense of competence. SECONDARY OUTCOMES: caregiver's quality of life, depressive symptoms, and burden, and patient's quality of life. Process measurements: intervention fidelity and caregiver's satisfaction with the quality of case management. RESULTS: Linear mixed model analyses showed no statistically significant and clinically relevant differences over time between the two groups. The process evaluation revealed that intervention fidelity could have been better. Meanwhile, informal caregivers were satisfied with the quality of case management. CONCLUSION: This study shows no benefits of case management for older adults with dementia symptoms and their primary informal caregivers. One possible explanation is that case management, which has been recommended among diagnosed dementia patients, may not be beneficial if offered too early. However, on the other hand, it is possible that: (1) case management will be effective in this group if more fully implemented and adapted or aimed at informal caregivers who experience more severe distress and problems; (2) case management is beneficial but that it is not seen in the timeframe studied; (3) case management might have undetected small benefits. This has to be established. Trial registration ISCRTN83135728.

A nationwide analysis of incidence and outcome of breast cancer in the country of Surinam, during 1994-2003.

In this study, we describe the incidence, treatment, and outcome of breast cancer (BC) during the period 1994-2003 in the South-American country of Surinam and compare these with those of BC in the Netherlands. Pathology reports and hospital charts from all BC cases diagnosed between 1994 and 2004 were retrieved from Surinam's single pathology laboratory and its five hospitals. Data on demographics, tumor characteristics, treatment, and follow-up were gathered. We compared our data to BC statistics of first generation immigrants from Surinam to the Netherlands. 421 patients were diagnosed with BC during the study period. The age-adjusted incidence rate was 26 per 100,000 compared to 65/100,000 in first generation Surinamese women in the Netherlands. The majority had a fairly advanced stage at presentation, with 60% of tumors larger than 2 cm, and 41.6% with lymph node involvement. Because of the absence of radiotherapy facilities, local treatment in most patients was radical mastectomy. Adjuvant hormonal therapy (51.6%) was administered more frequently than adjuvant chemotherapy (20.3%). A significant number of patients were lost to follow-up, resulting in a median follow-up duration of only 23 months. The 5-year overall survival was 79%. BC incidence in Surinam is low compared to that in the western world, but the advanced stage at diagnosis, the low utilization of systemic adjuvant therapy, and the inadequate follow-up may lead to poor outcomes. A number of steps are underway to improve the level of cancer care in Surinam.

Similarity in depressive symptom profile in a population-based study of migrants in the Netherlands.

OBJECTIVE: Depression is a clinical syndrome developed in Western Europe and North-America. The expression of symptoms and the impact of symptoms on functioning may therefore be expected to vary across cultures and languages. Our first aim was to study differences in depressive symptom profile between indigenous and non-Western immigrant populations in the Netherlands. We hypothesized that differences in expression of depressive symptoms would be more likely in the domains of mood and cognitions, and less likely in the domains of psychomotor and vegetative symptoms. Our second aim was to study ethnic differences in the association of depressive symptoms and general functioning. METHOD: In a random community sample stratified for ethnicity in Amsterdam, the Netherlands, depressive symptoms were assessed by bilingual interviewers using the Composite International Diagnostic Interview (CIDI 2.1) and the Symptom Checklist-90-Revised (SCL-90-R). Impairments in functioning were measured by the World Health Organization Disability Assessment Schedule II (WHODAS II). Results were obtained from 812 subjects: N = 321 native Dutch, N = 213 Turkish-Dutch, N = 191 Moroccan-Dutch, N = 87 Surinamese-Dutch. Differences in depressive symptom expression were tested by differential item functioning. RESULTS: The prevalence of DSM-IV depressive disorder and the overall level of depressive symptoms were higher in the Turkish and Moroccan immigrant groups compared to native Dutch subjects. Ethnic differences in item functioning of depressive symptoms were rare, and equally unlikely in all four symptom domains. Depression was equally associated with functional impairment across ethnic groups. CONCLUSION: Although depressive symptoms were more common among migrants than in the indigenous population, both the depressive symptom profile and the associated functional impairments were comparable. These findings may help diminishing concerns about the validity of using existing diagnostic procedures among ethnic minority groups.

Multivariate Discrete Hidden Markov Models for Domain-Based Measurements and Assessment of Risk Factors in Child Development.

Many studies in the social and behavioral sciences involve multivariate discrete measurements, which are often characterized by the presence of an underlying individual trait, the existence of clusters such as domains of measurements, and the availability of multiple waves of cohort data. Motivated by an application in child development, we propose a class of extended multivariate discrete hidden Markov models for analyzing domain-based measurements of cognition and behavior. A random effects model is used to capture the long-term trait. Additionally, we develop a model selection criterion based on the Bayes factor for the extended hidden Markov model. The National Longitudinal Survey of Youth (NLSY) is used to illustrate the methods. Supplementary technical details and computer codes are available online.

Raising awareness of carrier testing for hereditary haemoglobinopathies in high-risk ethnic groups in the Netherlands: a pilot study among the general public and primary care providers.

BACKGROUND: In the Netherlands no formal recommendations exist concerning preconceptional or antenatal testing for carriership of hereditary haemoglobinopathies. Those at highest risk may be unaware of the possibility of carrier screening. While universal newborn screening has recently been introduced, neither preconceptional nor antenatal carrier testing is routinely offered by health care services to the general public. A municipal health service and a foundation for public information on medical genetics undertook a pilot project with the aim of increasing knowledge and encouraging informed choice. Two groups were targeted: members of the public from ethnic groups at increased risk, and primary health care providers. This study examines the effectiveness of culturally specific 'infotainment' to inform high-risk ethnic groups about their increased risk for haemoglobinopathies. In addition, the study explores attitudes and intentions of primary care providers towards haemoglobinopathy carrier testing of their patients from high-risk ethnic groups. METHODS: Informational sessions tailored to the public or professionals were organised in Amsterdam, and evaluated for their effect. Psychological parameters were measured using structured questionnaires based on the Theory of Planned Behaviour. RESULTS: The pre-test/post-test questionnaire showed that members of the public gained understanding of inheritance and carriership of haemoglobinopathies from the "infotainment" session (p < 0.01). Perceived behavioural control, i.e. the feeling that they could actually get tested if they wanted to, increased in the targeted age group of 18-45 years (N = 41; p < 0.05). 191 surveys were collected from general practitioners or midwives. Their attitude towards the education programme for high-risk ethnic groups was positive, yet they did not show strong intention to effectuate carrier testing of their patients on the basis of ethnicity. The main factor which explained their (lack of) intention was social norm, i.e. their perception of negative peer opinion (41% variance explained). The majority of primary health care providers felt that policy change was unnecessary. CONCLUSION: The "infotainment" programme may have a positive effect on people from high-risk groups, but informed general practitioners and midwives were reluctant to facilitate their patients' getting tested. Additional initiatives are needed to motivate primary care providers to facilitate haemoglobinopathy carrier testing for their patients from high-risk backgrounds.

Relapse-remission and remission-relapse switches in rheumatoid arthritis patients were modeled by random effects.

OBJECTIVE: To compare statistical models for the analysis of two-state disease processes. STUDY DESIGN AND SETTING: A two-armed randomized trial of patients with early rheumatoid arthritis (RA) treated by either combination therapy (sulfasazaline, methotrextate, prednisolone) or monotherapy (sulfasazaline). Disease activity (remission or relapse) was analyzed with the logistic regression model, the proportional hazards regression model, and the continuous-time Markov process model for panel data. The dependence among the switching times was studied by (1) including correlated normal random patient effects for the relapse-remission and remission-relapse switching probabilities; (2) assuming the population to be a mixture of patients responsive and nonresponsive to therapy; (3) including separate parameters for the first and subsequent relapse-remission switch; and (4) combining (1) and (3). The four approaches were compared using parametric bootstrap checks. RESULTS: The logistic regression model, the proportional hazards regression model, and the continuous-time Markov process model for panel data yielded similar combination therapy effects. The inclusion of random patient effects (approaches 1 and 4) gave the best fit to the observed disease activity pattern. CONCLUSION: Models with correlated random effects can provide a satisfactory fit to two-state disease patterns.

Longitudinal tobit regression: a new approach to analyze outcome variables with floor or ceiling effects.

BACKGROUND: In many epidemiologic longitudinal studies, the outcome variable has floor or ceiling effects. Although it is not correct, these variables are often treated as normally distributed continuous variables. OBJECTIVES: In this article, the performance of a relatively new statistical technique, longitudinal tobit analysis, is compared with a classical longitudinal data analysis technique (i.e., linear mixed models). STUDY DESIGN AND SETTING: The analyses are performed on an example data set from rehabilitation research in which the outcome variable of interest (the Barthel index measured at on average 16.3 times) has typical floor and ceiling effects. For both the longitudinal tobit analysis and the linear mixed models an analysis with both a random intercept and a random slope were performed. RESULTS: Based on model fit parameters, plots of the residuals and the mean of the squared residuals, the longitudinal tobit analysis with both a random intercept and a random slope performed best. In the tobit models, the estimation of the development over time revealed a steeper development compared with the linear mixed models. CONCLUSION: Although there are some computational difficulties, longitudinal tobit analysis provides a very nice solution for the longitudinal analysis of outcome variables with floor or ceiling effects.

Anxiety symptoms and cognitive performance in later life: results from the longitudinal aging study Amsterdam.

OBJECTIVES: This study investigates whether, and if so how, anxiety symptoms are related to cognitive decline in elderly persons and whether anxiety symptoms precede cognitive decline. METHOD: Data were obtained from the Longitudinal Aging Study Amsterdam. Anxiety symptoms were measured with the Hospital Anxiety and Depression Scale. General cognitive functioning was measured with the Mini-Mental State Examination, episodic memory with the Auditory Verbal Learning Test, fluid intelligence with the Raven's Coloured Progressive Matrices and information processing speed with the coding task. Multilevel analyses were performed to investigate the relationship between anxiety symptoms and cognitive decline over 9 years, taking into account confounding variables. RESULTS: Although not consistent across all dimensions of cognitive functioning, a curvilinear effect of anxiety on cognitive performance was found. Furthermore, we found that previous measurement of anxiety symptoms were not predictive of cognitive decline at a later time-point. CONCLUSION: This study suggests that the effect of anxiety on cognition depends on the severity of the present anxiety symptoms with mild anxiety associated with better cognition, whereas more severe anxiety is associated with worse cognition. The effect of anxiety symptoms on cognitive functioning seems to be a temporary effect, anxiety is not predictive of cognitive decline.

A modified procedure for parallel analysis of ordered categorical data.

Parallel analysis has been well documented to be an effective and accurate method for determining the number of factors to retain in exploratory factor analysis. The O'Connor (2000) procedure for parallel analysis has many benefits and is widely applied, yet it has a few shortcomings in dealing with missing data and ordinal variables. To address these technical issues, we adapted and modified the O'Connor procedure to provide an alternative method that better approximates the ordinal data by factoring in the frequency distributions of the variables (e.g., the number of response categories and the frequency of each response category per variable). The theoretical and practical differences between the modified procedure and the O'Connor procedure are discussed. The SAS syntax for implementing this modified procedure is also provided.

Toward disentangling sources of individual differences in appraisal and anger.

A theoretical framework is presented to explain individual differences in situation-specific emotional experience in terms of three different sources of variance: (a) individual differences in how one appraises one's circumstances, (b) individual differences in how appraisals are related to the experience of emotion, and (c) individual differences independent from situation and appraisal. The relative contribution and nature of these sources was examined empirically for the experience of anger based on data from two directed imagery studies (total N=1,192). Consistent results across the two studies demonstrated that variability in anger experience primarily stems from variability in how a situation is appraised and to a smaller extent from individual differences in the relations between the appraisals and anger and individual differences independent of appraisal. The findings further identified frustration as the central appraisal involved in anger. Implications for emotion theories and anger management programs are discussed.

Depression in old age (75+), the PIKO study.

BACKGROUND: Old people (75+) are underrepresented in studies on the prevalence of and risk factors for depression while the number of elderly people suffering from this mood disorder may be considerably higher than previously assumed. The role--if any--of age and gender in 'Geriatric Depression' is still unclear. METHODS: In this community-based study, prevalence of depressive symptomatology and risk indicators were assessed in 2850 participants aged 75 years or more. A clinically relevant level of depressive symptoms was defined as a score of > or =16 on the Centre for Epidemiologic Studies Depression scale (CES-D). Demographic data and questions related to physical and psychological health were recorded. Simple and multiple logistic regression techniques were used to determine the risk indicators (Odds Ratios, OR, with 95% confidence intervals, CI) with apparent importance to this population. RESULTS: The prevalence of depressive symptoms was assessed to be 31.1%. This is considerably higher than what has been found in younger elderly samples. The bivariate age effect was OR 1.05 (95% CI=1.03 to 1.07). Controlling for confounding, the effect of gender and age on depressive symptoms disappeared. CONCLUSIONS: Depressive symptoms are highly prevalent in the elderly population and increase with age. This increase seems to be attributable to age-related changes in risk factors rather than to ageing itself. With regard to the risk factors found, attention should perhaps be paid to functional disability, loneliness and apprehensiveness for falling since these risk indicators are amenable for improvement.

Latent Class Models for Diary Method Data: Parameter Estimation by Local Computations.

The increasing use of diary methods calls for the development of appropriate statistical methods. For the resulting panel data, latent Markov models can be used to model both individual differences and temporal dynamics. The computational burden associated with these models can be overcome by exploiting the conditional independence relations implied by the model. This is done by associating a probabilistic model with a directed acyclic graph, and applying transformations to the graph. The structure of the transformed graph provides a factorization of the joint probability function of the manifest and latent variables, which is the basis of a modified and more efficient E-step of the EM algorithm. The usefulness of the approach is illustrated by estimating a latent Markov model involving a large number of measurement occasions and, subsequently, a hierarchical extension of the latent Markov model that allows for transitions at different levels. Furthermore, logistic regression techniques are used to incorporate restrictions on the conditional probabilities and to account for the effect of covariates. Throughout, models are illustrated with an experience sampling methodology study on the course of emotions among anorectic patients.

Qualitative longitudinal analysis of symptoms in patients with primary and metastatic brain tumours.

Primary and metastatic brain tumour patients are treated with surgery, radiation therapy and chemotherapy. Such treatments often result in short- and long-term symptoms that impact cognitive, emotional and physical function. Therefore, understanding the transition of symptom burden over time is important for guiding treatment and follow-up of brain tumour patients with symptom-specific interventions. We describe the use of a hidden Markov model with person-specific random effects for the temporal pattern of symptom burden. Clinically relevant covariates are also incorporated in the analysis through the use of generalized linear models.

Primary care patients with mental health problems: outcome of a randomised clinical trial.

BACKGROUND: The prevalence of patients with mental health problems in general practice is high, and at least one-third of these problems last for 6 months or longer. Patients with these problems take up more time during a consultation and attend more frequently. AIM: This study investigated the effectiveness of problem-solving treatment for primary care patients with mental health problems. The hypothesis was that patients receiving problem-solving treatment from a nurse would have fewer symptoms after 3 months, or a lower attendance rate, compared with patients receiving the usual care from the GP. DESIGN OF THE STUDY: Randomised clinical trial. SETTING: Twelve general practices in Amsterdam and 12 nurses from a mental healthcare institution. METHOD: A sample of patients aged >or=18 years were screened for mental health problems with the general health questionnaire (GHQ-12) in the waiting room of the general practices, and were randomised. Patients receiving the problem-solving treatment were required to complete four to six treatment sessions, while patients in the control group were treated as usual by the GP. RESULTS: No significant difference was found between the groups in terms of improved psychopathology or a decrease in attendance rate. Post-hoc analyses showed a sub-group of patients with more severe pathology who may benefit from problem-solving treatment. CONCLUSION: The main results show that problem-solving treatment provided by a nurse adds little to the usual care from the GP for frequent attenders with mental health problems. Post-hoc analyses show that there may be a sub-group of more severely depressed patients who could benefit from problem-solving treatment.