Sam Woolford

A Multiple-Threshold AR(1) Model

We consider the model Z, = +(0, k ) + +(I, k)Z,_, + a,(k) whenever r,_, < Z,_,S r,., 1S k k 1, with r, = -m and r, = m. Here {+(i, k); i =0 , l ; 1 5 k 5 1) is a sequence of real constants, not necessarily equal, and, for 1 5 k 5 I, {a,(k), t 2 1) is a sequence of i.i.d. random variables with mean 0 and with {a,(k), t 2 1) independent of {a,(j), t 2 1) for j # k. Necessary and sufficient conditions on the constants {+(i, k)} are given for the stationarity of the process. Least squares estimators of the model parameters are derived and, under mild regularity conditions, are shown to be strongly consistent and asymptotically normal. NON-LINEAR TIME SERIES; SETAR MODELS; AUTOREGRESSIVE MODELS; MARKOV CHAINS

Review of Three Latent Class Cluster Analysis Packages: Latent Gold, poLCA, and MCLUST

This article reviews three software packages that can be used to perform latent class cluster analysis, namely, Latent Gold®, MCLUST, and poLCA. Latent Gold® is a product of Statistical Innovations whereas MCLUST and poLCA are packages written in R and are available through the web site http://www.r-project.org. We use a single dataset and apply each software package to develop a latent class cluster analysis for the data. This allows us to compare the features and the resulting clusters from each software package. Each software package has its strengths and weaknesses and we compare the software from the perspectives of usability, cost, data characteristics, and performance. Whereas each software package utilizes the same methodology, we show that each results in a different cluster solution and suggest some rationales for deciding which package to use.

Identifying Groups: A Comparison of Methodologies

This paper describes and compares three clustering techniques: traditional clustering methods, Kohonen maps and latent class models. The paper also proposes some novel measures of the quality of a clustering. To the best of our knowledge, this is the first contribution in the literature to compare these three techniques in a context where the classes are not known in advance.

Physicians’ knowledge, beliefs, and use of race and human genetic variation: new measures and insights

BackgroundUnderstanding physician perspectives on the intersection of race and genomics in clinical decision making is critical as personalized medicine and genomics become more integrated in health care services. There is a paucity of literature in the United States of America (USA) and globally regarding how health care providers understand and use information about race, ethnicity and genetic variation in their clinical decision making. This paper describes the development of three scales related to addressing this gap in the literature: the Bonham and Sellers Genetic Variation Knowledge Assessment Index--GKAI, Health Professionals Beliefs about Race—HPBR, and Racial Attributes in Clinical Evaluation—RACE scales.MethodsA cross-sectional, web survey of a national random sample of general internists in the USA (N = 787) was conducted. Confirmatory factor analysis was used to assess the construct validity of the scales. Scale items were developed through focus groups, cognitive interviews, expert advisory panels, and exploratory factor analysis of pilot data.ResultsGKAI was measured as a count of correct answers (Mean = 3.28 SD = 1.17). HPBR yielded two domains: beliefs about race as a biological phenomenon (HPBR-BD, alpha = .69, 4 items) and beliefs about the clinical value of race and genetic variation for understanding risk for disease (HPBR-CD alpha = .61, 3 items). RACE yielded one factor (alpha = .86, 7 items).ConclusionsGKAI is a timely knowledge scale that can be used to assess health professional knowledge of race and human genetic variation. HPBR is a promising new tool for assessing health professionals’ beliefs about the role of race and its relationship with human genetic variation in clinical practice. RACE offers a valid and reliable tool for assessing explicit use of racial attributes in clinical decision making.

Measuring the international digital divide: an application of Kohonen self-organising maps

With the help of a Kohonen self-organising algorithm, this paper presents a mapping and analysis of the global digital divide along with its main drivers. Several broad groups and subgroups are identified, consisting of countries that are similar in their digital development and in a number of other attributes. We find that the digital divide seems to occur synchronously with divisions in income, social, demographic and infrastructure measures. By examining a large dataset of 160 countries over a short period of three years, we find evidence of both convergence and divergence among the countries over time. We expect these findings to inform the ongoing debate on drivers of the International Digital Divide (IDD). In addition, this paper provides a novel visualisation of the digital divide and its predictors on a two-dimensional grid. Extensions of this work, with the availability of more years of data, could investigate the potential convergence of countries to particular patterns of digital development.

Modelling decisions to undergo genetic testing for susceptibility to common health conditions: An ancillary study of the Multiplex Initiative

New genetic tests reveal risks for multiple conditions simultaneously, although little is understood about the psychological factors that affect testing uptake. We assessed a conceptual model called the multiplex genetic testing model (MGTM) using structural equation modelling. The MGTM delineates worry, perceived severity, perceived risk, response efficacy and attitudes towards testing as predictors of intentions and behaviour. Participants were 270 healthy insured adults aged 25–40 from the Multiplex Initiative conducted within a health care system in Detroit, MI, USA. Participants were offered a genetic test that assessed risk for eight common health conditions. Confirmatory factor analysis revealed that worry, perceived risk and severity clustered into two disease domains: cancer or metabolic conditions. Only perceived severity of metabolic conditions was correlated with general response efficacy (β = 0.13, p<0.05), which predicted general attitudes towards testing (β = 0.24, p<0.01). Consistent with our hypothesised model, attitudes towards testing were the strongest predictors of intentions to undergo testing (β = 0.49, p<0.01), which in turn predicted testing uptake (OR 17.7, β = 0.97, p<0.01). The MGTM explained a striking 48% of the variance in intentions and 94% of the variation in uptake. These findings support use of the MGTM to explain psychological predictors of testing for multiple health conditions.

A quality assessment tool for improvement planning

The identification and prioritisation of improvement opportunities can be a challenge for many organisations due to a variety of resource limitations. This can result in improvement efforts that do not have a clearly defined linkage to the return that they will generate or to the achievement of key organisation strategies intended to maintain or advance the organisations competitive position. The current research proposes the development of a decision support tool, based on the analytic hierarchy process and a self-assessment tool for evaluating current performance, which can effectively and efficiently identify and prioritise improvement opportunities for annual improvement planning. The decision support tool can be easily automated to facilitate implementation.

PUGS: A novel scale to assess perceptions of uncertainties in genome sequencing.

Expectations of results from genome sequencing by end users are influenced by perceptions of uncertainty. This study aimed to assess uncertainties about sequencing by developing, evaluating, and implementing a novel scale. The Perceptions of Uncertainties in Genome Sequencing (PUGS) scale comprised ten items to assess uncertainties within three domains: clinical, affective, and evaluative. Participants (n=535) from the ClinSeq® NIH sequencing study completed a baseline survey that included the PUGS; responses (mean = 3.4/5, SD=0.58) suggested modest perceptions of certainty. A confirmatory factor analysis identified factor loadings that led to elimination of two items. A revised eight‐item PUGS scale was used to test correlations with perceived ambiguity (r = −0.303, p < 0.001), attitudinal ambivalence (r = −0.111, p = 0.011), and ambiguity aversion (r = −0.093, p = 0.033). Results support nomological validity. A correlation with the MICRA uncertainty subscale was found among 175 cohort participants who had received results (r = −0.335, p < 0.001). Convergent and discriminant validity were also satisfied in a second sample of 208 parents from the HudsonAlpha CSER Project who completed the PUGS (mean = 3.4/5, SD = 0.72), and configural invariance was supported across the two datasets. As such, the PUGS is a promising scale for evaluating perceived uncertainties in genome sequencing, which can inform interventions to help patients form realistic expectations of these uncertainties.

On a class of threshold AR(k) processes

AbstractWe consider the model