On Testing Departure from the Binomial and Multinomial Assumptions

Paul, S. R.; Liang, Kung Yee; Self, Steven G.

doi:10.2307/2532048

Cited by 49 publications

(43 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Specifically, we tested whether the overall frequencies of hemorrhages within the four lobes were significantly different from the relative volumes of the four lobes. 15 A test for goodness of fit of the multinomial distribution against Dirichlet-multinomial alternatives 16 was used to determine whether the locations of hemorrhages within individuals were independent of one another (ie, no clustering of locations).…”

Section: Methodsmentioning

confidence: 99%

Spatial clustering of hemorrhages in probable cerebral amyloid angiopathy

et al. 2005

View full text Add to dashboard Cite

Cerebral amyloid angiopathy (CAA) is a common cause of symptomatic intracerebral hemorrhage (ICH), as well as small asymptomatic hemorrhage in the elderly. We used gradient-echo MRI to analyze spatial distribution of 321 hemorrhages in 59 patients with probable CAA-related ICH. Hemorrhagic lesions were found preferentially in the temporal (ratio of actual to expected hemorrhages = 1.37) and occipital lobes (ratio = 1.45, p < 0.0001). Within individuals, hemorrhages tended to cluster, regardless of lobe (p < 0.0001). Among subjects followed prospectively for recurrence, clustering of new symptomatic and asymptomatic hemorrhages was observed. These data suggest that regional differences within the brain play a role in the development of CAA-related hemorrhage.

show abstract

Section: Methodsmentioning

confidence: 99%

Spatial clustering of hemorrhages in probable cerebral amyloid angiopathy

et al. 2005

View full text Add to dashboard Cite

show abstract

“…In similar hypothesis testing situations (see, for example, Barnwal and Paul 1988;Paul et al, 1989;Breslow, 1991;Paul and Banerjee, 1998) a score test (Rao, 1947) or a more general C(α) test (Neyman, 1959) perform well in terms of level and power. Also, they require only estimates of the parameters under the null hypothesis and often produce a statistic that is simple to use.…”

Section: Shortcomings Of the Traditional Large Sample Testsmentioning

confidence: 99%

“…However, in similar contexts it was found earlier that both the Wald test and the likelihood ratio test suffer from the disadvantage that they often do not maintain level. See, for example, Paul et al (1989), Thall (1992), and Paul and Barnejee (1998). Here, we show such a disadvantage of the likelihood ratio statistic for testing H 0 : ρ = 0 against H A : ρ > 0 in a single clump finite mixture model.…”

Section: Shortcomings Of the Traditional Large Sample Testsmentioning

confidence: 99%

“…When the parameter value under H 0 is on the boundary of the parameter space they show that the correct distribution of the LR test is a 50:50 mixture of a point-mass at zero and χ 2 with one degree of freedom. See Paul et al (1989) for an application. Thus, the distribution of the LR statistic, under the hypothesis of a model with no clump against a model with one clump is a 50:50 mixture of a point-mass at zero and χ 2 with one degree of freedom.…”

Section: Shortcomings Of the Traditional Large Sample Testsmentioning

confidence: 99%

See 1 more Smart Citation

The multi‐clump finite mixture distribution and model selection

Paul¹,

Banerjee²,

Balasoorya³

2009

Environmetrics

View full text Add to dashboard Cite

SUMMARYIn practical data analysis, often an important problem is to determine the number of clumps in discrete data in the form of proportions. This can be done through model selection in a multi-clump finite mixture model. In this paper, we propose bootstrap likelihood ratio tests to test the fit of a multinomial model against the single clump finite mixture distribution and to determine the number of clumps in the data, that is, to select a model with appropriate number of clumps. Shortcomings of some traditional large sample procedures are also shown. Three datasets are analyzed.

show abstract

“…In addition, locally most powerful score tests can be used for testing against the null hypothesis. 16 The above discussion concerns results as the tracking/correlation decreases. Conversely, as the a, approach 0,8 approaches co , and p approaches 1.…”

Section: Introductionmentioning

confidence: 99%

Parametric Models to Assess Tracking of Categorical High-Risk Sexual Behaviour for Hiv Infection

1996

View full text Add to dashboard Cite

Prevalence and tracking (within-individual correlation) of high-risk behaviour measured in longitudinal studies provide a complementary description of how behaviour is maintained over time. Without tracking, when population prevalences are approximately constant over time, the steadiness of the aggregate pattern is likely to mask the dynamic changes of the response occurring on an individual level. In this paper, a parametric model is proposed for the estimation of these features when the observed response is a time-stationary categorical measurement. The model extends the standard Dirichlet-multinomial model in the spirit of Prentice (1986, JASA) by allowing both negative and positive correlation among repeated categorical measurements. In addition, both the marginal category probabilities and the tracking can be modelled as a function of individual-level covariates. Details regarding likelihood based estimation and inference are provided and illustrated with data from the Multicenter AIDS Cohort Study (MACS).

show abstract

On Testing Departure from the Binomial and Multinomial Assumptions

Cited by 49 publications

References 9 publications

Spatial clustering of hemorrhages in probable cerebral amyloid angiopathy

Spatial clustering of hemorrhages in probable cerebral amyloid angiopathy

The multi‐clump finite mixture distribution and model selection

Parametric Models to Assess Tracking of Categorical High-Risk Sexual Behaviour for Hiv Infection

Contact Info

Product

Resources

About