Dataset Information

The Performance of Multivariate Methods for Two-Group Comparisons with Small Samples and Incomplete Data.

ABSTRACT: In intervention studies having multiple outcomes, researchers often use a series of univariate tests (e.g., ANOVAs) to assess group mean differences. Previous research found that this approach properly controls Type I error and generally provides greater power compared to MANOVA, especially under realistic effect size and correlation combinations. However, when group differences are assessed for a specific outcome, these procedures are strictly univariate and do not consider the outcome correlations, which may be problematic with missing outcome data. Linear mixed or multivariate multilevel models (MVMMs), implemented with maximum likelihood estimation, present an alternative analysis option where outcome correlations are taken into account when specific group mean differences are estimated. In this study, we use simulation methods to compare the performance of separate independent samples t tests estimated with ordinary least squares and analogous t tests from MVMMs to assess two-group mean differences with multiple outcomes under small sample and missingness conditions. Study results indicated that a MVMM implemented with restricted maximum likelihood estimation combined with the Kenward-Roger correction had the best performance. Therefore, for intervention studies with small N and normally distributed multivariate outcomes, the Kenward-Roger procedure is recommended over traditional methods and conventional MVMM analyses, particularly with incomplete data.

SUBMITTER: Pituch KA

PROVIDER: S-EPMC7093229 | biostudies-literature | 2019 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

The Performance of Multivariate Methods for Two-Group Comparisons with Small Samples and Incomplete Data.

Pituch Keenan A KA Joshi Megha M Cain Molly E ME Whittaker Tiffany A TA Chang Wanchen W Park Ryoungsun R McDougall Graham J GJ

Multivariate behavioral research 20190925 5

In intervention studies having multiple outcomes, researchers often use a series of univariate tests (e.g., ANOVAs) to assess group mean differences. Previous research found that this approach properly controls Type I error and generally provides greater power compared to MANOVA, especially under realistic effect size and correlation combinations. However, when group differences are assessed for a specific outcome, these procedures are strictly univariate and do not consider the outcome correlat ...[more]

PMID: 31552754

Similar Datasets

Project description:Next-generation sequencing has emerged as an essential technology for the quantitative analysis of gene expression. In medical research, RNA sequencing (RNA-seq) data are commonly used to identify which type of disease a patient has. Because of the discrete nature of RNA-seq data, the existing statistical methods that have been developed for microarray data cannot be directly applied to RNA-seq data. Existing statistical methods usually model RNA-seq data by a discrete distribution, such as the Poisson, the negative binomial, or the mixture distribution with a point mass at zero and a Poisson distribution to further allow for data with an excess of zeros. Consequently, analytic tools corresponding to the above three discrete distributions have been developed: Poisson linear discriminant analysis (PLDA), negative binomial linear discriminant analysis (NBLDA), and zero-inflated Poisson logistic discriminant analysis (ZIPLDA). However, it is unclear what the real distributions would be for these classifications when applied to a new and real dataset. Considering that count datasets are frequently characterized by excess zeros and overdispersion, this paper extends the existing distribution to a mixture distribution with a point mass at zero and a negative binomial distribution and proposes a zero-inflated negative binomial logistic discriminant analysis (ZINBLDA) for classification. More importantly, we compare the above four classification methods from the perspective of model parameters, as an understanding of parameters is necessary for selecting the optimal method for RNA-seq data. Furthermore, we determine that the above four methods could transform into each other in some cases. Using simulation studies, we compare and evaluate the performance of these classification methods in a wide range of settings, and we also present a decision tree model created to help us select the optimal classifier for a new RNA-seq dataset. The results of the two real datasets coincide with the theory and simulation analysis results. The methods used in this work are implemented in the open-scource R scripts, with a source code freely available at https://github.com/FocusPaka/ZINBLDA.

Project description:The weaning of dairy calves is a significant stressor which can impact on calf performance and welfare. However, many traditional methods of assessing the effects of stressors can be invasive and impractical for farmers. This study aimed to use a combination of non-invasive monitoring technologies alongside traditional measures of calf performance to examine the impact of two contrasting weaning methods commonly used on dairy farms in the United Kingdom. Ninety group-housed Holstein Friesian calves were allocated to one of two weaning methods: (i) gradual weaning (GW) with volume of milk replacer (MR) stepping down from 36 days of age and complete withdrawal of MR at 57 days of age and (ii) abrupt weaning (AW) with consistent daily volume of milk replacer and complete withdrawal of MR at 50 days of age. Feeding regimes were such that calves from both treatments were offered the same total amount of milk powder. Gradually weaned calves displayed increased solid feed intake at an earlier age when compared with AW calves. Feed conversion efficiency (FCE) was reduced in gradually weaned calves between days 36 and 49. However, there was no difference in live weight (LWT) or average daily gain (ADG) throughout this period. Abrupt weaning at 50 days of age resulted in decreased ADG and FCE between days 50 and 56. However, there were no treatment differences in ADG between days 57 and 62. Live weight tended to be increased by 2.2 kg in GW calves when compared with AW calves at the end of experiment on day 63. Frequency of unrewarded visits to the milk feeder throughout the pre-wean period was consistently increased in GW calves. Daily lying time was reduced in AW compared with GW calves in the days following abrupt weaning (days 50-55). However, these differences did not persist between days 57 and 62. Heart rate variability (HRV) tended to be decreased in GW compared with AW calves in the period following complete withdrawal of milk replacer. Findings from the current study suggest that calves offered the same total amount of milk powder can be weaned either gradually from 36 days of age or abruptly at 50 days of age without significant impact to live weight at 63 days of age. However, both behavioural and physiological data collected using the methods described could suggest that gradual weaning of calves from 36 days of age results in an increase in underlying frustration. This study highlights the potential of using a combination of non-invasive monitoring technologies in assessing calf response to common management practices.

Dataset Information

The Performance of Multivariate Methods for Two-Group Comparisons with Small Samples and Incomplete Data.

Publications

The Performance of Multivariate Methods for Two-Group Comparisons with Small Samples and Incomplete Data.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets