ijsmr
Abstract : Survival Analysis of Duration of Breastfeeding and Associated Factors of Early Cessation of Breastfeeding in Ethiopia (12)
|
|
Abstract: The purpose of this study was to assess the duration of breastfeeding among women of reproductive age in Ethiopia and to identify determinants associated with early cessation of breastfeeding. Data for the study were drawn from the Ethiopia Demographic and Health Survey 2005. The study included mothers of 9,066 children from nine regional states and two city administrations. The Kaplan-Meier and stratified Cox’s hazard model were employed for the analysis of breastfeeding-related data. The Kaplan-Meier survival estimate showed that the probability of mothers who continue to breastfeeding was high (97.3%) for the first month. The breastfeeding rates then declined to 92.5% at 6 months, 78.4% at 12 months, 37% at 24 months and 8.3% at 48 months. The mean and median duration of breastfeeding in Ethiopia were 25.64 and 24.00 months respectively. The stratified Cox regression analysis revealed that younger mothers, mothers who had lived in urban area, mothers having higher education, higher maternal parity, early pregnant and being a Muslim and protestant were significant determinants of early cessation of breastfeeding in Ethiopia. Then, we recommend that the breastfeeding-promotion programs in Ethiopia should give special attention to young mothers, those who lived in urban areas, mothers with higher education, those who have higher parity, those who have early pregnancy and who are Muslims and Protestants since these mothers tend to breastfeed their child for a relatively shorter period of time. Keywords: Breastfeeding duration, Kaplan-Meier estimator, Determinants, Stratified- Cox regression model, Hazard-Ratio, Ethiopia.Download Full Article |
Abstract: A Comparison of Parametric and Semi-Parametric Models for Microarray Data Analysis
|
|
Abstract: Microarray technology has revolutionized genomic studies by enabling the study of differential expression of thousands of genes simultaneously. Parametric, nonparametric and semi-parametric statistical methods have been proposed for gene selection within the last sixteen years. In an effort to find the “gold standard", the performance of some common parametric and nonparametric methods have been compared in terms of power to select differentially expressed genes and other desirable properties. However, no such comparisons have been conducted between parametric and semi-parametric models. In this study, we compared a semi-parametric model based on copulas with a parametric model (the quantitative trait analysis or QTA model) in terms of power and the ability to control the Type I error rate. In addition, we proposed a simple algorithm for choosing an optimal copula. The two approaches were applied to a publicly available melanoma cell lines dataset for validation. Both methods performed well in terms of power but the copula approach was notably the better. In terms of the Type I error rate control, the two methods were comparable. More methods for selecting an optimal copula for gene expression data need to be developed, as the proposed procedure is limited to copulas that permit both negative and positive dependence only. Keywords: Copula, Goodness-of-fit, Melanoma, Microarray, Power, Type I error. |
Abstract: The MAX Statistic is Less Powerful for Genome Wide Association Studies Under Most Alternative Hypotheses
|
|
Abstract: Genotypic association studies are prone to inflated type I error rates if multiple hypothesis testing is performed, e.g., sequentially testing for recessive, multiplicative, and dominant risk. Alternatives to multiple hypothesis testing include the model independent genotypic c2 test, the efficiency robust MAX statistic, which corrects for multiple comparisons but with some loss of power, or a single Armitage test for multiplicative trend, which has optimal power when the multiplicative model holds but with some loss of power when dominant or recessive models underlie the genetic association. We used Monte Carlo simulations to describe the relative performance of these three approaches under a range of scenarios. All three approaches maintained their nominal type I error rates. The genotypic c2 and MAX statistics were more powerful when testing a strictly recessive genetic effect or when testing a dominant effect when the allele frequency was high. The Armitage test for multiplicative trend was most powerful for the broad range of scenarios where heterozygote risk is intermediate between recessive and dominant risk. Moreover, all tests had limited power to detect recessive genetic risk unless the sample size was large, and conversely all tests were relatively well powered to detect dominant risk. Taken together, these results suggest the general utility of the multiplicative trend test when the underlying genetic model is unknown. Keywords: Armitage test, case-control study, efficiency robust statistics, MAX statistic, multiple comparisons;, Type I error. |
Abstract: Sample Size Calculation in Clinical Studies: Some Common Scenarios
|
|
Abstract: Determining the optimal sample size is crucial for any scientific investigation. An optimal sample size provides adequate power to detect statistical significant difference between the comparison groups in a study and allows the researcher to control for the risk of reporting a false-negative finding (Type II error). A study with too large a sample is harder to conduct, expensive, time consuming and may expose an unnecessarily large number of subjects to potentially harmful or futile interventions. On the other hand, if the sample size is too small, a best conducted study may fail to answer a research question due to lack of sufficient power. To draw a valid and accurate conclusion, an appropriate sample size must be determined prior to start of any study. This paper covers the essentials in calculating sample size for some common study designs. Formulae along with some worked examples were demonstrated for potential applied health researchers. Although maximum power is desirable, this is not always possible given the resources available for a study. Researchers often needs to choose a sample size that makes a balance between what is desirable and what is feasible. Keywords: Sample Size Calculation, Power, Hypothesis Test, Level of Significance, Mean, Proportion. |
