Generalized lambda distribution for flexibly testing differences beyond the mean in the distribution of a dependent variable such as body mass index.

Generalized lambda distribution for flexibly testing differences beyond the mean in the distribution of a dependent variable such as body mass index.
复制标题

DOI:
10.1038/ijo.2017.262
复制
发表时间:
2018-04
期刊:
International journal of obesity (2005)
影响因子:
--
通讯作者:
Allison DB
Allison DB
中科院分区:
其他
文献类型:
--
作者:
Ejima K;Pavela G;Li P;Allison DB

文献摘要

参考文献

相似文献

传统的统计方法通常在特定的分布假设下检验分布中单个参数的组间差异,通常是条件均值(例如,按教育类别计算的平均体重指数[BMI; kg/m2]的差异)。然而,平均值以外的参数可能是有意义的,并且在某些情况下可能违反传统统计方法的分布假设。我们描述了广义lambda分布(GLD)的一个应用,GLD是一个灵活的分布,可以用来模拟连续的结果;并同时描述一个似然比检验[LRT],用于多个分布参数的差异,包括集中趋势、分散、不对称和陡峭度的测量。我们通过使用健康与退休研究(HRS)数据集测试不同教育类别的BMI分布的多个参数的差异,证明了我们方法的价值。我们提出的方法表明,在完整数据集(N=13571) (P<0.001)和随机重新抽样数据集(每个类别N=300)中,BMI分布的至少一个参数因教育类别而异,以评估在较小功率情况下的方法(P=0.044)。采用正态分布替代GLD的类似方法表明,在完整数据集之间存在显著差异(P<0.001),但在较小的随机重采样数据集中没有显著差异(P=0.968)。此外,所提出的方法使我们能够分别指定在完整子样本和随机子样本中,哪些教育类别的BMI分布参数存在显著差异。我们的方法提供了一种灵活的统计方法来比较感兴趣的变量的整个分布,这可以作为传统方法的补充,这些方法经常需要不满足的假设,并且只关注单个分布参数。
Conventional statistical methods often test for group differences in a single parameter of a distribution, usually the conditional mean (e.g., differences in mean body mass index [BMI; kg/m2] by education category) under specific distributional assumptions. However, parameters other than the mean may of be interest, and the distributional assumptions of conventional statistical methods may be violated in some situations. We describe an application of the generalized lambda distribution (GLD), a flexible distribution that can be used to model continuous outcomes; and simultaneously describe a likelihood ratio test [LRT] for differences in multiple distribution parameters, including measures of central tendency, dispersion, asymmetry, and steepness. We demonstrate the value of our approach by testing for differences in multiple parameters of the BMI distribution by education category using the Health and Retirement Study (HRS) dataset. Our proposed method indicated that at least one parameter of the BMI distribution differed by education category in both the complete dataset (N=13571) (P<0.001) and a randomly resampled dataset (N=300 from each category) to assess the method under circumstances of lesser power (P=0.044). Similar method using normal distribution alternative to GLD indicated the significant difference among the complete dataset (P<0.001) but not in the smaller randomly resampled dataset (P=0.968). Moreover, the proposed method allowed us to specify which parameters of the BMI distribution significantly differed by education category for both the complete and the random subsample, respectively. Our method provides a flexible statistical approach to compare the entire distribution of variables of interest, which can be a supplement to conventional approaches that frequently require unmet assumptions and focus only on a single parameter of distribution.
DOI: 10.1186/s12963-016-0076-2
发表时间: 2016
影响因子: 3.3
作者:
Ng M;Liu P;Thomson B;Murray CJ
通讯作者: Murray CJ
影响因子: --
作者:
Neyman, J
通讯作者: Neyman, J
DOI: 10.1016/j.ehb.2011.07.004
发表时间: 2011-12-01
影响因子: 2.5
作者:
Jolliffe, Dean
通讯作者: Jolliffe, Dean
史诗般的帕纳奇研究中,教育与体重指数和腰围的关联。
DOI: 10.1186/1471-2458-11-169
发表时间: 2011-03-17
期刊: BMC public health
影响因子: 4.5
作者:
Hermann S;Rohrmann S;Linseisen J;May AM;Kunst A;Besson H;Romaguera D;Travier N;Tormo MJ;Molina E;Dorronsoro M;Barricarte A;Rodríguez L;Crowe FL;Khaw KT;Wareham NJ;van Boeckel PG;Bueno-de-Mesquita HB;Overvad K;Jakobsen MU;Tjønneland A;Halkjær J;Agnoli C;Mattiello A;Tumino R;Masala G;Vineis P;Naska A;Orfanos P;Trichopoulou A;Kaaks R;Bergmann MM;Steffen A;Van Guelpen B;Johansson I;Borgquist S;Manjer J;Braaten T;Fagherazzi G;Clavel-Chapelon F;Mouw T;Norat T;Riboli E;Rinaldi S;Slimani N;Peeters PH
通讯作者: Peeters PH
DOI: 10.1038/oby.2008.432
发表时间: 2008-12-01
期刊: OBESITY
影响因子: 6.9
作者:
Beyerlein, Andreas;Toschke, Andre M.;von Kries, Ruediger
通讯作者: von Kries, Ruediger