Literature DB >> 27829378

Erratum to: A Monte Carlo simulation study comparing linear regression, beta regression, variable-dispersion beta regression and fractional logit regression at recovering average difference measures in a two sample design.

Christopher Meaney1, Rahim Moineddin2.   

Abstract

Year:  2016        PMID: 27829378      PMCID: PMC5103514          DOI: 10.1186/s12874-016-0256-6

Source DB:  PubMed          Journal:  BMC Med Res Methodol        ISSN: 1471-2288            Impact factor:   4.615


× No keyword cloud information.

Erratum

After publication of the original article [1], the authors noticed an error in Fig. 1. The legend included in the original sub-plot of Fig. 1 was labelled “phi = 500 (p = 25, q = 475)”; however, the figure title suggested phi = 1000.
Fig. 1

Various forms of the beta density for varying shape parameters {p,q}. Top left panel: We fix the mean equal to 0.5 and plot the resulting beta densities for varying dispersion parameters. Top right panel: We fix the mean equal to 0.05 and plot the resulting beta densities for varying dispersion parameters. Bottom left panel: We fix the dispersion parameter equal to 100 and plot the resulting beta densities for varying mean parameters. Bottom right panel: We fix the dispersion parameter equal to 5 and plot the resulting beta densities for varying mean parameters

Various forms of the beta density for varying shape parameters {p,q}. Top left panel: We fix the mean equal to 0.5 and plot the resulting beta densities for varying dispersion parameters. Top right panel: We fix the mean equal to 0.05 and plot the resulting beta densities for varying dispersion parameters. Bottom left panel: We fix the dispersion parameter equal to 100 and plot the resulting beta densities for varying mean parameters. Bottom right panel: We fix the dispersion parameter equal to 5 and plot the resulting beta densities for varying mean parameters An updated version of Fig. 1 is published in this erratum, where the legend has been updated to “phi = 1000 (p = 50, q = 950)” to be consistent with the figure title.
  1 in total

1.  A Monte Carlo simulation study comparing linear regression, beta regression, variable-dispersion beta regression and fractional logit regression at recovering average difference measures in a two sample design.

Authors:  Christopher Meaney; Rahim Moineddin
Journal:  BMC Med Res Methodol       Date:  2014-01-24       Impact factor: 4.615

  1 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.