PHatonesidedTestStatistics {reproducer}

PHatonesidedTestStatistics

Description

This function is a helper function for meta-analysis of experiments using PHat as an effect size. It returns the 100*(1-alpha)

Usage

PHatonesidedTestStatistics(
  effectsize,
  effectsize.variance,
  effectsize.df = 0,
  alpha = 0.05,
  alternative = "greater"
)

Arguments

effectsize

The overall estimate of the centralized PHat (i.e. Phat-0.5) from a group of effect sizes to be meta-analysed

effectsize.variance

The estimate of the variance of the overall estimate ofPHat

effectsize.df

The total degrees of freedom for the set of effect sizes. If effectsize.df>0, the confidence intervals, pvalues and significance test use the t-distribution probability values. If effectsize.df=0 (default), the confidence intervals, the pvalues and significance test use the normal distribution probability values.

alpha

The significance level (default 0.05) used to control the significance tests and calculation of confidence limits.

alternative

Specifies the type of significance test and can take the values "less" or "greater" (default).

Value

ES.test The value of the t-statistic

ES.pvalue The p-value of the two-sided t-test if the parameter d.df>0, or the normal probability value if d.df=0

ES.sig The significance of the statistical test of the d.tvalue return value at the alpha level for one sided tests and aplha/2 for two sided tests as specified by the input parameter alternative.

ES.ci.lower The lower 100*(1-alpha/2)

ES.ci.upper The upper 100*(1-alpha/2)

Author(s)

Barbara Kitchenham and Lech Madeyski

Examples

PHatES=mean(c(0.92,0.6,0.48,0.72,0.88))-0.5
PHatESvar=sum(c(0.01,0.04,0.05,0.04,0.01))/25
PHatdf=sum(c(6.63,6.63,5.08,5.61,8))
#PHatonesidedTestStatistics(effectsize=PHatES,effectsize.variance=PHatESvar,effectsize.df=PHatdf)
# A tibble: 1 x 5
#  ES.test ES.pvalue ES.sig ES.ci.lower ES.ci.upper
#    <dbl>     <dbl> <lgl>        <dbl>       <dbl>
#1    2.84   0.00389 TRUE        0.0888       0.351
#PHatonesidedTestStatistics(effectsize=PHatES,effectsize.variance=PHatESvar,effectsize.df=0,
# alternative="less")
# A tibble: 1 x 5
#  ES.test ES.pvalue ES.sig ES.ci.lower ES.ci.upper
#    <dbl>     <dbl> <lgl>        <dbl>       <dbl>
#1    2.84     0.998 FALSE       0.0926       0.347