test: Calculates the statistics for hypotesis tests

In cgu-dados/auditsampling: A auxiliary package to calculate some basic things in sampling. Most used para CGU auditors.

Description

This function calculates de statistics for hypotesys tests considering the following hypotesis:

H_{0}:\text{ } μ \neq μ_{0}

H_{0}:\text{ } \bar{p} \neq p_{0}

H_{0}:\text{ } μ ≤q \text{ or } ≥q μ_{0}

H_{0}:\text{ } \bar{p} ≤q \text{ or } ≥q p_{0}

Usage

test(sample.data, num.col = NULL, prop.col = NULL, mu0 = NULL,
  p0 = NULL, alpha, alternative = NULL, labels = NULL,
  plot.graph = FALSE)

Arguments

sample.data	The source data.frame from which the funtion will search the columns.
num.col	Vector of the numeric column numbers.
prop.col	Vector of the binary column numbers.
mu0	Hypothesized mean.
p0	Hypothesized proportion.
alpha	Level of significance.
alternative	1 if is a greater/lesser test and 2 for equal/different one.
labels	Label vector for the outputs.
plot.graph	Plot a confidence intervals for proportions only. Not implemented yet.

Value

data.frame containing the following:

var

The label of the variable/parameter.

point

Point estimate.

tolerance

The hypothesized mean/proportion.

sdd_err

Standard error.

z_value

Statistics z.

pval

P-value.

power

Test power, defined as (1 - β).

Examples

n=168
set.seed(101)
dados <- data.frame(CONTROLE1=rbinom(n,1,prob = 0.4),
                    CONTROLE2=rbinom(n,1,prob = 0.02),
                    CONTROLE3=rbinom(n,1,prob = 0.1),
                    CONTROLE4=rbinom(n,1,prob = 0.01),
                    CONTROLE5=rbinom(n,1,prob = 0.2),
                    CONTROLE6=rnorm(n,10,3),
                    CONTROLE7=rnorm(n,250,20))
test(sample.data = dados, num.col = 6:7, prop.col = 1:5, alpha = 0.05,
     mu0 = c(10,100),p0=c(0.05,0.05,0.10,0.10,0.1),
     alternative = c(rep(1,5),rep(2,2)))

cgu-dados/auditsampling documentation built on March 24, 2021, 9:38 a.m.