tigerstats (version 0.3.2)

ttestGC: t-Procedures

Description

t-tests and confidence intervals for one and two samples.

Usage

ttestGC(x=NULL,data=parent.frame(),mean=numeric(),sd=numeric(),n=numeric(),
 mu=NULL,alternative=c("two.sided","less","greater"),var.equal=FALSE,
 conf.level=0.95,graph=FALSE,first=NULL,verbose=TRUE)

Arguments

x

If not NULL, then must be a formula. If a formula, then data must be a dataframe. For one sample t-procedures, x is of the form ~var. For two-sample procedures, x is of the form resp~exp, where exp is factor with two values. If x is of form ~var1-var2, then matched pairs procedures are performed.

data

A data frame containing variables in formula x. If some variables are not in data, then they are searched for in the parent environment.

mean

When not NULL, contains sample mean(s). Length 1 for one sample t-procedures, Length 2 for two-sample procedures.

sd

When not NULL, contains sample standard deviation(s).

n

When not NULL, contains sample size(s).

mu

Contains the null value for the parameter of interest. If not set, no test is performed.

alternative

"two.sided" requests computation of a two-sided P-value; other possible values are "less" and "greater".

var.equal

When FALSE, use Welch's approximation to the degrees of freedom.

conf.level

Number between 0 and 1 indicating the confidence-level of the interval supplied.

graph

If TRUE, plot graph of P-value.

first

If assigned, gives the value of the explanatory variable that is to count as the first sample.

verbose

Indicate how much output goes to console

Value

A list of class "GCttest" Components of the list that may be usefully queried include: "statistic", "p.value", and "interval".

Examples

# NOT RUN {
#One-sample t, 95%-confidence interval only:
ttestGC(~fastest,data=m111survey)

#For other confidence levels, set argument conf.level as desired.  For 90%-interval:
ttestGC(~fastest,data=m111survey,conf.level=0.90)

# One-sample t, 95%-confidence interval and two-sided test with H_0:  mu = 100:
ttestGC(~fastest,data=m111survey,mu=100)

#Two-sample t, 95%-confidence interval only:
ttestGC(fastest~sex,data=m111survey)

#control order of groups with argument first:
ttestGC(fastest~sex,data=m111survey,first="male")

# Matched pairs, confidence interval with one-sided test, H_0: mu-d = 0:
ttestGC(~ideal_ht-height,data=m111survey,mu=0,alternative="greater")

#Summary data, one sample, one-sided test with H_0:  mu = 52.5:
ttestGC(mean=55,sd=4,n=16,mu=52.5,alternative="greater")

#Summary data, two samples:
ttestGC(mean=c(50,55),sd=c(3,4),n=c(25,40),mu=0)
# }