asbio (version 1.9-7)

ci.mu.z: Z and t confidence intervals for mu.

Description

These functions calculate t and z confidence intervals for \(\mu\). Z confidence intervals require specification (and thus knowledge) of \(\sigma\). Both methods assume underlying normal distributions although this assumption becomes irrelevant for large sample sizes. Finite population corrections are provided if requested.

Usage

ci.mu.z(data, conf = 0.95, sigma = 1, summarized = FALSE, xbar = NULL,
fpc = FALSE, N = NULL, n = NULL, na.rm = FALSE)
ci.mu.t(data, conf = 0.95, summarized = FALSE, xbar = NULL, sd = NULL, 
fpc = FALSE, N = NULL, n = NULL, na.rm = FALSE)

Value

Returns a list of class = "ci". Default printed results are the parameter estimate and confidence bounds. Other invisible objects include:

Margin

the confidence margin.

Arguments

data

A vector of quantitative data. Required if summarized = FALSE

conf

Confidence level; 1 - P(type I error).

sigma

The population standard deviation.

summarized

A logical statement specifying whether statistical summaries are to be used. If summarized = FALSE, then the sample mean and the sample standard deviation (t.conf only) are calculated from the vector provided in data. If summarized = FALSE then the sample mean xbar, the sample size n, and, in the case of ci.mu.t, the sample standard deviation st.dev, must be provided by the user.

xbar

The sample mean. Required if summarized = TRUE.

fpc

A logical statement specifying whether a finite population correction should be made. If fpc = TRUE the population size N must be specified.

N

The population size. Required if fpc=TRUE

sd

The sample standard deviation. Required if summarized=TRUE.

n

The sample size. Required if summarized = TRUE.

na.rm

Logical, indicate whether NA values should be stripped before the computation proceeds.

Author

Ken Aho

Details

ci.mu.z and ci.mu.t calculate confidence intervals for either summarized data or a dataset provided in data. Finite population corrections are made if a user specifies fpc=TRUE and provides some value for N.

References

Lohr, S. L. (1999) Sampling: Design and Analysis. Duxbury Press. Pacific Grove, USA.

See Also

pnorm, pt

Examples

#With summarized=FALSE 
x<-c(5,10,5,20,30,15,20,25,0,5,10,5,7,10,20,40,30,40,10,5,0,0,3,20,30)
ci.mu.z(x,conf=.95,sigma=4,summarized=FALSE)
ci.mu.t(x,conf=.95,summarized=FALSE)
#With summarized = TRUE
ci.mu.z(x,conf=.95,sigma=4,xbar=14.6,n=25,summarized=TRUE)
ci.mu.t(x,conf=.95,sd=4,xbar=14.6,n=25,summarized=TRUE)
#with finite population correction and summarized = TRUE
ci.mu.z(x,conf=.95,sigma=4,xbar=14.6,n=25,summarized=TRUE,fpc=TRUE,N=100)
ci.mu.t(x,conf=.95,sd=4,xbar=14.6,n=25,summarized=TRUE,fpc=TRUE,N=100)