Preliminaries

This is a brief guide on how to use R and functions in tigerstats and related packages to do some very basic inferential statistics. We will give “templates” for the functions, accompanied by no-frills examples of their use. Consult the function tutorials or other Help documents to learn more about the options for each function.

One Factor Variable

Two Values

We recommend binomtestGC(). The template is:

\[binomtestGC(\sim variable,data=MyData,p=NullBelief).\]

binomtestGC(~sex,data=m111survey,
            p=0.50,
            alternative="two.sided",
            success="female")

More Than Two Values

chisqtestGC() is the function to use. The template is:

\[chisqtestGC(\sim variable,data=MyData,p=NullProbs).\]

chisqtestGC(~seat,data=m111survey,p=c(1/3,1/3,1/3))

Two Factor Variables

Both Variables Have Two Values

If both the explanatory and response variable have two values each, then you might want proptestGC(). The template is:

\[proptestGC(\sim explanaotry + response, data=MyData,p=0).\]

proptestGC(~sex+love_first,data=m111survey,
                p=0,
                alternative="two,sided")

One or Both Variables Have More Than two Values

If one or more of the variables has more than two values, stick with chisqtestGC(). The format is:

\[chisqtestGC(\sim explanatory + response, data=MyData).\]

chisqtestGC(~sex+seat,data=m111survey)

One Numerical Variable

The function is ttestGC(). Template:

\[ttestGC(\sim variable, data=MyData,mu=NullBelief).\]

ttestGC(~fastest,data=m111survey,mu=100,
          alternative="two.sided")

If the single numerical variable of interest is a difference of two numerical variables, then you are probably interested in the “paired” t-test:

\[ttestGC(\sim secondVar - firstVar, data=MyData,mu=0).\]

ttestGC(~height - ideal_ht,data=m111survey,
          mu=0,
          alternative="two.sided")

One Numerical and One Factor Variable

We assume that the factor variable has two values.

You probably want the “2-sample t-test”. The format is:

\[ttestGC(numerical \sim factor, data=MyData,mu=0).\]

ttestGC(fastest~sex,data=m111survey,
          mu=0,
          alternative="two.sided")

Two Numerical Variables

This topic is not extensively provided for in the tigerstats package, but if you encounter this situation you will want to learn about lm(), the “grown-up” version of lmGC().

R: Inferential Statistics