Function name

What it gives you

Arguments [square brackets denote optional arguments]

Comments

normalpdf

Height of the probability curve at a data point (x)

x [, mean , s.d.]

If mean and s.d. are omitted, they are assumed to be 0 and 1.

Note that s.d. does not necessarily mean s . For a sampling distribution of xbar (a very common situation), you’d punch in s /Ö n for s.d. For the normal approx. to a binomially distributed variable, you’d punch in Ö (np(1–p)) for s.d.

normalcdf

Cumulative area between two points

start, end [, mean , s.d. ]

Same comments as for normalpdf.

invNorm

Data value that gives you a certain cumulative area (starting from –¥ )

area [, mean , s.d. ]

For example, enter invNorm(.975) to find the upper z critical value for a 95% confidence interval.

tpdf

Height of the t distribution at a data point (x)

x , df

Rarely used.

tcdf

Cumulative area under the t distribution

start , end , df

Since we are usually computing tail probabilities, "end" is nearly always 99999.

PRGM INVT

The t score required to give a certain cumulative area (starting from –¥ )

Menu-driven (this is a program we wrote in class).

WARNING: Some versions of this program only display the result without storing it into Ans, so if you immediately start doing computations without retyping, you will get wrong results.

c ²pdf

Height of the chi-square distribution at a data point (x)

x , df

Rarely used.

c ²cdf

Cumulative area under the chi-square distribution

start , end , df

Since we are usually computing tail probabilities, "end" is nearly always 99999.

binompdf

Height of the binomial distribution at a point (i.e., the probability of getting exactly x successes in n trials)

n , p , x

binomcdf

Cumulative binomial probability of getting 0 through "goesthrough" successes in n trials

n , p , goesthrough

geometpdf

Probability of having first success on trial #x (where trials are binomial-style, indep.)

p , x

geometcdf

Probability of having first success on or before trial #x

p , x