Math 143 F
Probability and Statistics
Spring 2000

What do I do with this Data?

When deciding how to analyze data (and how to present it graphically), one of the first questions to ask is "what kind of data do I have?" Most graphical presentations, numerical summaries, and inference procedures only work with certain kinds of data. Keeping track of this kind of information will greatly reduce the number of possibilities and help you avoid trying to apply a procedure in inappropriate setting.

One variable

The simplest situation is when we are dealing with only one variable at a time. That variable may be either quantitative or categorical, and that determines what types of tools are available to us:
Categorical Quantitative
pie/bar charts
calculate proportions
1-proportion Z- procedures
histogram
stemplot
boxplot (5-number summary)
mean, standard deviation
1-sample T- and Z- procedures

Two (or more) variables

If we have two (or more) variables to analyze at the same time, we need to look at the types of all the variables involved. Bracketed items below imply more than one quantitative variable.
Categorical Quantitative
C
a
t
 side-by-side pie/bar charts
two-way tables
Chi-square
2-sample procedures for proportions
 side-by-side boxplots
back-to-back stemplots
side-by-side histograms
2-sample T- and Z- procedures
[ANOVA]
sign-test
Q
u
a
n
t
 (same as upper right corner) scatter plot
paired tests (convert to one variable)
regression
[multiple regression]

Back to Math 143 Home Page

This page maintained by:
Randall Pruim
Department of Mathematics and Statistics
Calvin College
rpruim@calvin.edu

Last Modified: Thursday, 11-Jan-2001 16:02:49 EST