Table/Graph Format and Interpretation

This page is designed to be a quick reference for the preparation of various kinds of tables and graphs for courses in the department.  It will be expanded as time goes on.  Be sure that any tables or graphs that you create follow the format specified below.

Basic Cross-Tabulations  

Your independent variable should always be the column variable and your dependent variable should always be the row variable.  You should use column percentaging, so that the percentages add to 100 reading down the columns.  In the table below, Education is the independent (column) variable and Commun.Sp (whether the respondent thinks that communists should be allowed to make a speech in their community) is the dependent (row) variable.  In other words, the cross tabulation is designed to test whether support for free speech by communists is related to level of education. The correlation statistic used in a cross-tabulation like this is Cramer's V. In the Student Versions of MicroCase, the significance level is indicated by one or two asterisks (.05 or .01 levels respectively). However, in the professional versions of MicroCase, asterisks are not used--you must click on Statistics and examine the "p" value separately. If Cramer's V is significant, a rough rule of thumb for interpreting the strength of the relationship is as follows:

Having trouble remembering the distinction between an independent and a dependent variable?  Click on the sound icon above to hear a RealPlayer audio clip explaining the difference.

Note: the table is highly significant (p=0.00).  Support for free speech by communists increases with education. The correlation is moderate.

Scatterplots

For scatterplots your independent variable should be situated along the horizontal axis and your dependent variable along the vertical axis.  The correlation statistic generally used for scatterplots is Pearson's r. If the Pearson r statistic meets the test of significance, a rough rule of thumb for interpreting the strength of the relationship is:

In the scatterplot above, we see that nations with higher levels of inequality (measured by the Gini index) tend to provide lower levels of physical quality of life for their citizens. The correlation is moderate.