Contingency Table
Frequency Tables
Frequency tables show distribution just for one table
| none | small | big | Total |
|---|---|---|---|
| 549 | 2827 | 545 | 3921 |
When we replace counts with rations, we get a Relative Frequency Table:
If we replaced the counts with percentages or proportions, he table would be called a relative frequency table.
| none | small | big | Total |
|---|---|---|---|
| 549 / 3921 | 2827 / 3921 | 545 / 3921 | 1 |
Contingency Tables
A contingency table is a table that summarizes data for two Categorical Variables
- It describes the Joint Distribution of these variables
E.g. consider
- Spam = { Spam, Not Spam } and
- Number = { none, small, big }
| none | small | big | Total | |
|---|---|---|---|---|
| spam | 149 | 168 | 50 | 367 |
| not spam | 400 | 2659 | 495 | 3554 |
| Total | 549 | 2827 | 545 | 3921 |
In R a contingency table can be obtained by table command
table(var1, var2)
So, a contingency table of 2 categorical variables is a table that shows the frequency of observations for the two variables as a combination.
2x2 Contingency Tables Analysis
Confusion Matrix? http://stats.stackexchange.com/questions/107895/proper-analyses-for-2x2-contingency-tables