Earthquakes can be broken into two classes based on the directions the earth moves when they fracture. The classes can be compared across time, for whether earthquakes occur in a given region in a small time period, and if they occur in the next small time period, and it can be added up across several time periods, below is a table for a region off Indonesia over a 30 year period:
| | Second time period, earthquake happens | Second time period no earthquake happens | Marginal Sums |
| :--- | :---: | :---: | :---: |
| First time period, Earthquake happens | 148 | 274 | 422 |
| First time period no earthquake happens | 276 | 2626 | 2902 |
| Marginal sums | 424 | 2900 | 3324 |
Is happening of an earthquake in one time period statistically independent of happening in another time period? Test at the .01 level
These are earthquakes of the same type, Which cells have higher than expected occurrence if independence is true. (Use the deviation table).