Statistics: The Addition Rule

Understanding Statistics: The Addition Rule

A reliable way to reason about combined events is to start with actual outcomes instead of memorising a rule. Imagine a survey of one hundred students. Fifty-eight play a sport, forty-six play an instrument, and twenty-one do both.

The students in the shared group appear in each of the first two totals. Counting every student who does at least one activity gives fifty-eight plus forty-six minus twenty-one, which is eighty-three students.

Since the survey contains one hundred students, the probability is eighty-three out of one hundred. This counting view makes the subtraction feel necessary rather than arbitrary.

The shared group must come from the same sample space and the same situation as the two individual events. For example, a school may record students who take biology, students who take chemistry, and students who take both subjects during one term. It would not make sense to combine a biology total from one year with a chemistry total from another year.

In real data, unclear definitions cause many mistakes. Students should check what each category means, who was included in the data, and whether a person can belong to both groups. A person may have more than one label, so categories that sound different are not automatically separate.

Independence and mutual exclusivity are often confused. Independent events do not affect each other's chance, but they can still happen together. Getting a head on a coin toss and rolling an even number on a die are independent, yet both can occur in the same trial.

Their overlap is real and must be accounted for when finding the chance of either event. Mutually exclusive events are different because one outcome blocks the other. On one roll of a standard die, rolling a two and rolling a five cannot occur together.

The key habit is to ask whether the same trial can produce both event labels. If it can, there is an overlap to find.

Tables can make these problems much easier. A two-way table separates outcomes into four groups, those in A only, those in both, those in B only, and those in neither. To find the combined event, add the first three groups once each.

This method works especially well for survey results, medical screening data, sports records, and factory inspections. In a factory, an item might have a scratch, a wrong colour, or both defects. Counting the overlap twice would exaggerate the defect rate.

As a final check, the result cannot be larger than the whole sample. It should be at least as large as either individual event, because it includes every outcome from each event. These simple checks catch many calculation errors.