1
$\begingroup$

Let's say, we have a table

$$\begin{array}{c|c|c|} & \text{Smoking} & \text{Non-Smoking} \\ \hline \text{Male} & 70 & 30 \\ \hline \text{Female} & 40 & 10 \\ \hline \end{array}$$

shows the number of patients.

I would like calculate the joint occurrence probability of being a male and being smoking. Moreover, I assume gender and smoking events are independent of each other.

Is the requested probability = $\frac{70+30}{70+30+40+10}.\frac{70+40}{70+30+40+10}$=$\frac{100}{150}.\frac{110}{150}$ ?

OR is it $\frac{70}{150}$ ? 70 is the number in the cell of intersection of M and S.

$\endgroup$
  • 1
    $\begingroup$ If you pick out one of the patients then the probability that it is a smoking man is simply $\frac7{15}$. Also your table is a strong indication that gender and smoking are not independent at all. $\endgroup$ – drhab Nov 24 '16 at 11:39
1
$\begingroup$

We have $P(male) = \frac{100}{150}$ and $P(smoker)=\frac{110}{150}$. If the variables were really independent then $P(male \cap smoker) = P(smoker) \cdot P(male) = \frac{110}{150} \cdot \frac{100}{150} $. However, I think the variables are not independent. In the case of independence, you would expect the value $\frac{100 \cdot 110}{150}$ in the upper left corner of the matrix, i.e. about 73 insted of 70. If they were independent, both of your suggested methods would give the same result.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.