Counting the Significant Findings

A screening lab runs $20$ independent hypothesis tests, each at significance level $\alpha = 0.05$.

Unknown to the lab:

• in $12$ of the tests the null hypothesis is genuinely true — there is no real effect;
• in the remaining $8$ tests a real effect is present, and at that effect size each of those tests has power $0.80$.

A test is called statistically significant when it rejects $H_0$.

What is the expected number of the 20 tests that come out statistically significant?