Ling has developed a new kind of antibiotic that she expects to kill $90\%$ of harmful bacteria when applied. She applied her antibiotic to a Petri dish full of bacteria, waited for it to take effect, and took a random sample of $200$ bacteria. She found that $87\%$ of them were dead.
Let's test the hypothesis that **the actual percentage of dead bacteria is $90\%$** versus the alternative that the actual percentage is *lower* than that.
The table below sums up the results of $1000$ simulations, each simulating a sample of $200$ bacteria, assuming there are $90\%$ dead bacteria.
**According to the simulations, what is the probability of getting a sample with $87\%$ dead bacteria or less?**
$\qquad$[[☃ input-number 1]]
Let's agree that if the observed outcome has a probability *less* than $1\%$ under the tested hypothesis, we will reject the hypothesis.
**What should we conclude regarding the hypothesis?**
[[☃ radio 1]]
Measured $\%$ of dead bacteria | Frequency
:-: | :-:
$84$ | $1$
$85$ | $9$
$86$ | $36$
$87$ | $43$
$88$ | $108$
$89$ | $163$
$90$ | $190$
$91$ | $178$
$92$ | $131$
$93$ | $77$
$94$ | $43$
$95$ | $17$
$96$ | $4$