Monkey Investigation6 Step Method: Steps 1-2
In-Class Activity Worksheet
Step 1: Ask a Research Question
1. In this investigation, what is the real-world question we want to answer?
Does Luc, our chimpanzee, prefer raw or cooked potatoes?
2. Write out three possible answers to the real-world question.
1. Yes, prefers cooked potatoes
2. No, prefer raw potatoes
3. No, has no preference between the 2
3. Which of the three answers claims that nothing interesting is happening? This answer is your null
hypothesis (H0).
The third one, No the monkey has no preference between the 2
4. Which of the remaining two answers is the answer that we would like to prove is correct? This answer
is your alternative hypothesis (Ha).
Yes, prefers cooked potatoes
Step 2: Design a Study
Import the dataset for your group from the Rguroo Dataset Repository and View it. Each row in the
dataset represents an observational unit and each column represents a variable.
5. What is the name of the chimpanzee whose results you are analyzing?
Luc the chimpanzee 0.5
2. Which of the following statements is most correct? Explain your reasoning.
a. Small values of the statistic (near 0%) provide a lot of evidence that the alternative hypothesis is
correct, while large values (near 100%) do not provide a lot of evidence
b. Large values of the statistic (near 100%) provide a lot of evidence that the alternative hypothesis
is correct, while large values (near 0%) do not provide a lot of evidence
c. Both small values of the statistic (near 0%) and large values of the statistic (near 100%) provide a
lot of evidence that the alternative hypothesis is correct
3. Sketch a number line showing values of your statistic. Draw an arrow in the direction of “more
evidence that the alternative hypothesis is correct.”
4. In Rguroo, perform a simulation-based test of significance. Insert below the “P-Value Graph.” This dot
plot shows the null distribution of your statistic; that is, the number of times we simulated 0/12, 1/12,
2/12, …, up to 12/12 cooked potatoes chosen by a chimpanzee for which we know the null hypothesis is
true.
5. What does the x-axis show on this graph? What does the y-axis show on this graph?
X-axis is the proportion of cooked potatoes, and the y-axis reveals the number of simulations/samples per
proportion
6. This graph has three parts: a green triangle, a region of blue dots/bars, and a region of red dots/bars.
Which parts represent outcomes from the chance model? Which parts represent the outcome from your
actual (observed) chimpanzee?
The part that represents the chance model is the blue bars. Outcome from our observation is the green
triangle
7. The p-value is a number between 0 and 1 that represents the strength of the evidence contained in the
observed sample. Fill in the blanks in the paragraph explaining how to compute the p-value for this test.
Since the alternative hypothesis contains a _____greater than____ sign, the further a statistic value is to
the right (look at your answer to Question #3) on the graph of the null distribution, the more evidence
that sample provides that the null hypothesis is false and the alternative hypothesis is true
.
Therefore, to find the p-value, find the proportion of simulated sample proportions that are _____as
much or more than________ your observed value of
8/12 .
8. Rguroo has already computed the p-value. How is this process illustrated on the graph? (Think about
where the dividing line is between blue and red bars.)
By the amount of times in which the graph chose that they had 8/12 or more times of choosing cooked
potatoes
9. What is the p-value for your test, according to Rguroo?
0.1933
Step 5: Make Conclusions
We typically use a “significance level” as our threshold for determining whether we have “strong enough”
evidence to conclude that our null hypothesis is not a plausible (believable) explanation
for what we observed.
1. Suppose we have strong enough evidence to reject our null hypothesis. Does this imply that
the alternative hypothesis is correct? Why or why not? Yes it does imply that thee alternative is correct, because
it is either the null or the alternative hypothesis and if it is not the null, then it must be the alternative.
2. Suppose we do not have strong enough evidence to reject our null hypothesis. Does this imply
that the null hypothesis is correct? Why or why not? Yes because we do not have enough evidence to reject
The null hypothesis.
3. Do smaller or larger p-values indicate stronger evidence that the null hypothesis is not a
plausible explanation? Explain your reasoning. Larger p-values lead us to fail to reject the null, thus making
It a non-plausible explanation.
4. Open your Rguroo work from last class. According to Rguroo, what is the p-value for your
Test? P value equals 0.1933
5. Typically we use 5% (the cutoff for “strong evidence”) as our significance level. Using this
significance level, do you have “strong enough” evidence to reject the null hypothesis? Why or
why not? No we do not have strong enough evidence to reject the null as our p-value is greater than 5%.
6. Is it plausible that your chimpanzee has no preference and so was “randomly choosing”
between the cooked and raw potato? Why or why not? Yes it is very likely that he was randomly choosing
Between the two, because the p-value reveals that it is greater than 5%.
7. Is it plausible that your chimpanzee prefers cooked potatoes over raw potatoes? Why or why
Not? We do not know, because we we are failing to reject the null hypothesis
8. Use your answers to Questions #6 and #7 to answer your research question.
We do not know
Step 6: Look Back and Ahead
9. Could you have committed a Type I Error (false alarm) with your conclusion in Question #8? Why or why not?
Originally we did by thinking we could reject the null, but after doing the coin toss we realized that it was
Likely due to chance.
10. Could you have committed a Type II Error (missed opportunity) with your conclusion in
no
Question #8? Why or why not?
