STEP 1: Ask a research question

1. State the research questions the researchers investigate in this study.

Can free ranging Rhesus monkeys understand human cummunicative gestures?

STEP 2: Design a study and collect dat.

2. What are the observational units and sample size of the individual experiments in this study?

The observational units are each instance of the monkeys responding in a trial. The sample size is N = 40 for each experiment.

3. What variables did the researcher collect? State whether each variable is quantitative or categorical.

The variable collected is whether or not the monkeys pick the right box and it is categorical.

4. Summarize the procedure used in the experiment. Your explanation should summarize the description in the article in a clear and concise manner.

To determine how rhesus monkeys react to different communicative gestures, the researchers set up boxes to get the monkeys to focus on one area and then presented them with different communicative gestures that are both from their species and from humans. The researchers observed if the monkeys responded correctly to the communicative gestures to analyze whether or not the monkeys understood the gesture.

5. Why did the researchers conduct trials with an apple slice and without an apple slice?

To understand if the monkeys actually understood the gesture or if the food made them understand.

6. Why did the researchers use multiple gesture types?

To see if they only understood species specifci gesture or if they could infer the meaning of other gestures they aren’t used to.

Questions 7-17 refer to the food-presnt communicating gesture condition experiment.

7. Describe the parameter of interest in words. Assign the proper mathematical symbol to the parameter of interest.

The long run proportion that the monkeys picked the correct box based on the gesture.

8. In words, state the null and alternative hypothesis tested.

Null: The monkeys are randomly guessing which box is correct.

Alternative: The monkeys actually understand the gesture and are not guessing the box.

9. Using symbols, state the null and alternative hypothesis tested.

Null: π=0.50

Alternative: π>0.50

STEP 3. Explore the data.

10. When you look at Figure 2 do you believe that rhesus monkeys are being influenced by experimenter’s actions?

Yes, more often than not when the researcher made a gesture the monkey would follow it to the correct box.

11. Calculate the relevant statistic for this portion (food-present communicating gesture condition) of the experiment. Use the appropriate symbol.

p̂=30/40

12. Is it possible to observe the above sample statistic even if all of the rhesus monkeys are randomly selecting a box? Explain.

Yes, 30/40 is possible if they are guessing by random chance but it is unlikely.

STEP 4: Draw inferences.

13. Use the 3S strategy to investigate how much evidence the sample data provide to support your conjecture (hypothesis).

a) Statistic: The monkey chose the correct box 3/4 of the time in the 40 trials.

b) Simulate: Use the One Proportion applet to simulate 1000 repetitions of this study. Provide a screen shot of your completed simulation.

Screenshot of Applet

Screenshot of Applet

i. What are you assuming when you simulate your null distribution?

The monkey is just guessing the box.

ii. What is the center of your null distribution and does it make sense that this is the center? Explain.

Centered at about 0.50 or 20/40 which makes sense as if the monkey were to be just guessing that would be the proportion of correct guesses and this is the random chance model.

c) Strength of evidence:

i: When looking at the null distribution you have simulated, is your observed statistic a very unlikely result if your null hypothesis is true?

Yes.

ii. Determine the p-value from your simulation analysis. Interpret what the p-value represents.

p=.0020. We expect to observe a sample proportion of 0.75 or greater in 0.20% of trials.

14. Use the theory-based test to determine the p-value for the food-present communicating gesture condition experiment. Provide the R code and output for the theory-based test.

prop.test(x = 30, n = 40, p = 0.5, alternative = "greater", correct = F)
## 
##  1-sample proportions test without continuity correction
## 
## data:  30 out of 40, null probability 0.5
## X-squared = 10, df = 1, p-value = 0.0007827
## alternative hypothesis: true p is greater than 0.5
## 95 percent confidence interval:
##  0.6240271 1.0000000
## sample estimates:
##    p 
## 0.75

a) Are the validity conditions met for this test? Explain.

Yes, the validity conditions are met becasue there are atleast 10 successes and failures.

b) What is the p-value?

p=0.00078

c) How does your p-value compare to the p-value presented in the paper?

Our p-value of 0.00078 is more precise than their p-value of 0.0010 but they are extremely close.

STEP 5: Formulate conclusions.

15. Based on your p-values, how much evidence does the data provide against your null hypothesis? Why?

We have very strong evidence against he null hypothesis because our p-value of 0.00078 is less than 0.01 which is the baseline threshold for verys strong evidence for or against a null.

16. Summarize conclusions you can draw from your analysis. Also explain the reasoning process behind your conclusion.

The monkeys are unlikely to be guessing when presented with food-present communicative gestures because there is very strong evidence against the null hypothesis.

17. How broadly are you willing to generalize your conclusions? Would you be willing to genrealize your conclusions to all rhesus monkeys, primates, or mammals?

This conclusion can be generalized to all rhesus monkeys because the introduction stated that apes were incapable of understanding communicative gestures and our sample size was comprised of only rhesus monkeys so it wouldn’t make sense to generalize our conclusion any further than just rhesus monkeys.

STEP 6: Look back and ahead.

18. How do the researcher’s findings support or contradict previous research? You only need to discuss one comparison discussed in the article.

Hauser, Glynn, and Wood reference the work done by Gergely and Csibra in 2003 when they did work to understand the teleological stance that actions can be infered by human children even if the intentions are ambiguous and this data is consistent with rhesus monkeys as well.