Lab

Name1*

First NameLast Name
Name2*

First NameLast Name
E-mail*
example@example.com
Section

Section 1 (9-10am)Section 2 (10-11am)

Investigation 1.2: Can wolves understand human cues?

Previous research has demonstrated that domesticated dogs can be trained to understand human cues, such as looking or pointing at an object. But it was not clear how the animals would perform with “behavioral cues” showing intent such as reaching for, but not obtaining, or trying to open an object. Lampe et al. (2017) gave captive wolves, pack dogs living in identical conditions to the wolves, and pet dogs living with families a series of “object-choice tasks.” A table was placed outside a fenced compartment, and a container was placed at each end of the table, one containing food and one empty. The experimenter would give the cue as to which container had the food and then the animal would touch one of the two targets next to the containers. A 6-year-old female timber wolf, Yukon, chose the intended container in 6 of the 8 trials with behavioral cues.

(a) Identify the observational units for Yukon. [Hint: What is the sample size? Deciding that will help you define the observational units.]
(b) Identify the variable of interest for these observational units. [Hint: The variable is the question posed to each observational unit.] Is the variable quantitative or a categorical?
(b) Identify the variable of interest for these observational units. [Hint: The variable is the question posed to each observational unit.] Is the variable quantitative or a categorical?
(c) If Yukon was purely guessing, how often would you expect her to identify the correct container?

We will think of the sample of 8 attempts as identical observations from a random process, and we are choosing between two possibilities:
(1) Yukon can understand behavioral cues,
(2) Yukon does not consistently understand behavioral cues and is guessing randomly.

Definition: In drawing conclusions beyond our sample data to the underlying random process, we will often be choosing between two competing claims about the underlying process:
• The null hypothesis, which is the “by chance alone” explanation;
• The alternative hypothesis, which is usually what the researchers are hoping to show.

In Investigation 1.1, the null hypothesis was that infants (in general) choose equally among the two toys in the long run. The alternative hypothesis was that infants have a genuine preference for the helper toy.

(d) Identify the above possibilities as the null and alternative hypotheses.

	Hypothesis
(1) Yukon can understand behavioral cues
(2) Yukon does not consistently understand behavioral cues and is guessing randomly.

(e) So our simulation model is going to assume the null hypothesis is true and we will see how unusual it is to find 6 successes and 2 failures in 8 attempts. Design the simulation:

	Simulation model
One trial represents
Success corresponds to
Hypothesized probability of success
Number of trials in each repetition

Let's turn to the computer to repeat this random process many many times.

• Use the One Proportion Inference applet to carry out 1,000 repetitions of the simulation.

(f) Check the Summary Statistics box and report the mean and standard deviation of your simulated distribution of “could have been” outcomes.

Mean Standard deviation

Number of successes
To measure how often or rarely results like this happen by chance alone, we aren't only going to see how often we found 6 successes in the simulation (because we can't really compare this value across studies with different sample sizes), we will count how often we see 6 or something more extreme (in the direction of the alternative hypothesis).
(h) In the applet, enter 6 in the Count samples box and keep the “as extreme as” button set to greater than or equal to. Report the proportion of the repetitions with 6 or more heads (correct choices).
(i) In the applet, check the Exact Binomial box. How does this "exact probability" match up to your earlier simulation results?
OPTIONAL: Include a screen capture of your applet output. Show the input and the output details.

Cancelof
(j) Write a one-sentence interpretation of this probability in this context. Key components to include: what random process is being repeated, what is the null hypothesis that we are assuming to be true, what is the observed result we are considering, which outcomes are we considering to be more extreme?
(k) Based on this analysis, do these data provide convincing evidence that wolves can understand behavioral cues? Explain your reasoning in detail/provide supporting evidence.
Should be Empty:

Investigation 1.2: Can wolves understand human cues?