1. Names and Faces Here is an interesting study exploring whether someone’s face tends to “match” their name:
https://www. apa. org/news/press/releases/2017/02/loo k-like-name

This Writing Homework helps us think through how one might give statistical proof for this claim.
We design the following experiment:
We select a random sample of students from Shoreline and show them a picture of another student from Shoreline along with 5 choices for possible names for that person (assume that the person in the experiment doesn’t already know the person pictured). That person then tries to guess the name of the person in the picture from the choices.

Here is an example picture from the study:

A. Jacob
B. Dan
C. Josef
D. Nathaniel
E. Michael

State the Null and alternative hypotheses:
If everybody involved just randomly guessed for each picture, what percent of names would they get correct?
This is the null hypothesis. Write this as a statement involving “p” and an equals sign “=“.
We think that there actually is a relationship between names and faces. If this were true, do we think people would guess at a better rate or worse rate then the amount above?
This is the alternative hypothesis. Translate this into a statement using “p” and an inequality sign.
Maybe here is a good time to choose an alpha-level also.
Check Assumptions and conditions:
We get a randomly chosen group of students from Shoreline Community College.
We show each of them several pictures of faces of other students (that they don’t know).
Overall, we have n = 150 total trials of this experiment.

Does a sample of 150 trials satisfy the assumptions and conditions necessary to use the Central Limit Theorem? Why or why not?
Construct our model:
Assume the null hypothesis is true. Construct the normal distribution that the central limit theorem guarantees us.
Analyze our observation:
In our study, we observe that of our 150 trials, 38 guesses were correct.
Plot this observation on our normal model, and find the area in the tail of this distribution. This area is called the “p-value.”

Interpret the conclusion:
Is this “p-value” large or small? Is this event usual or unusual? Is it unusual enough that you doubt the null hypothesis?
In other words, do you think the rate at which people got the answer correct is significantly different than if they were just guessing randomly?
Interpret this conclusion to decide if there is an apparent relationship between names and faces.
Types of Errors
Describe what Type I and Type II errors are for this hypothesis test and describe the possible consequences of making them.
Experiment Meta-analysis
Even though the step-by-step statistics allow us to do the analysis from a numerical perspective, it is still up to us, the researchers, to think about our experiment design.

Give three possible sources of bias in this experiment design. That is, name three things about the way we are doing this experiment (lurking or confounding variables) that may produce results that give us a skewed perception of reality.