Hypothesis Testing III

Lecture 23

Dr. Elijah Meyer

Duke University
STA 199 - Spring 2023

April 7th, 2023

Checklist

– Clone ae-22

– Lab 9 Due Tuesday (April 11th)

– HW-6 (Statistics Experience) due April 28th

– Project Draft Report due Friday (April 7th)

– HW 5 Due Thursday (April 13th)

– Good Conversation in Slack! Check it out

– Second Peer Evaluation (April 10th)

– Exam 2 starts April 14th

Research Participation

– Research team creating an intro data science assessment

– We would like to have students in an intro data science course take the assessment

– Pin down how students think about things so that we can better phrase our questions

– Located at the end of HW 5

– Your responses matter!

Project Announcement

– Include literature review in report

Lab - Peer Review

– Will provide feedback via issues tab on 2 projects

– Attendance will be taken to track participation

– If you can not attend, you are expected to create a separate issue tab and provide additional feedback

Warm Up

\(\mu_s = \mu_v\)

\(\mu_s \neq \mu_v\)

Please write out an appropriate interpretation, decision, and conclusion in the context of the problem.

Goals for Today

– Use theory methods for 1 categorical explanatory 1 quantitative response

– Use permutation and theory methods for 2 categorical variables

Standardized Statistic

Standardized test statistics are used in hypothesis testing

Standardized test statistic: (statistic - null value)/(standard error of the statistic).

Standardized test statistics are a way for you to compare your results to a “normal” distribution of statistics.

The standardized test statistic represents the number of SEs your statistic is away from the center of the null distribution.

ae-22

Two Categorical Variables

– What is our parameter of interest in proper notation?

– What is our sample statistic in proper notation?

Two Categorical Variables

\(\pi_1 - \pi_2\)

\(\hat{p_1} - \hat{p_2}\)

Two Categorical Variables

– Our parameters and statistic change

– We model our statistic using a normal distribution

Wrap Up

We can calculate confidence intervals and conduct hypothesis tests for a combination of categorical and quantitative variables