Course Website STAT 101: Data Analysis and Statistical ... · Data Analysis and Statistical Inference ... Lecture slides will be posted on the course website ... Project 1 individual

8/28/2012

1

Statistics: Unlocking the Power of Data Lock5

STAT 101: Data Analysis and

Statistical Inference

Professor Kari Lock Morgan

[email protected]


Course Website: http://stat.duke.edu/courses/Fall12/sta101.002/

Sakai: https://sakai.duke.edu/portal/site/STAT101_Fall12

Course Website

Syllabus


Lecture Slides

Lecture slides will be posted on the course website

I’ll try to post them the night before, so you can print them out if you want

The slides posted the night before will NOT be the complete version (I want you to think during class, so won’t give you all the answers). Complete slides will be posted after class.


Textbook

Statistics: Unlocking the Power of Data

by Lock, Lock, Lock Morgan, Lock, and Lock

(Not published yet – you get an advance custom version)


Other Course Materials

Clicker: i>clicker

Available at the bookstore, Amazon, or from

previous students at this google doc

Clicker grading will begin on 9/10

Register your clicker at

http://www.iclicker.com/support/registeryourclicker/

for Student ID use your NetID.

Calculator

must support normal, t, 2, and F distributions


Support

My Office Hours: (in Old Chemistry 216) 2 – 4 pm Monday 1 – 2:30 pm Thursday

Statistics Education Center: (in Old Chem 211A) 4 – 9 pm Sunday – Thursday in Old Chem 211A

Email: Email your TA or [email protected] NOTE: I will only reliably reply to email 4:30 – 5:30pm

on weekdays.

mailto:[email protected]

http://stat.duke.edu/courses/Fall12/sta101.002/

https://sakai.duke.edu/portal/site/STAT101_Fall12

https://docs.google.com/spreadsheet/ccc?key=0AtJ5X5rxFtfqdEpEaXpBNFR2SmU3VllRRktTV1Voc3c



http://www.iclicker.com/support/registeryourclicker/

mailto:[email protected]

8/28/2012

2


Grade Breakdown

Labs 5%

Clicker Questions 10%

Homework 10%

Projects (2 10%) 20%

Midterm Exams (2 15%) 30%

Final Exam 25%

Grades ≥ 90 are guaranteed at least an A- Grades ≥ 80 are guaranteed at least a B- Grades ≥ 70 are guaranteed at least a C- Grades ≥ 60 are guaranteed at least a D-


Labs

Labs are on Monday in Old Chem 01

The goal of labs is to familiarize you with statistical software, and to give you guided experience analyzing data

Labs will be group based

If you missed lab yesterday, make sure to work through it on your own by Thursday (it’s on the course website)


Clickers Clicker grade will be split equally between:

Review “Quiz” Questions: Credit only for answering correctly

Goal: motivate you to keep up with the material

Review questions will usually happen at the beginning of class – it is your responsibility to arrive on time.

New Questions: Credit simply for clicking in

Goal: motivate you to think actively about new material as it is being presented


Class Year

What is your class year?

(a) First-year

(b) Sophomore

(c) Junior

(d) Senior


Major

Your primary major (or potential future major) best falls under the category…

(a) Natural Sciences

(b) Arts and Humanities

(c) Social Sciences

(d) Math/Statistics/CS

(e) Other


Homework Weekly homework due, usually on Tuesdays

Point of homework: to LEARN! to make sure you are keeping up with the material to prepare you for projects and exams

Graded problems and practice problems

Grading Graded on a 10 point scale Lowest homework grade dropped Penalties for late homework

8/28/2012

3


Projects

Project 1 individual

confidence intervals, hypothesis tests

written report up to 5 pages in length

Project 2 with your lab group

regression

10 minute presentation

written report up to 10 pages in length


Exams

Midterm Exams Thursday, October 11, in class Thursday, November 15, in class

Final Tuesday, December 11th, 2 – 5 pm

Exams are mandatory and cannot be made-up


Keys to Success

Come to class

Come to lab

Do the homework

Read the textbook

Do lots of practice problems

Stay on top of the material


8/28/12

Introduction to Data

SECTION 1.1 • Data • Cases and variables • Categorical and quantitative variables • Using data to answer a question


Why Statistics?

Statistics is all about DATA

Collecting DATA

Describing DATA – summarizing, visualizing

Analyzing DATA

Data are everywhere! Regardless of your field, interests, lifestyle, etc., you will almost definitely have to make decisions based on data, or evaluate decisions someone else has made based on data


Data

Data are a set of measurements taken on a set of individual units

Usually data is stored and presented in a dataset, comprised of variables measured on cases

8/28/2012

4


Cases and Variables

We obtain information about cases or units.

A variable is any characteristic that is recorded for each case.

Generally each case makes up a row in a dataset, and each variable makes up a column


Countries of the World

Country Land Area Population Rural Health Internet

Birth Rate

Life Expectancy HIV

Afghanistan 652230 29021099 76 3.7 1.7 46.5 43.9

Albania 27400 3143291 53.3 8.2 23.9 14.6 76.6

Algeria 2381740 34373426 34.8 10.6 10.2 20.8 72.4 0.1 American Samoa 200 66107 7.7

Andorra 470 83810 11.1 21.3 70.5 10.4

Angola 1246700 18020668 43.3 6.8 3.1 42.9 47 2

Antigua and Barbuda 440 86634 69.5 11 75

Argentina 2736690 39882980 8 13.7 28.1 17.3 75.3 0.5


Intro Statistics Survey Data


Diet Coke and Calcium Drink Calcium Excreted

Diet cola 50

Diet cola 62

Diet cola 48

Diet cola 55

Diet cola 58

Diet cola 61

Diet cola 58

Diet cola 56

Water 48

Water 46

Water 54

Water 45

Water 53

Water 46

Water 53

Water 48


Data US News and World Report National University Rankings

Stock Market

Duke Basketball

Obama versus Romney

Unemployment Rate

Hybrid Cars

Antidepressants and Alzheimer’s


Data Applicable to You

Think of a potential dataset (it doesn’t have to actually exist) that you would be interested in analyzing

What are the cases?

What are the variables?

What interesting questions could it help you answer?

http://colleges.usnews.rankingsandreviews.com/best-colleges/rankings/national-universities/data

http://www.google.com/finance

http://goduke.statsgeek.com/basketball-m/seasons/season-stats.php?season=2011-12

http://fivethirtyeight.blogs.nytimes.com/

http://www.allabouthybridcars.com/comparison-chart.htm

http://www.google.com/publicdata?ds=usunemployment

http://www.google.com/publicdata?ds=usunemployment

http://www.hybridcars.com/top-hybrid-cars-list

http://www.hybridcars.com/top-hybrid-cars-list

http://www.google.com/finance

http://www.sciencenews.org/view/generic/id/333548/title/Antidepressants_show_signs_of_countering_Alzheimer%E2%80%99s

http://hurricanes.nhl.com/club/stats.htm?gameType=2&season=20102011&srt=p



8/28/2012

5


Kidney Cancer

Source: Gelman et. al. Bayesian Data Anaylsis, CRC Press, 2004.

Counties with the highest kidney cancer death rates


Kidney Cancer

Counties with the lowest kidney cancer death rates

Source: Gelman et. al. Bayesian Data Anaylsis, CRC Press, 2004.


Kidney Cancer

If the values in the kidney cancer dataset are rates of kidney cancer deaths, then what are the cases?

(a) The people living in the US

(b) The counties of the US

A person either has kidney cancer or doesn’t… a rate must apply to a group of people, such as a county


Kidney Cancer

If the values in the kidney cancer dataset are yes/no, then what are the cases?

(a) The people living in the US

(b) The counties of the US

A person either has kidney cancer or doesn’t. Yes/no doesn’t make sense for a county.


Categorical versus Quantitative

• A categorical variable divides the cases into groups

• A quantitative variable measures a numerical quantity for each case

Variables are classified as either categorical or quantitative:


Categorical Quantitative

8/28/2012

6


Kidney Cancer

If the cases in the kidney cancer dataset are counties, then the measured variable is…

(a) Categorical

(b) Quantitative

Rates are numbers (quantitative).


Kidney Cancer

If the cases in the kidney cancer dataset are people, then the measured variable is…

(a) Categorical

(b) Quantitative

Either having kidney cancer or not is categorical.


Variables

For each of the following situations: What are the variables? Is each variable categorical or quantitative?

1. Can eating a yogurt a day cause you to lose weight?

2. Do males find females more attractive if they wear red?

3. Does louder music cause people to drink more beer?

4. Are lions more likely to attack after a full moon?

(the answer to all of these questions is yes!)


Let’s Collect Some Data!

QUESTION: If you are romantically interested in someone, should you be obvious about it, or should you play hard to get?

Using Data to Answer a Question


Romance

What type of person are you generally more romantically interested in?

(a) Someone who is obviously into you

(b) Someone who plays heard to get


Romance

MALES ONLY: What type of person are you generally more romantically interested in?



8/28/2012

7


Romance

FEMALES ONLY: What type of person are you generally more romantically interested in?




One or Two Variables

Sometimes we are interested in one variable, as in whether people prefer obvious romantic interest or hard to get

Other times we are interested in the relationship between two variables, such as

1) prefer obvious interest or hard to get?

2) gender


What do you want to know?

We’ll do a class survey, collecting data you are interested in.

What do you want to know about your peers?

Is this a question about one variable or two variables?

What are the variables?

Are they categorical or quantitative?


What do you want to know?

Write a question to measure each variable of interest. Write questions so the resulting data will be accurate and easy to analyze.

Quantitative variable? Give units.

Categorical variable? Give the possible categories (no more than 5).

Be clear and specific.


Summary

Data are everywhere, and pertain to a wide variety of topics

A dataset is usually comprised of variables measured on cases

Variables are either categorical or quantitative

Data can be used to provide information about essentially anything we are interested in and want to collect data on!


To Do

Read Section 1.1

If you haven’t already…

Get the textbook

Get a clicker and register it

Do Lab 0 by Thursday

Documents

Course Website STAT 101: Data Analysis and Statistical ... · Data Analysis and Statistical Inference ... Lecture slides will be posted on the course website ... Project 1 individual