ST 372          Introduction to Statistical Inference and Regression

Dr. Herle McGowan
Email: hmmcgowa@ncsu.edu
Office: 5262 SAS

Office Hours:

Syllabus

TI89 Stat List Editor Manual.pdf


Project Materials

Assignment

Proposal

Protocol Guidelines

How To Write About Randomization

Peer review guidelines

Report Guidelines

Very good—not perfect—example report from previous class.

 

Day

Week

Date

DOW

The Plan

What we did, HW assigned

Anything happening today or due today that was pre-announced?

1

1

Aug 18

H

Experiment Design Learning Outcomes

1.       Went over syllabus (link above)

2.       Took attendance and had you give me your calculator type and your major

3.       Told you what to do if you were on the waitlist or if you have not taken ST371 as your prereq

4.       Introduction to ST372 Notes with blanks

4.1.    No blanks version

5.       HW:

5.1.    Fill in Table of Analogies (remember your prerequisite material)

5.2.    Get TI 83, 84, 89, etc.

5.3.    Go get JMP statistical software on your own computer (free to NCSU students):

 http:/www.ncsu.edu/software/available-software/product.php?software=209

 

2

1

Aug 23

T

1.       Met with your groups for the first time—name, major, embarrassing.

2.       Table of Analogies

2.1.    solutions handed out

2.2.    Taken up for grade (effort)

3.       Announced exploratory statistics with calculator and JMP learning outcomes (what you should be able to do)

3.1.    Exploratory Statistics Calcultor Learning Outcomes

3.2.    Exploratory Statistics JMP Learning Outcomes

4.       HW:

4.1.    Learn how to use your calculator using the following files—you will have a calculator quiz on Aug 30, and you will be given a limited amount of time to answer questions, so I will assume that you have practiced enough to be efficient.

4.1.1. Exploratory Stats 83 84 how to   Key—compare your answers with your group members or other classmates.

4.1.2. Exploratory Stats 89 how to Key—compare your answers with your group members or other classmates.

4.2.    JMP HW due Tues Aug 30

4.2.1. ExploratoryStatisticsJMPHowTo

5.       Experiment Design Notes 

5.1.     ExperDesNotes1IntroThruExperErrorBlanks

5.2.    ExperDesNotes1IntroThruExperErrorNoBlanks

6.       In-class exercise identifying parts of experiment—handed in at end of class

Table of Analogies

3

2

Aug 25

H

1.       New room again.

2.       Handed back Table of Analogies

3.       Group member quiz

4.       More on experimental units: Nascar vs. Detroit  

4.1.    Blanks

4.2.    No blanks

7.       Continued Experiment Design Notes Part 1

Group Member Quiz (names, majors, embarrassing)

4

2

Aug 30

T

1.       Seating chart for new room (to facilitate paper handing in/out and working in groups)

2.       Turned in JMP HW Key

3.       Took calc quiz

4.       Handed back Group Member Quiz

5.       Experiment Design Notes Part 2—but you still need Part 1 for the examples!

5.1.    Blanks

5.2.    No blanks

1.       JMP HW due today Key to HW

2.       Calculator quiz today—what fun!

5

3

Sept 1

H

1.       Assigned Projects

2.       Gave Proposal template

3.       Two sample proposals

3.1.    Bleach Pen---Nail comments on it

3.2.    Bending Metal—Nail comments on it

6

3

Sept 6

Sampling Distributions  Learning Outcomes  Final

1.       Transition from Experiment Design to Sampling Distributions (Devore Section 5.3)

1.1.    HW: In-Class Assignment REvisited

2.       Started Section 5.3: Sampling Distributions Notes  (this is Section 5.3 in Devore text).

3.       In Class Assignment on Widget problem

4.       Class time to discuss projects with groups—last 15 min

7

4

Sept 8

H

1.       Review solutions to In-Class Assignment Revisited (Oops, I forgot to take it up.)

2.       Sampling Dist notes continued from last time (this is Section 5.3 in Devore text).

3.       HW:

3.1.    For ,

3.1.1. Graph the PMF under the PMF of

3.1.2. Find

3.1.3. Find  --write this underneath where you wrote out , and compare the values of the deviations and their weights for the two random variables.

3.2.    For ,

3.2.1. Use a table to find the PMF like we did for, but don’t write out x1, x2, x3, x4, unless you are masochistic, but instead try writing  as a function of  , and then write out combinations of two values that  could take.

3.2.2. Put the PMF of  in a table underneath the tables for the PMF’s for  and .

3.2.3. Graph the PMF of  underneath the graphs for the PMF’s for  and .

3.2.4. Find

3.2.5. Find --write this underneath where you wrote out  and , and then compare the values of the deviations and their weights again.

3.3.    Think: what is the point of doing this? Can you answer this question? This is a learning demonstration. What do I want you to learn?

4.       Class time to discuss projects with groups—last 15 min

In-Class Assignment REvisited

8

4

Sept 13

T

1.       Go over HW from last Thursday.

2.       Finished Sampling Dist notes Part 1—the handout (see above). Note this was essentially Section 5.3 from Devore text.

3.       Did In-Class Assignment: Sampling Distributions Assignment 3 and turned it in. It went along with the notes and homework.

4.       Assigned Sampling Distributions Assignment 2, due Thursday. Represents HW from Section 5.3.

5.       Sampling Distributions Notes Section 5.5: Linear Combinations of Random Variables

5.1.    Copied notes from the board (yes, really!)

6.       Class time to discuss projects with groups—last 15 min

1.       HW from Thursday

2.       Project Proposal due by 11:59pm tonight

Email nail@stat.ncsu.edu, subject: GX: Meaningful Title of Proposal

9

5

Sept 15

H

1.       Passed out solutions to HW and discussed—at length! No, I don’t have a link to this because I basically did it by hand and photocopied it, and I didn’t feel like scanning it. So get it from a friend, or ask me later if you missed class.

2.       Came up with some nice sample test questions.

3.       Continued notes on Section 5.5: linear combinations of random variables

4.       Talked about chicken poop.

5.       Assigned HW due Tues:  Section 5.5 HW: 58, 59, 60, 64, 72—lots o’ work!!

SamplingDistributionsAssignment2 (This represents HW from Section 5.3 in Devore text).

10

5

Sept 20

T

1.       Went over HW at length—which means test questions will look like these.

1.1.    Discussed how to do normal probability calculations with TI83, 84, 89.

1.2.    Discussed that you have to remember how to do Expected Value calculations and variance calculations from your prerequisite course—this is expected.

1.3.    It’s OK to do integration on your TI89.

2.       Told stories (believe it or not, I always tell these stories when I teach—it’s part of the plan):

2.1.    The Story of Laura

2.2.    The Story of Michael

3.       Started the theorems in Section 5.5.

Section 5.5 HW: 58, 59, 60, 64, 72—lots o’ work

11

6

Sept 22

H

1.       Discussed change of schedule for detailed protocol and peer review—do not be absent the Tuesday before fall break!

2.       Gave you an extra credit opportunity—5 points on final exam—if you read the book by Bill Walsh: The Score Takes Care of Itself. Sometime over the course of the semester. I’ll ask some open-ended question on the final that you will answer to convince me that you read it.

3.       Continued theorems about sampling distributions of statistics—culminating in the Central Limit Theorem—did I mention how important this theorem is?

4.       How large must n be? Discussed with handouts.

5.       Assigned HW: Examples 5.24, 5.25, 5.26 from the handout (my questions are different from the ones in the book, though you can use those examples for help). Plus problems 46, 48, 50, 52.

Section 5.5 HW will be taken up for a grade—correctness!! I’ll randomly choose one problem to grade

Nick owes me a TI89 cable.

12

6

Sept 27

T

1.       Handed out HW solutions.

2.       Answered questions about HW.

3.       Discussed Test 1

3.1.    some reminders on Exp Design,

3.2.     how to study

3.3.    Look at learning outcomes for Exp Design and Samp Dists (links in the column to the left of this column)

3.4.    Make sure you can work HW problems—that’s what test problems look like

3.5.    I will supply Z-table

3.6.    If I want you to work with a funky distribution—other than Normal or Binomial—I’ll give you the PMF, expected value, and variance if you need them to do the problem.

4.       Discussed timeline for upcoming class days

4.1.    Test Thurs (I won’t be here)

4.2.    Work on protocols with groups Tues ( I won’t be here, Ms. Rice will take attendance.)

4.2.1. I will have graded your proposals by then.

4.2.2. I will somehow supply you with some guidelines for the protocol by then.

5.       Discussed how the normal approximation to the binomial distribution is an application of the CLT—did I mention how important that theorem is?

5.1.    Binomial Approx to Normal via CLT Blanks

5.2.    No Blanks

6.       Awkwardly proceeded with bead sampling activity—I need to figure out the logistics on that for a class this large better….Hmm….

Section 5.4 HW: Problems 46, 48, 50, 52, and

Examples 5.24, 5.25, 5.26 from the handout

13

7

Sept 29

H

Test 1:

You can use the following:

1.       Calculator

2.       Table of Analogies

3.       Cheat sheet: both sides of 8.5 by 11 piece of paper

I’ll supply:

1.       Z-table

2.       Info on funky distributions if needed. That means distributions other than Normal and Binomial.

Test 1:

Experiment Design

 Sampling Distributions

 

 

14

7

Oct 4

T

Do not miss class today! You will lose 2 points on your project grade!

Today in class, you will spend the entire class working with your groups on your detailed protocol. Somebody bring a laptop.

1.       Turn off all cell phones, pagers, ipods, etc.

2.       Focus 100% on working with your group to make as much progress as possible on your protocol.

3.       You will not get tests back today.

 

Protocol Guidelines

How To Write About Randomization

 

Oct 6

H

No Class Fall Break!

No Class: Fall Break!

15

8

Oct 11

T

Learning Outcomes CI and PI Sigma known

1.       Went over tests

1.1.    Test 1 Solutions Out of Order

1.2.    Blank Test 1 for exam prep

Get graded Test 1 back today.

16

8

Oct 13

H

1.       Exchanged protocols with other groups

2.       Peer review guidelines discussed.

3.       Handed out z-table—bring this to class every day!

4.       Notes: Transition from Sampling Distributions to Confidence Intervals—part typed, part hand-written. Stopped at bulls-eyes.

5.       HW: do peer review

Detailed protocol due.

 

Last day to drop a course without a grade.

17

9

Oct 18

T

1.       Notes:  Inference with Sigma known: confidence intervals (also point estimates and estimators, predictions, and predictors).

2.       Ended by drawing four bulls-eyes.

 

Peer review of another group’s protocol due.

18

9

Oct 20

H

1.       In class assignment on point estimators and estimates, predictors and predictions.

2.       Wrote solutions on board.

3.       Handwritten notes on confidence intervals for sigma known.

19

10

Oct 25

T

1.       HW due Tues Nov 1

2.       More handwritten notes.

2.1.    Derived the formula for a C% confidence interval for mu, with sigma known.

2.2.    Wrote down the formula for a C% confidence lower/upper bound for mu, with sigma known.

2.3.    Discussed why you might want a lower/upper bound instead of a symmetric interval.

20

10

Oct 27

H

1.       CI’s on calculator handout.

2.       Prediction Intervals handout.

2.1.    Blanks

2.2.    No blanks

3.       Summary of CI and PI for Sigma known

3.1.    Word version

21

11

Nov 1

T

Learning Outcomes Hypoth Tests Sigma known

1.       Took up HW--solutions

2.       Finished Prediction Interval notes (see last class for link)

3.       Notes: Intro To Hypothesis Tests Sigma known

3.1.    Blanks

3.2.    No Blanks

HW due on CI and PI with Sigma known

Solutions posted in column to left.

22

11

Nov 3

H

1.       Finished Intro To Hypoth Test Notes

2.       Started Notes 2 Hypoth Tests Sigma Known

2.1.    Blanks

2.2.    No Blanks

3.       We got through the end of page 2 of the notes.

4.       You should have tried to do page 3 on your own.

23

12

Nov 8

T

1.       Discussed data submission guidelines

2.       Discussed test coverage (see Nov 10 list  below)

3.       Finish Notes 2 Hypoth Tests Sigma known

4.       Handed out summary of hypoth test formulas

4.1.    Pdf version

4.2.    Word version

5.       Do Hypothesis test problems sigma known in class.

5.1.    Solutions

Project data due. Submit by 11:59 pm.

Data submission guidelines

1.       Email

1.1.    me and the TA

1.2.    CC all group members

1.3.    Subject “Group XX Data: Name of Project”

1.4.    Attach: excel spreadsheet

1.4.1. Name of file: “Group XX Data: Name of Project.xlsx”

1.4.2.  With data following the example below.

2.       Data format example

24

12

Nov 10

H

Test 2:

1.       I will provide Z-table

2.       Formula sheets:

2.1.    Table of Analogies

2.2.    8.5 by 11 piece of paper, both sides

2.3.    You’ll turn these in with your test, but you’ll get them back.

3.       Be able to do problems like the true/false problems on Test 1 (even if they aren’t true false again). (From Sampling Distribution LO’s)

4.       Point estimates

4.1.    From the notes on October 18 and the In Class Assignment on October 20.

4.2.    Be able to answer questions like those in the In Class Assignment.

5.       CI and PI for sigma known—see LO’s posted above

6.       Hypoth tests for sigma known—see LO’s and Problems posted above.

Test 2

Solutions

Distribution

25

13

Nov 15

T

ANOVA

1.       Discuss project timeline and grading

2.       Introduction to two-factor ANOVA via Wood-joint-type-tensile strength experiment

2.1.    Notes on Notation

2.2.     Table with data

2.2.1. New notation for random variable

2.2.2. Notation for and how to calculate treatment means

2.2.3. Notation for and how to calculate marginal means

2.2.4. Ditto grand mean

2.2.5. Ditto effect of Factor A at level i

2.2.6. Ditto effect of Factor B at level j

2.3.    Boxplots for each factor

2.4.    Define additive model

2.5.    Define and calculate (put in a table) effect of interaction between Factor A at level i and factor B at level j.

2.6.    Define full model

2.7.    Summary of formulas and notation

3.       Gave back tests

 

Homework:

4.       For your project data, make a table of estimates of treatment means, marginal means, the grand mean, and the effects of factors A and B as we did in class

4.1.     by hand (or coding it yourself in excel or some other software tool)—and no, I’ll never make you do this again

4.2.    using a built-in feature in excel.

5.       Make all the plots we made in class using software of your choice (boxplots for now)

27

14

Nov 22

T

1.       Pick up where we left off discussing interactions.

2.       Make an ANOVA table for the wood-joint-type-tensile strength example.  Blank notes.  Filled in notes.

2.1.    Why?

2.2.    SST

2.3.    Decompose SST—formulas and interpretations of pieces

2.4.    Make table for full model: df, ss, ms, f, p-vlaue

2.5.    Create additive model anova table from full model anova table.


Homework:

3.    For your project data, use StatCrunch (or Excel or another statistical software) to create the ANOVA table for the full model.  Bring this to class on Tuesday.

Unstacked Wood-Joint Data in StatCrunch

How to do ANOVA table in StatCrunch -- Updated 11/29

Unstacked Wood-Joint Data  in Excel

How to do ANOVA table in Excel (old)

Nov 24

H

No Class, Thanksgiving!

No Class, Thanksgiving!

28

14

Nov 29

T

 1. Go over Pencil brand/soak time example and discuss how to evaluate the ANOVA model.  Blank Notes.  Filled in notes.
2. Hypothesis Testing for ANOVA.  Blank notes.  Filled in notes.

 

HW: Create the ANOVA table for the full model using your project data.

29

Dec 1

H

Last day of class

1.  Regression example.  Blank copy.  Filled in copy.  

            1.1. Lecture material for regression: http://www.stat.ncsu.edu/people/woodard/courses/ST372/

2.  Passed out overview of topics for the final exam.

            2.1. I added a little more detail about confidence/prediction intervals based on questions from class.

            2.2. Notes on each of these topics are already posted on the course website.

            2.3. I will post practice problems so that you can see how I might ask questions about these topics.

Project report due--Extended to December 5 at NOON

Report Guidelines

Good example report from a previous class.

Last day of class

Dec 15

H

Final Exam 8-11am, same classroom

Practice problems on sampling distribution, confidence intervals, and hypothesis tests for a mean (sigma known):  Blank.  Completed.

Practice problems on experiments, 2-factor ANOVA, and regression:  Blank.  Completed.

Practice problems--multiple choice format: Blank. Completed.

Final Exam 8-11am, same classroom