EDUR 8132 Fall 2010 Instructor: Bryan W. Griffin (bwgriffin@GeorgiaSouthern.edu) 
Announcements 
Dec 1  Several Updates:
Nov 30  Added notes on Standardized Coefficients, see Lectures notes 9.i.
Nov 20  A number of updates:
Session 15 Nov 30 (Last Chat)  will focus on Standardized Regression Coefficients. I will create a summary set of notes on these, but in the meantime read Agresti text on standardized regression coefficients pages 270, 351354, 529532. Also, see my discussion in Notes 8a Regression with One Quantitative Predictor (see page 10) and Notes 8b Regression with Two Quantitative Predictors (also starting on page 10).
PDF Files
Submissions in this course must be as a PDF attachment. Below are links to creating PDF files:
Here are a few free web page that convert files to PDF over the internet:
 http://convert.neevia.com/
 http://www.gohtm.com/
 http://www.pdfonline.com/
 http://www.pdfdownload.org/
Links to free software to create PDF files (it leaves no watermark):
http://www.primopdf.com/
http://www.nitropdf.com/free/index.htmI also use OpenOffice to create some free PDF files. Open Office is a free Office Suite similar to Microsoft Office (Open Office leaves no watermark): http://www.openoffice.org/
If you want further tips and links for converting to free PDFs, read this site http://www.pruittfamily.com/paul/freepdf.htm or this one http://www.masternewmedia.org/2002/03/31/creating_pdfs_without_adobe_acrobat_part_ii.htm
Course Index (e.g., tests, activities, reading and other material to review for each class session)
 Regression
 Regression: One Qualitative Predictor with Two Categories
 One Qualitative Predictor with Three or more Categories
 Multiple Qualitative Predictors
 Multiple Quantitative Predictors
 Both Qualitative and Quantitative Predictors
 ANOVA
 See course notes below for exercises embedded within notes.
 Others may be added here if needed.
 Introductory notes (modeling behavior, central tendency, variability, correlation, ttest, hypothesis testing)
 Simple Regression with One Quantitative Predictor
 Summary notes for chat: Notes 8a Simple Regression
 Detailed notes: Notes 8a Regression with One Quantitative Predictor
 Model Fit: R^{2}, SEE and MSE, and Adjusted R^{2}
 Table of Critical tratios: Critical t
 Multiple Regression with Multiple Quantitative Predictors
 Summary notes for chat: Notes 8b Multiple Regression
 Detailed notes: Notes 8b Regression with Two Quantitative Predictors (Minor update to Notes 8b: the value of ΔR^{2} is actually the squared semipartial correlation)
 Reading Published Research Results:
 Regression Coefficient Interpretation
 Semipartial Correlation (ΔR^{2})
 Summary notes for chat: Notes 8c Regression Semipartial Correlation
 Detailed notes: Read Notes 8b page 5 section entitled "ΔR^{2}, Semipartial Correlation, and the Partial F Test of ΔR^{2}"
 Video: ΔR^{2} Discussed and Illustrated in SPSS; ΔR^{2} Notes used in video; Test Score Data (in SPSS format) used in video
 (Note Fall 2010: SPSS syntax offers easy way to obtain several ΔR2 values and F ratios simultaneously: change Method=enter x1 x2 x3 x4 to Method=test (x1) (x2 x3) (x4). For Video data above, this command change produces summary output: "/METHOD=ENTER Study_Time IQ Griffin Moore" to this "/METHOD=TEST (Study_Time) (IQ) (Griffin Moore)" Add video segment to video above demonstrating this.
 Table of Critical Fratios: Critical F at .05 and .01
 Regression with One Qualitative Predictor
 Summary notes for chat: One qualitative dichotomous predictor (2 categories)
 Summary notes for chat: One qualitative predictors with 3+ categories
 Detailed notes: Notes 8d Regression with One Qualitative Predictor
 Detailed notes: Notes 8e Regression Multiple Comparisons
 Table of Bonferroni Critical tratios: Dunn's Critical Values for Bonferroni t
 SPSS Data: Cars.sav
 Regression with Multiple Qualitative and Quantitative Predictors
 Summary notes for chat: Notes 8f Regression with Two Qualitative Predictors Summary
 Summary notes for chat: Notes 8g Regression with Both Qualitative and Quantitative Predictors Summary
 Detailed notes: Notes 8f Regression with Two Qualitative Predictors
 Detailed notes: Notes 8g Regression with Both Qualitative and Quantitative Predictors
 Effect sizes in regression
 Sample size for regression
 Standardized Regression Equation: Notes 8h Regression Standardized Coefficients (Note  update formatting, also check standardized equation subscripts)
 ANOVA Models
 Notes 9 ANOVA
 Homework answers to problems found in Notes 9 ANOVA
 Some ANOVA Exercises with APA answers (note table missing R^{2} values, these should be added)
 Notes 9b ANOVA without Interactions
 Notes 9c ANOVA with Interactions (Video: ANOVA with Interactions in SPSS)
 Notes 9d ANCOVA
 Supplemental Reading: Stevens Chapter 7 ANCOVA
 Excel Spreadsheet for Interactions
 Video: Oneway ANCOVA with three groups
 Video: Twoway ANCOVA
 Video: Oneway ANCOVA with Factor*Covariate Interaction (Unfortunately sound spikes and breaks occasionally)
 EdS Read Aloud, Think Aloud Data Results (Chat 15 worked example of ANCOVA with interaction)
 ANCOVA Data Examples
 Sample Data 1: Fictional salary discrimination
 Sample Data 2: Gun control and homicides by state
 Sample Data 3: Fictional achievement data  homework with parental support
 Sample Data 4: Fictional data showing ANCOVA power vs ANOVA
 Sample Data 5: Car MPG by country of origin controlling for horsepower and weight
 Spreadsheet to Calculate Bonferroni and Scheffé Confidence Intervals (based on Editgrid.com)
 Same as above, except in Excel format
 Multiple Comparisons in Excel: Bonferroni and Scheffe (speadsheet may provide incorrect CIs)
 Multiple Comparisons in Zoho (online spreadsheet): Bonferroni and Scheffe (speadsheet may provide incorrect CIs)
 Multiway ANOVA
 Effect sizes in ANOVA
 Sample size for ANOVA
 Interactions in Regression
 Polynomial Regression
 Model fitting and assessment
 Common Research Designs and Related Statistical Analysis (e.g. Posttest only control, Pretestposttest control, Nonequivalent control group, etc.)
 Logistic regression or Factor Analysis
 Notes 1 Descriptive Statistics
Excel file: 1. SD, Variance, and Z scores Notes 2 Normal Distribution and Standard Scores
 Notes 3 Statistical Inference
Excel file: 2. Population, Sample, and Sampling Distributions Notes 4 Hypothesis Testing and One Sample Z Test
Excel file: 3. Coin Flip Notes 5 ttests
Excel files:
4. One Sample ttest,
5. Two Sample ttest,
6. Correlated Sample ttest ;
Critical tvalues table:
ttables.pdf Notes 6 Correlation
Excel files:
7. Correlation Notes 7 Chisquare
Excel files:
8. Goodness of fit chisquare,
9. Association chisquare
 One and Two Samples ttest (answers)
 Correlated Samples ttest (answers)
 Correlation (answers)
 Chisquare (with answers)
 ANOVA (answers note this is a Word Document)
 Regression (see regression notes)
 ANCOVA (to be added)
Topic 
Readings 
1. Hypothesis Testing Logic  Chapters 5, 6 
2. Regression  Chapter 9 
3. Multiple Regression  Chapters 10, 11 
4. ANOVA  Chapter 12 
5. ANCOVA/Regression  Chapter 13 
6. Modeling building in Regression/ANOVA  Chapter 14 
7. Logistic Regression  Chapter 16 
Course Calendar (will be revised throughout term)
Session 1 (8/17): 5pm to 7:45pm (Note: Fall 2010 covered through correlation review; begin hypothesis testing Session 2)
 Syllabus
 Review of modeling behavior/outcomes with statistics, descriptive statistics, ttest, and correlation (SPSS use illustrated throughout)
 Logic and practice of hypothesis testing (fairness of coin), Errors in hypothesis testing
 Statistical Inference, Hypothesis Testing (read this: logic of hypothesis testing)
 Central Limit Theorem Illustrated
 Excel file showing 95% CI and Standard Error of the Mean: http://www.bwgriffin.com/gsu/courses/edur8131/content/Example_Calculating_SE_of_Mean_adv.xls
 Some additional readings on the logic of hypothesis testing:
Session 2 (8/24) (Note: Fall 2010 covered through student ratings example; resume at Additional Examples for Interpretation p. 7)
 Logic and practice of hypothesis testing (fairness of coin), Errors in hypothesis testing
 Statistical Inference, Hypothesis Testing (read this: logic of hypothesis testing)
 Central Limit Theorem Illustrated
 Excel file showing 95% CI and Standard Error of the Mean: http://www.bwgriffin.com/gsu/courses/edur8131/content/Example_Calculating_SE_of_Mean_adv.xls
 Some additional readings on the logic of hypothesis testing:
 Linear Regression (one quantitative predictor)
Session 3 (8/31) (Note: Fall 2010 covered through notes 8a.)
 Linear Regression  resume at Additional Examples for Interpretation; add reading simple regression tables in literature
Session 4 (9/7) (Note: Fall 2010 covered summary notes 8b but did not illustrate APA presentation of final example)
 Multiple Regression (two or more quantitative predictors)
Session 5 (9/14)
 Review Multiple Regression (two or more quantitative predictors) and work through example to illustrate APA styled presentation
 Confidence intervals  calculating 99% by hand and SPSS
 Review Model Fit Summary
 Begin Semipartial Correlation (ΔR^{2})
Session 6 (9/21)
 Finish Semipartial Correlation (ΔR^{2}) if needed
 Begin Notes 8d Regression with One Qualitative Predictor
Session 7 (9/28) (Note: Fall 2010 covered 8d One Qualitative Predictor)
 Notes 8d Regression with One Qualitative Predictor if needed
 Begin Notes 8e Regression Multiple Comparisons
 Possibly begin Notes 8f Regression with Two Qualitative Predictors
Session 8 (10/5) (Note: Fall 2010 completed one qual. predictor, briefly discussed interpretation with two qual IVs, began multiple comparisons)
 Resume Notes 8d Regression with One Qualitative Predictor at "The Regression Equation for One Categorical IV" (p. 13)
 Begin Notes 8e Regression Multiple Comparisons
 Possibly begin Notes 8f Regression with Two Qualitative Predictors
Session 9 (10/12) (Note: Fall 2010 covered notes on multiple comparisons)
 Resume notes Notes 8e Regression Multiple Comparisons
 Begin Notes 8f Regression with Two Qualitative Predictors
 Begin Notes 8g Regression with Both Qualitative and Quantitative Predictors
Session 10 (10/19) (Note: Fall 2010 covered notes on multiple comparisons and two qual IVs.)
 Resumed multiple comparisons
 Covered Notes 8f Regression with Qual IVs Qualitative IVs (Added Examples with two IVs with 3+ categories; reference use with ANOVA easier at this point)
 Possibly begin Notes 8g Regression with both Qual and Quan Predictors.
Test 2 likely sometime after 10/19 or 10/26
Session 11 (10/26)
 Begin Notes 8g Regression with both Qual and Quan Predictors.
 Possibly begin notes on ANOVA.
Test 2 after 10/26
Session 12 (11/2)
 Resume, if needed, Notes 8g Regression with both Qual and Quan Predictors
 Begin notes on ANOVA.
Session 13 (11/9)
 ANOVA
 Multiway ANOVA (plus interactions)
Session 14 (11/16)
 Multiway ANOVA (plus interactions)
 ANCOVA
Thanksgiving 11/23
Session 15 (11/30)
 ANCOVA
 Standardized regression coefficients
Session 16 (12/7) Test 3 due
NOTE  Material below this point is not yet organized into sessions
Course Assessment Questionnaire
General Readings on ANOVA:
ANCOVA
Note to self  add links to online statistical programs, e.g.
http://home.ubalt.edu/ntsbarsh/Businessstat/otherapplets/SampleSize.htm#rmenu
http://noppa5.pc.helsinki.fi/koe/flash/flash.html
Office Information
Hours
By chance or appointment. Best to contact me via email to arrange an appointment.
Telephone Numbers
Office (Room 2128 College of Education Building): 9124780488
Department of Curriculum, Foundations, and Research: 9124785091Use GeorgiaView to contact me. If GeorgiaView is not working, my regular email address is bwgriffin@GeorgiaSouthern.edu, but please use GeorgiaView for courserelated communications.
Department of Curriculum, Foundations, and Reading
P.O. Box 8144
College of Education
Georgia Southern University
Statesboro, GA 30460
Course Description
This course is will provide cover modeling relations among variables with a
focus on regression and various analysis of variance models. If time permits,
other modeling methods will be covered such as logistic regression or factor
analysis.
Texts
Required  Agresti, A. & Finlay, B. (2009). Statistical methods for the social sciences (4th ed.). Pearson/Prentic Hall. (Note that 3rd edition will also suffice for this course and may be cheaper if found used online)
Web Pages
http://www.bwgriffin.com/gsu/courses/edur8132/
Software (Required)
SPSS (PASW) Version 10.0 or higher. Version 18 of the Graduate Pack Base can be rented for 6 months for $35 here IBM/SPSS Statistics Standard GradPack 18 for Windows can be rented for 6 months here for $85:
Other student options and pricing for SPSS/PASW can be found here (the Grad Pack can be bought for $200 and the license is for 4 years, e.g., www.journeyed.com ):
http://www.spss.com/vertical_markets/education/online.htm
Course Activities
To facilitate learning of statistical analysis of data and reporting of
statistical results, numerous computer replications of examples from readings
and outofclass exercises will be provided.
Content Covered
Examinations
There will be three tests administered during the term. Each test will focus
on conceptual components of statistical analysis, computer applications, choice
of statistical procedures, and written results and interpretations.
Grading
Each test will be weighted equally at 1/3 of the final grade. Final grades
will be assigned based on the following table:

90
and above 
=
A 


80
to less than 89.999 
=
B 


70
to less than 79.999 
=
C 


60
to less than 69.999 
=
D 


59.999
and below 
=
F 

You will be allowed to take any missed test for any absence (no excuse is
necessary). Should you not provide responses to a missed test before the end of
the term, an I (incomplete) will be assigned as your final course grade and
will remain until all missed tests are completed, or for one year and then
become an F.
If you fail to take the final test on the scheduled date, you may take it
during a time that is convenient for the instructor. Tests cannot be taken
early.
Attendance
You may come and go as you please.
Withdrawing from Class
The university sets a specific date in which you may withdraw from a course without an academic penalty. In this course, however, you may withdraw without an academic penalty (i.e., you will received a WP) until the last day of regular class (this excludes exam week), no questions asked, no matter what your current performance. My policy of assigning WPs is contingent upon the approval of the CFR department chair and COE Dean (i.e., a WP is not guaranteed).
To withdraw after the drop date, contact the registrar's office to learn what form is needed (it may be called "petition to withdraw" or something similar). Complete that form and submit to me so I may sign and forward to my departmental chair. It may also be possible to withdraw via emailagain, contact the registrar's office to learn if possible and how.
Academic Integrity Expectations
Students are expected to abide by the GSU student conduct code and
regulations regarding academic integrity. Academic misconduct such as cheating
and plagiarism will be reported to the Office of Judicial Affairs and
appropriate penalties imposed, such as a grade of zero on the targeted
activity. See information on student
conduct in the relevant student handbook for details.
Disability Accommodations
If a student has a documented and declared disability, reasonable
accommodations will be provided if requested by the student according to the
recommendations of the GSU Disabled Student Services office.
How This Course Supports the Colleges Conceptual Framework
The College of Educations conceptual framework advances the theme of reflective educators for diverse learners. In this course information will be learned that should make each student a more knowledgeable consumer of educational research. With this knowledge you will be better able to evaluate critically current and recommended practices when analyzed empirically.
Types of Statistical Procedures and Their Characteristics
Statistical
Test 
Independent Variable 
Dependent Variable 
Special Feature 
Example Hypothesis 
(1) Pearson's r (correlation coefficient) 
Quantitative 
Quantitative 

There is a positive relationship between intelligence and mathematics
achievement scores. 
(2) ttest for independent groups 
Qualitative 
Quantitative 
IV has only 2 categories 
There will be a difference between boys and girls on mathematics
achievement scores. 
(3) ChiSquare c2 
Qualitative 
Qualitative 
Both IV and DV may have more than two categories 
Males will be more likely to drop out of school than females. 
(4) Oneway ANOVA (Analysis of Variance) 
Qualitative 
Quantitative 
IV may have 2 or more categories. 
There will be a difference among Black, Hispanic, and White students in
mathematics achievement scores. 
(5) Multiway ANOVA (Analysis of Variance) 
Several Qualitative 
Quantitative 
IVs may have 2 or more categories. 
There will be a difference among Black, Hispanic, and White, and
between male and female, students in mathematics achievement scores. 
(6) ANCOVA (Analysis of Covariance) 
1. Qualitative 2. Covariate may be either
qual. or quan., but usually quantitative 3. May also have several
additional qualitative and quantitative variables 
Quantitative 
Qualitative IVs may have 2 or more categories. Covariates used to make
adjustments to DV means. 
There will be a difference among Black, Hispanic, and White students in
mathematics achievement scores after taking into account levels of
motivation. 
7) Regression 
Both Qualitative and Quantitative 
Quantitative 

Any of the above except for chisquare. 
(9) Logistic Regression 
Both Qualitative and Quantitative 
Qualitative 
DV is qualitative with only 2 catgories 
Any of the above, except the DV will be binary (e.g., dropout/persist;
pass/fail; favor/oppose) 
Copyright 2005, Bryan W. Griffin
Last revised on