# SCS: Seminars 2008-09

### From MathWiki

- For the list of books that were in contention see SCS Reads 2008

- For links related to the book we selected, see Gelman & Hill (2007)

- R: Gelman & Hill makes intensive use of R (
*http://cran.r-project.org/*). There is a wiki page on getting started with R.

Table of contents |

[edit]

## **Week 1**: Gelman & Hill Chapters 1 and 2

- Friday, September 19, 2008 at 2:30 pm

[edit]

### Data sets

- coop (
*http://www.stat.columbia.edu/~gelman/arm/examples/coop*): tallies from an election for the board of a cooperative housing organization, used in Chapter 2 for testing the hypothesis of underdispersion in count data

- death.polls (
*http://www.stat.columbia.edu/~gelman/arm/examples/death.polls*): time series of Gallup polls on the death penalty, used in Chapter 2 to show simple confidence intervals

[edit]

## **Week 2**: Gelman & Hill Chapters 3, 4, 5

- Friday, October 3, 2008 at 2:30 pm

[edit]

### Data sets

- beauty (
*http://www.stat.columbia.edu/~gelman/arm/examples/beauty*): Daniel Hamermesh's data on beauty and teaching evaluations, used in regression exercises in Chapters 3 and 4

- child.iq (
*http://www.stat.columbia.edu/~gelman/arm/examples/child.iq*): study of children's IQ's used in Chapter 3 to illustrate linear regression

- earnings (
*http://www.stat.columbia.edu/~gelman/arm/examples/earnings*): data on height, weight, earnings, and other variables from a national survey (Chapters 4 and 13)

- girls (
*http://www.stat.columbia.edu/~gelman/arm/examples/girls*): the proportion of girl births in Vienna for 24 consecutive months, from Mises (1953), used in an exercise in Chapter 2 for testing the binomial model

- pyth (
*http://www.stat.columbia.edu/~gelman/arm/examples/pyth*): simple dataset for a linear regression exercise in Chapter 3

- arsenic (
*http://www.stat.columbia.edu/~gelman/arm/examples/arsenic*): data on switching drinking water wells in Bangladesh, used in Chapter 5 to illustrate logistic regression

- nes (
*http://www.stat.columbia.edu/~gelman/arm/examples/nes*): data from the National Election Study, used in Chapters 4 and 5 to show a time series of linear and logistic regressions and in an exercise in Chapter 5 to illustrate nonidentifiability in logistic regression

- pollution (
*http://www.stat.columbia.edu/~gelman/arm/examples/pollution*): old dataset of mortality and pollution in 60 U.S. cities, used in an exercise in Chapter 4 to illustrate transformations in linear regression

[edit]

## **Week 3**: Gelman & Hill Chapters 5,6

- Friday, October 17, 2008 at 2:30 pm

[edit]

## **Week 4**: Gelman & Hill Chapters 7,8

- Friday, October 31, 2008 at 2:30 pm

[edit]

## **Week 5**: Gelman & Hill Chapters 8, 9

- Friday, February 13, 2009 at 2:30 pm

[edit]

## **Week 6**:Gelman & Hill Chapters 9 and 10

- Friday, February 27, 2009 at 2:30 pm

Hugh McCague to give a presentation on circular statistics

[edit]

## **Week 7**:Gelman & Hill Chapters 10 and 11

- Friday, March 13

[edit]

## **Week 8** Lord's Paradox: Gain scores or ANCOVA

- Friday, April 3 (MOVED -- no longer March 27)

Readings on Lord's paradox previously circulated via email:

- Lord, F. (1967). A Paradox in the Interpretation of Group Comparisons.
*Psychological Bulletin, 68,*304-305. (*http://csaweb109v.csa.com.ezproxy.library.yorku.ca/ids70/view_record.php?id=1&recnum=5&log=from_toc&SID=4qd8tsaa4ughl8tjd06m7fbq17&mark_id=cache:0,0,13*) - Lord, F. (1969). Statistical Adjustments When Comparing Preexisting Groups.
*Psychological Bulletin, 72,*336-337. (*http://csaweb109v.csa.com.ezproxy.library.yorku.ca/ids70/view_record.php?id=5&recnum=8&log=from_toc&SID=4qd8tsaa4ughl8tjd06m7fbq17&mark_id=cache:2,0,12*) - Daniel B. Wright, D. B. (2006). Comparing groups in a before–after design: When t test and ANCOVA produce different results,
*British Journal of Educational Psychology, 76,*663-675. (*http://www.ingentaconnect.com.ezproxy.library.yorku.ca/content/bpsoc/bjep/2006/00000076/00000003/art00014*) - Friedman, L. and W. Wall. (2005). Graphical Views of Suppression and Multicollinearity in Multiple Linear Regression.
*The American Statistician, 59,*127-136. (*http://proquest.umi.com.ezproxy.library.yorku.ca/pqdweb?index=7&did=831692351&SrchMode=3&sid=1&Fmt=6&VInst=PROD&VType=PQD&RQT=309&VName=PQD&TS=1237789833&clientId=5220&aid=1*) - "Emerging+Themes+in+Epidemiology":And:LQE%3D(DA,None,8)20080122:And:LQE%3D(VO,None,1)5$&sgHitCountType=None&inPS=true&sort=DateDescend&searchType=PublicationSearchForm&tabID=T002&prodId=AONE&searchId=R1¤tPosition=1&userGroupName=yorku_main&docId=A175408362&docType=IAC Yu-Kang Tu, Y.-K., D. Gunnell and M. S. Gilthorpe. (2008). Simpson's Paradox, Lord's Paradox, and Suppression Effects are the same phenomenon – the reversal paradox.
*Emerging Themes in Epidemiology, 5(2)*(*http://find.galegroup.com.ezproxy.library.yorku.ca/itx/retrieve.do?contentSet=IAC-Documents&resultListType=RESULT_LIST&qrySerId=Locale(en,US,):FQE%3D(JN,None,33)*)

[edit]

## Some references

- Rogosa, D. R., & Willett, J. B. (1983). Demonstrating the reliability of the difference score in the measurement of change.
*Journal of Educational Measurement*, 20, 335-343.

- Rogosa, D. (1988). Myths about longitudinal research. In K. W. Schaie, R. T. Campbell, W. M. Meredith, & S. C. Rawlings (Eds.),
*Methodological issues in aging research*(pp. 171-209). New York, NY: Springer Publishing Company.

- Maris, E. (1998). Covariance adjustment versus gain scores--revisited.
*Psychological Methods*,**3**, 309-327.

- Maxwell, S. E., & Delaney, H. D. (1990).
*Designing experiments and analyzing data: A model comparison approach*. Belmont, CA: Wadsworth.

[edit]

## Links

[edit]

### Multilevel models

- A First Look at Multilevel and Longitudinal Models (
*http://www.math.yorku.ca/~georges/Courses/Repeated/CourseNotes.pdf*) (userid:fisher password:cohen) - Longitudinal Data Analysis with Mixed Models: A Graphical Overview (
*http://www.math.yorku.ca/~georges/Slides/Workshop-v1-0Slides.pdf*)

[edit]

## **Week 9**: Gelman & Hill Chapters 11,12

- Friday, April 24, 2008 at 2:30 pm