An Introduction to Logistic Regression
Analysis and Reporting
CHAO-YING JOANNE PENG
KUK LIDA LEE
GARY M. INGERSOLL
& Kravitz, 1994; Tolman & Weisz, 1995) and in educational researchespecially in higher education (Austin, Yaf
GSR 719 Research Report Seminar, Spring, 2012
Instructor: Michael Wood 772-5572; <[email protected]>.
Ofce and hours: (T 3:30-6:30pm; Th 4:00-6:00pm) and by appointment
Wednesdays afternoons are typically concerned with
GSR719 sp2003: data dreams p.
Dr. Wood's Easy Guide to Statistical Procedures commonly used in Analysis of
Important note: this exercise assumes that you have done some conceptual work,
including at least some literature review, which guides
Approximately one in five adults smokes cigarettes in America. In 2011, adult
cigarette smoking prevalence ranged from 11.8% in Utah to 29% in Kentucky, with a
median of 21.2% across all states. Similarly, the distribution of youth s
Ordinal Logistic Regression with SPSS
Revised: Tuesday, March 26, 2013
Ordinal logistic regression
With ordered categories, there is some underlying construct that accounts for the order of the categories. F
Much of the material in this hand-out was taken from Marija Norusis’s book
Advanced Statistics Procedures Companion.
Ordinal regression is used With ordinal dependent (response) variables,
where the 1ndependents may be craggoﬁg
Binary Logistic Regression
Main Effects Model
Logistic regression will accept quantitative, binary or categorical predictors and will code the latter two in various
ways. Heres a simple model including a selection of variable types - the criterion variabl
GSR719, sp13 update MWood p1
HINTS FOR CATEGORICAL AND METRIC DATA ANALYSIS
The term metric data is typically applied to instances where there is some kind of count -i.e.
a true zero value and a range of non-zero values. Examples would be age in years, he
GSR719 Feb. 2009 p 1
Overview and Strategies for Missing Data
Having considered sampling the next topic for us to consider is the quality of the data itself.
Below are some condensed (as always) ideas about missing data.
Missing Completely at Random, Miss
DATA ANALYSIS AND CODING ISSUES
GSR719 Spring 13
The rst stage of data analysis (after addressing nonresponse and missing data is combining and re-combining
of questionnaire items. I begin here with consideration of levels of measurement (we return to thi
GSR719 07March M.Wood p1
REPORTING ODDS RATIOS VS. MEASURES BASED ON PERCENTAGES.
Odds ratios, although useful as a component of multivariate analysis via maximum
likelihood techniques (e.g. logistic, multinomial, and ordinal regression) are