|1st & 4th floors of the Mathematical Sciences Building, UC Davis|
|9-5 (closed 12-1 for lunch)|
Statistics at UC Davis is a small/intimate major and is considered great preparation for careers in several fields ranging from business to science. Because of the relatively small number of course requirements, students often choose to double major in statistics and their chosen field of application.
At UC Davis statistics majors have a few different degree options. The department offers a B.S. in Statistics and a B.S. in Statistics with a computational emphasis. The new B.S. in Applied Statistics is similar to the former B.A. in Statistics, which offers a more flexible set of course work, ideal for double majors in the social sciences. Here students may even switch stats 131AB for the less rigorous stats 130AB.
As for graduate degrees one can either earn a M.S. or a Ph.D. The masters degree in statistics actually requires only a few core graduate courses. Since many graduate students were mathematics or other science majors as undergraduates, the first year or so of this degree may include several undergraduate statistics classes. For a Ph.D. in statistics or a Ph.D. in Biostatistics one must take a more rigorous set of courses and conduct extensive research.
The department also offers a minor in statistics. Requirements are the core stats courses such as 106 (Analysis of Variance), 108 (Linear Regression), 130A/131A (Probability Theory), 130B/131B (Mathematical Statistics) and one upper division course with 130B or 131B as a prerequisite. If you are a math major intending on going to grad school in statistics or a bio major intending on going to grad school in biostatistics, these courses are considered to be the key preparation you need, so that you don't enter having to take a bunch of undergrad classes. Specifically, many people who plan on going to grad school in biostats are either bio majors who minor in statistics or math/stats majors who minor in Quantitative biology and bioinformatics.
Dr. Jie Peng, Undergraduate Adviser (For Fall 2010, while Dr. Chris Drake is on sabbatical)
Elizabeth Dudley, Undergraduate Program Coordinator
Alejandra Garibay, Peer Advisor (2010-2011)
Check the UC Davis Statistics department's Course Descriptions
13 - Elementary Statistics. Learn the basics of statistics including but not limited to: probability distributions, hypothesis testing, confidence intervals, combinatorics, simple linear regression, one-way/two-way ANOVA. Possibly the easiest class in mankind. Most people can learn this material on the job, but it might be good to take it anyways since employers look for it for any discipline.
32 - Statistical Analysis through Computers. Nearly equivalent to STA 13 in terms of statistical concepts covered; yet, there is more emphasis in the usage of computer packages (R). For stat majors, this is a lot more useful course than STA 13.
100 - Applied Statistics for Biological Sciences. Nearly equivalent to STA 13 in terms of statistical concepts covered. The emphasis is on biological applications. Basically the same course as STA 13 and 32, but it is considered an upper division course.
103 - Applied Statistics for Business and Economics. Goes a little deeper than STA 13, applied problems in business and economics.
104 - Nonparametric Statistics. Many statistical analyses are based on common properties of known statistical models. Nonparametric statistics focus on parameterization via the data. These parameters are flexible and thus distribution free. This class teaches you how to apply the most common nonparametric statistical tests. Fit for more unusual problems. This potentially can be a difficult course, but usually the students are non-majors and that dilutes its rigorousness. There used to be an upper division non parametrics class that was much more rigorous, but this is no longer the case.
106 - Analysis of Variance. Teaches the mathematics of basic ANOVA. Considered one of the easiest classes that one can take in the major. Stats 106 and 108 have a reputation of being more or less plug and chug classes. Topics include 1-way and 2-way ANOVA, complete randomized block designs, Analysis of Covariance, and nested ANOVA.
108 - Linear Regression. Teaches the mathematics (and data analysis depending on Prof.) of simple linear regression. Unfortunately, it doesn't teach you much more than that. The statistics department desperately needs an undergraduate class for nonlinear regression. Topics include simple linear regression, multiple linear regression, ANOVA approach to regression, model selection criteria (AIC, Adjusted R2, Mallows' Cp), backwards elimination, and forward selection model building.
120 - Probability and Random Variables for Engineers
An easier version of STA 131A. This is the old requirement for EE/CE majors, with the new requirement being EEC 161 starting Fall 08. CSE majors must take the more challenging STA 131A class.
130AB - Brief Mathematical Stats and Probability Theory. Supposedly easier than STA 131ABC, but depending on the Prof., that is not always the case.
131A - Probability Theory. Intro to probability theory. Learn about continuous and discrete probability distributions, CLM, moments, expected values, etc. Possibly the most important course in the stats major. Everything else (like hypo. testing) follows from the base knowledge of probabilities.
This class is often considered to be better than its sister class math 135A. The stats class focuses more on applications/problem solving, where the math class does deep into the theory. Most people tend to agree that the stats version is also easier. -MattHh
The amount of theory in this class largely depends upon the teacher. Some professors stick more to applications of probability, while others go deep into the mathematics behind probability, such as Roussas.
131BC - Mathematical Statistics. You get taught the mathematics behind estimation, hypo. testing, simple linear regression, ANOVA, convergence and nonparametric statistics. The mathematical rigor not withstanding, the subjects covered here are quite boring (IMO).
135 - Multivariate Data Analysis. Most of the material in undergraduate statistics courses are taught in a univariate setting. Multiple variables arise commonly in real world situations. Hence, this class tends to deal with more realistic data sets. This class does not give as much rigor as the similar STA232C course, but for an undergraduate course, you take what you can get.
137 - Applied Time Series Analysis. You get to learn the basics of time series analysis (starting with AR and MA models). Get to use Shumway's time series software for Windows, ASTSA. No other undergraduate course deals with time series, widely used in economics and biostatistics (longitudinal analysis).
138 - Categorical Data Analysis. Learn the analysis of categorical data. Most of the times, you use partitioned count data. The class has been taught by Rahman Azari for the past several years, and is a required course for those pursuing the B.S. Statistics option, unless you can get signed off on taking a different 130-level class instead (135, 137).
141 - Statistical Computing. Traditionally taught by Temple Lang. Class consists of multiple computing assignments (in R) with a final project at the end. There are no exams, but the assignments are time consuming enough to the point that you rather take an exam than finish an assignment. You get preached that math is really not that useful on the broad scale. Assignments are incredibly open-ended and allow for a total exploration of the data, with the focus being on how to succinctly express large volumes of data and deal with human error in your data sets.
145 - Bayesian Statistics. Bayesian statistics is a completely different way of doing statistics. Applications in the real world has increased in recent days thanks to the increase in computing power. Used to be taught by Wes Johnson who emphasized its application using WinBUGS or JAGS. Johnson is at UCI, so Samaniego takes over and he emphasizes theory, which is cryptic at best.
CLIMB - is a program that focuses on mathematical and statistical modeling in biology. If you are interested in going to grad school in biostatistics, applied statistics or mathematics, or biology, this research program is for you.
Wolfgang Polonik (Nonparametric Statistics, Probability Theory, Mathematical Statistics)
One of the most helpful professors, somewhat easy too
Thick German accent
Since becoming department chair, he no longer teaches at the undergraduate level, but he still teaches STA231C, Graduate-level mathematical statistics.
George Roussas (Probability Theory, Mathematical Statistics)
One of the oldest professors on campus whose courses are quite rigorous. Likes to prove everything.
Thick Greek accent
Makes you buy his textbook, which is quite expensive
Very different teaching style in lecture versus office hours. If you find yourself having trouble in his lectures, give his office hours a chance. He sits down and really makes sure you understand every step.
Prabir Burman (Biostatistics, Analysis of Variance, Regression Analysis, Multivariate Statistics)
One of the best professors
Has a bowl of chocolates in his office for students to take whenever he's around
I took Regression Analysis with him. He's a genius, explains concepts clearly and exams are straight forward. Wish he taught more undergrad stats courses.
Fushing Hsieh (Biostatistics, Analysis of Variance, Regression Analysis)
One the easiest professors
Duncan Temple-Lang (Computational Statistics)
He is one of the developers for R
Talks fast, but really easy to get along with.
Responds very quickly to emails and class mailing list questions
Chris Drake (Biostatistics, Sampling Theory)
Francisco Samaniego (Bayesian Analysis, 130A/B)
His STA 145 is hard (which is fine), but not useful in terms of applying the knowledge in the future.
His STA 130A/B class is VERY theoretical
Jane-Ling Wang (Longitudinal Data Analysis, Survival Analysis)
Somewhat methodical, but good teacher and person nonetheless.
Good to take STA 131A from her.
Useful Statistical Packages
R Wikipedia - R Homepage - The somewhat hard to understand command-line statistical package. It does not have the neat GUI features as the commercial version (S-PLUS); yet the ever-growing community of R developers provide add-ons to facilitate unique routines which make this a cunning edge program for research. R is for people who have decent knowledge of programming and a constant supply of novel problems in data analysis.
SAS Wikipedia - SAS Homepage - Arguably the most used (and coveted by employers) program for people in the field of business and economics. The programming language is even less intuitive then R, but there are many resources and professionals that can help you in learning the language. It is also only used for data analysis, so one can't get as creative in its analysis as R, but it's very fast. Good to learn for people who are looking to work in companies which require data analysts. It also has the ugliest graphics engine ever.. (but again, it's fast).
SPSS Wikipedia - Statistics Package for the Social Sciences Homepage, another popular commercial statistics software, but only in the social sciences. Basically unused outside of the departments of sociology, human development, psychology, etc.