Spring Graduate Courses in Survey Research
The following two classes may be of interest to graduate students in the social and behavioral sciences. The courses are independent and may be taken in any order.
SOC (PL SC) 518 Survey Methods I: Survey Design
Tuesdays, 6:30 - 9:30 PM
Instructor, Kurt Johnson (Director, PSU Survey Research Center)
This course is intended to provide graduate students the background to both evaluate published research using survey methods, and -- when combined with additional training -- to design their own surveys to collect data for their own research. Students will learn the essentials of sampling, questionnaire design, and how surveys may be implemented in different modes: telephone, face to face interviews, mail or other self-administered modes, and the internet. The course will emphasize how decisions of research design have important implications for the validity, reliability, and quantity of data that will be analyzed to answer key questions in the social, behavioral and health sciences. Topics typically include sample design, questionnaire design and item analysis, Telephone Surveys; Face to face surveys; Self administered and mail surveys; Internet Surveys; and ethics and human subjects protection: 1 week
PL SC (SOC) 519 Survey Methods II: Analysis of Survey Data
Monday & Wednesday, 4:15 - 5:30
Instructor, Eric Plutzer (Academic Director, PSU Survey Research Center)
This is an intermediate level course in quantitative analysis. It is intended for graduate students who have completed 1-2 semesters of graduate-level statistics (not general research methods) and who are interested in the application of social statistics to the unique aspects of data collected by way of surveys. Surveys have a combination of qualities that represent challenges to valid inference. These include cluster and stratified sampling, under-representation of some groups due to differential response rates, missing data due to item non-response, cross-sectional design, and coarse measurement. Quite often we use surveys to test theories that the original survey designer did not intend to address, raising issues of validity and reliability of measurement. At the same time, surveys offer a number of opportunities and, when combined with other surveys (pooled cross sections) or merged with contextual data, can address a wide range of theoretical puzzles in the social sciences.
This course provides an introduction to techniques in applied statistics that have developed specifically to address the special features of survey data. Examples of such techniques are: use of design weights, post-stratification weights, merging surveys with other surveys or auxiliary data, missing data imputation, challenges of causal inference. The class will blend an understanding of the core statistical issues with an emphasis on acquiring an intuition for the theory underlying the statistical models rather than focusing on proofs and estimation.
Last modified: 11/05/07 | Contact Webmaster
