PS 133 is a 4-unit graded undergraduate seminar in which students will design, carry out, analyze, and write up a survey-experiment. Students will learn about: the varying designs and objectives of survey experiments, the creation of a survey instrument and randomization procedures using the Qualtrics software program, human subjects protections and the filing of a research protocol, running experiments using workers from Amazon Mechanical Turk as subjects, the formulation of data analysis plans, the basics of data analysis using STATA, the creation of tables and charts to display results, and the writing of an empirical research paper. Initially, class sessions will be focused on helping students obtain a deep appreciation of survey in political science through presentations by the professor and invited graduate students, readings, and discussion. Subsequently, class sections be focused on helping students develop, implement, and analyze their experiments.
Note: Students should waitlist in Phase 2
Students interested in taking the class must have taken PS 3 with Professor Stoker in Fall 2016 or Fall 2014. Students should sign up on the waitlist for the class before the end of the fall term, 2016. In mid to late December, Professor Stoker will email the waitlisted students with a link to the online application form for the class. The application requires students to answer background questions, provide a personal statement, and elaborate a research proposal that the student might want to pursue in the class. Professor Stoker will review the applications and make admissions decisions before the start of the spring semester.
Grades will be based on class participation (10%), weekly or bi-weekly written assignments (totaling 30%), class presentations (totaling 15%), and a final paper (~20 pages plus tables and appendices) containing the write-up of the experiment (45%).
One book is required for the course: Diana C. Mutz’ 2011 Population-Based Survey Experiments, Princeton University Press. Other readings come from journal articles, edited volumes, and websites, and will be made available on the Bcourses website.
In addition to purchasing the Mutz book (approximately $20), students will need to use their own funds to pay their subjects on MTurk. However, this should not run to more than $50-$60 (and I may be able to find some $ to subsidize this). Students may also decide to purchase the STATA program so that they can analyze data using STATA on their own computers. A six-month license costs between $38 and $75, depending on the size of the dataset. https://www.stata.com/order/new/edu/gradplans/student-pricing/