Hierarchical Linear Modeling II EDLD 610 4 Credits CRN 36079 University of Oregon, Department of Educational Methodology, Policy, and Leadership Spring 2010 Term Syllabus Rev. Date 28 March 2010 Subject to Change Meeting Days/Time: TR 10:00 a.m. - 11:50 p.m. Location: 146 HEDCO Instructor: Akihito Kamata, Ph.D. Professor, Educational Methodology, Policy, and Leadership E-Mail: kamata@uoregon.edu Phone: (541) 346-5065 Fax: (541) 346-5174 Address: 102S Lokey Education Bldg 5267 University of Oregon Eugene, OR 97403-5267 Office Hours: Tu 12-2pm, or by appointment DESCRIPTION This second course of hierarchical linear modeling (HLM) continues to provide an introduction to multilevel models. Data with a nested structure occur often in the context of social science research (e.g. students nested within classrooms or schools, repeated measures nested within subject). When data are nested, the assumption of independence required of many standard analysis techniques is violated. Multilevel models allow for (and account for) the dependency present in nested data. Students will continue to learn about a variety of multilevel or hierarchical models appropriate for a broad range of applications. The class will begin with an introduction to regression models for binary and ordered category outcome data. The focus on the HLM will be followed with coverage of hierarchical generalized linear model (HGLM) for binary and ordered category data, as well as models for nominal category data and count data, multilevel measurement models, and cross-nested multilevel models. Topics discussed within the context of each multilevel model include hypothesis testing, evaluation of model fit, and computer packages that can be used to estimate the various multilevel models. COURSE PREREQUISITES EDLD 629 (HLM I) OBJECTIVES In this course, you will become familiar with Logistic regression model for binary outcome data, and ordinal logistic regression model for ordered category outcome data. HGLM for binary, ordinal, nominal, and count outcome data. Multilevel measurement modeling. Multilevel modeling for latent variables. Multilevel modeling for cross-nested design. TEXTBOOKS & READING MATERIALS Required textbooks: Raudenbush, S. W., & Bryk, A.S. (2002). Hierarchical Linear Models (2 nd ed.). Thousand Oaks, CA: Sage Publications. [This book is denoted R&B in the readings list below.] 1 P age
Raudenbush, S. W., Bryk, A. S., Cheong, Y. F., & Congdon, R. (2004). HLM 6: Hierarchical linear and nonlinear modeling. Scientific Software International. [A version of this manual also downloads along with student version of HLM.] Required readings: Kamata, A. (2001). Item analysis by the hierarchical generalized linear model. Journal of Educational Measurement, 38, 79-93. Kamata, A., Bauer, D. J., & Miyazaki, Y. (2008). Multilevel Measurement Model. In A. A. O Connell & D. B. McCoach (Eds.). Multilevel Analysis of Educational Data. (pp.345-388). Charlotte, NC: Information Age Publishing. [This article is denoted K,B&M in the readings list below.] Other Suggested readings: O Connell, A. A., & McCoach, D. B. (2008). Multilevel Modeling of Educational Data. Charlotte, NC: Information Age. COURSE STRUCTURE The course will use a traditional lecture format. Lecture slides will be made available on Blackboard, and students are expected to download and bring them to the class. Student participation is highly expected. Therefore, students are expected to read assigned readings and slides before each class meeting and bring questions. ASSIGNMENT PROJECTS The course will require students to complete one worksheet and two assignment projects. The worksheet is a review exercise of HLM, which will be due on Tuesday of the 2 nd week in class. The first project will be due in the 6 th week of the quarter. It will require data analysis to analyze a selected data set by using HGLM and interpret the results. The second assignment project will be due on Friday of the 9 th week. It is another data analysis project for a selected data set to apply Multilevel Measurement Model. Except for the worksheet, you may work in a group of two or three on these homework assignments; a team of students should turn in one copy of the group homework with all names listed. Every group member will receive the same grade on an assignment completed as a group. FINAL PROJECT The final project will consist of a paper (approximately 10 double-spaced pages in APA style) presenting an HLM analysis of a substantive nature, that applies either HGLM, MMM, or cross-nested model. For extra credit, a 20-minute presentation summarizing the paper can be presented on the final day of the class (Thursday, June 3). The paper and presentation are due on Thursday, June 3. You are encouraged to work together in groups of two or three on the project. Various datasets are publicly available, or you can use your own data if you have some with a hierarchical structure. GRADING POLICY Your final grade for this course will be determined based on attendance/participation (5%), one worksheet (5%), two data analysis assignment (65%), and the final project (30%). Your final grades will be based on the total number of points accrued during the term. There will not be a curve. Final letter grades for the course will be calculated as follows: A+ 97-100% A 93-96.9% A- 90-92.9% B+ 87-89.9% B 83-86.9% B- 80-82.9% C+ 77-79.9% C 73-76.9% C- 70-72.9% D+ 67-69.9% D 63-66.9% D- 60-62.9% F < 59.9% 2 P age
Please note that if this class is taken P/NP, 80% or higher is required to pass the class. COURSE INCOMPLETES Students are expected to be familiar with university policy and procedures which result in failing to complete the course by the end of the term in which it is offered. Please see http://interact.uoregon.edu/pdf/sas/aincgrdcon.pdf. 3 P age
TENTATIVE SCHEDULE OF TOPICS AND ASSIGNMENTS Week Day Projects & Exams Topics Readings 1 3-30 4-1 Introduction Logistic Regression 1 Binary Outcome Model 2 4-6 4-8 3 4-13 4-15 Worksheet due (4-6) HGLM 1 Introduction Binary Model Logistic Regression 2 Ordinal Category Outcome Model (pp.291-309) 4 4-20 4-22 HGLM 2 Ordinal Model Multinomial Model (pp.317-332) 5 4-27 HGLM, continued Poisson Model Binomial Model (pp.309-317) 6 (AERA: no class on 4-29) (AERA: no class on 5-4) 5-6 7 5-11 5-13 Project 1 due (5-7) Multilevel Measurement Model Relationship to IRT Multilevel Measurement Model, continued Relationship to CTT, IRT, and SEM Kamata (2001) R&B Ch.11 (pp.365-371) K,B&M (2008) 8 5-18 5-20 Multilevel Measurement Model, continued HLM for latent variables R&B Ch.11 (pp.336-364) 9 5-25 5-27 Cross-nested designs R&B Ch.12 (pp.373-398) 10 6-1 Project 2 due (5-28) Cross-nested designs, continued 6-3 Final Project due (6-3) Presentation of Final Papers 4 P age
ATTENDANCE POLICY Attendance at all class and participation to discussions is required. ABSENCE POLICY Students must contact the instructor in case of illness or emergencies that preclude fulfilling course requirements scheduled or attending class sessions. Messages can be left on the instructor's voice mail or e-mail at any time of the day or night, prior to class. If no prior arrangements have been made before class time, the absence will be unexcused. If you are unable to take a quiz or exam due to a personal and/or family emergency, you should contact your instructor or discussion leader as soon as possible. On a case-by-case basis, the instructor will determine whether the emergency qualifies as an excused absence. ACADEMIC MISCONDUCT POLICY All students are subject to the regulations stipulated in the UO Student Conduct Code (http://www.uoregon.edu/~conduct/). This code represents a compilation of important regulations, policies, and procedures pertaining to student life. It is intended to inform students of their rights and responsibilities during their association with this institution, and to provide general guidance for enforcing those regulations and policies essential to the educational and research missions of the University. CONFLICT RESOLUTION Several options, both informal and formal, are available to resolve conflicts for students who believe they have been subjected to or have witnesses bias, unfairness, or other improper treatment. It is important to exhaust the administrative remedies available to you including discussing the conflict with the specific individual, contacting the Department Head, or within the College of Education you can contact Joe Stevens, Associate Dean for Academic Affairs, at 346-2445 or stevensj@uoregon.edu or Surendra Subramani, Diversity Coordinator, at 346-1472 or surendra@uoregon.edu. Outside the College, you can contact: UO Bias Response Team: 346-1139 or http://bias.uoregon.edu/whatbrt.htm Conflict Resolution Services 346-0617 or http://studentlife.uoregon.edu/programs/crs/ Affirmative Action and Equal Opportunity: 346-3123 or http://aaeo.uoregon.edu/ DIVERSITY It is the policy of the University of Oregon to support and value diversity. To do so requires that we: respect the dignity and essential worth of all individuals. promote a culture of respect throughout the university community. respect the privacy, property, and freedom of others. reject bigotry, discrimination, violence, or intimidation of any kind. practice personal and academic integrity and expect it from others. promote the diversity of opinions, ideas and backgrounds which is the lifeblood of the university. 5 P age