Psychology 5520-00

Multilevel Longitudinal Analysis / Rasch Modeling

Spring 2016

COURSE: PSY 5950 – 01 23280

TITLE: Multilevel and Longitudinal Analysis plus Rasch IRT Models

CLASS SCHEDULE: Wednesdays1:00 PM – 3:30 PM; Holt 303

CREDIT: 3 hours graduate credit

FACULTY: Michael Biderman

Holt 350F

423-425-4268 / 423-316-9504

By appointment

ADA STATEMENT: Attention: If you are a student with a disability (e.g. physical, learning, psychiatric, vision, hearing, etc.) and think that you might need special assistance or a special accommodation in this class or any other class, call the Office for Students with Disabilities at 425-4006, come by the office - 102 Frist Hall or see http://www.utc.edu/OSD/

If you find that personal problems, career indecision, study and time management difficulties, etc. are adversely affecting your successful progress at UTC, please contact the Counseling and Career Planning Center at 425-4438 or http://www.utc.edu/Administration/CounselingAndCareerPlanning/.

REQUIRED TEXTS:

Heck, R. H., Thomas, S. L., & Tabata, L. N. (2010). Multilevel and Longitudinal Modeling with IBM SPSS. 2nd Edition. New York: Routledge.

Bond, R. G., & Fox, C. M. (2007). Applying the Rasch Model: Fundamental Measurement in the Human Sciences. 3nd Ed. New York: Routledge. ISBN: 978-0-8058-5462-6 (paper); 978-1-4106-1457-5 (e book); 978-0-8058-5461-9 (cloth). (New purchases come with a disk with data and a student version of WINSTEPS, a Windows based IRT program.)

RECOMMENDED

Singer, J. D., & Willett, J. B. (2003). Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. New York: Oxford University Press.

This course is an advanced graduate course designed to present the fundamentals of multilevel and longitudinal analysis using SPSS and to introduce concepts of item response theory through the Rasch model. PSY 5100/5110 and PSY 5130 are prerequisite for the course. PSY 5950-Advanced SPSS is a strongly recommended prerequisite. Students enrolled in this course should be well-versed in multiple regression analysis, including the use of regression techniques to compare group means as done in traditional analysis of variance. Students should be familiar with analysis of repeated measures data using SPSS. Students should be familiar with basic measurement concepts, such as reliability and validity. The analyses taught in this course are more complex than analyses presented in the prerequisite courses and in any undergraduate Psychology courses at UTC. A higher than average level of intellectual capability, maturity, and interest in data analysis will be required to perform successfully in this course.


Topics

HLM - Heck

Chapter 1: Introduction

Chapter 2: Preparing and Examining Data

Chapter 3: Defining a Basic Two-Level Multilevel Regression Model

Chapter 5: Examining Individual Change with Repeated Measures Data

Singer & Willett Chapters 3,4

Singer & Willett Chapter 5

Rasch Model – Bond & Fox

Chapter 1: Why Measurement is Fundamental

Chapter 2: Important Principles of Measurement Made Explicit

Chapter 3: Basic Principles of the rash Model

Chapter 4: Building a Set of Items for Measurement

Chapter 5: Invariance: A Crucial Property of Scientific Measurement

Chapter 6: Measurement Using Likert Scales

Chapter 7: The Partial Credit Rasch Model

Chapter 10: The Rasch Model Applied Across the Human Sciences

Chapter11: Rasch Modeling Applied: Rating Scale Design

Dimmed chapters are recommended but not required for the course at the present time.

Grading: 1 homework assignment worth 10 points per week consisting of 1-4 parts per assignment.

Longer take-home final – About 10 problems each worth 10 points. Given out on last meeting.

Due date will be ______

Finals will be returned for possible resubmissions on ______

Retakes due at 9 AM on ______.

Regular assignments graded on 10 point scale.

Take-home final worth about 100 points.

1 resubmission of each assignment allowed. Cost is 1 point. So don’t bother to resubmit a 9.

Late submissions and Resubmission must be by next class period after the due date.

Original submissions must accompany resubmissions.

Submissions must be easy to follow, to-the-point, neat, addressing all the relevant issues.

Submissions must be in “portrait” mode.

Email submissions will cost you 1 point per submission.

Submit only enough computer output to show you've done the analyses or to make whatever point you intend – save a tree!! Superfluous output will result in loss of points.