Advanced Educational Statistics
Spring Semester, 2014
Section 091, Thursday5:30-8:15 PM, College of Education Building 005
Richard G. Lambert, Ph.D. College of Ed. Bldg. 280 704-687-8867
Course website:Class / Date / Topic / Reading / Assignment
1 / 1/9/14 / Experimental Methods / See website / 1
2 / 1/16/14 / t-tests / Chapter 10 / 2
3 / 1/23/14 / One Way ANOVA / Chapter 11 / 3
4 / 1/30/14 / Post Hoc and Planned Comparisons / Chapter 12 / 4
5 / 2/6/14 / Two Way ANOVA / Chapter 13 / 5
6 / 2/13/14 / Interactions, Simple Effects, and Planned Comparisons / Chapter 13 / 6
7 / 2/20/14 / In class work on Midterm examination
8 / 2/27/14 / Repeated Measures ANOVA / Chapter 14 / 7
9 / 3/13/14 / Mixed Models / Chapter 14 / 8
10 / 3/20/14 / Mixed Models / Chapter 14 / 9
11 / 3/27/14 / Analysis of Covariance / Chapter 15 / 10
12 / 4/3/14 / Analysis of Covariance / Chapter 15 / 11
13 / 4/10/14 / Statistical Power / Chapter 8 / 12
14 / 4/17/14 / Data collection for food tasting studies
15 / 4/24/14 / Final Presentation of Projects, In class work on Final Examination
There will not be class on 3/6/13 for spring break.
Class on 3/13/13 and 4/3/14 will be online.
Those of you who are not comfortable with the fundamentals of hypothesis testing, or need some review of the concepts from 6110 / 8110 are encouraged to read Chapters 1-7.
The midterm and final will be posted on the website. The midterm is due by 2/27/14 and the final is due on 5/1/13.
Required Text and Materials
Huck, S. (2012). Reading Statistics & Research. (6th Ed.) Boston, MA.: Allyn & Bacon.
If you are able to purchase earlier editions (3rd, 4th, or 5th) they will also be acceptable.
40% Weekly Assignments
20% Group project and presentation
Advanced topics in probability and statistics as a basis for objective decision-making in educational research; one-way and n-way analysis of variance and covariance, advanced ANOVA designs, introduction to multivariate statistical procedures. Emphasis on understanding concepts through analyses of prepared data.
RSCH 8110 or equivalent.
Today's helping professions (educators, nurses, counselors, etc.) are called upon to select and justify practices, to verify predictions of student/patient/client success, and to evaluate programs. These demands require a thorough understanding of scientific research methods for the investigation and solution of problems by rational and empirical means. Speculation, trial-and-error, introspection, common sense, and persuasive argument are no longer adequate to defend or evaluate educational practice. Researchers cannot design or implement research methods, however, without a solid foundation in probability and statistics. Research 8120 covers advanced topics in educational statistics, including: multiple correlation and regression, analysis of variance, analysis of covariance, advanced ANOVA methods, and advanced nonparametric analysis methods. The course also introduces multivariate statistical procedures. Problems to illustrate the logic and typical application of the techniques will be selected from many disciplines, with a focus on the behavioral and social sciences. The course is designed with the assumption that students should develop an appreciation of the logic of inferential statistics, rather than merely develop mechanical proficiency. In fact, the requisite numerical operations are better left to calculators and computers. This course will avoid rote memorization and stress the foundations, purposes, assumptions, appropriate application, and proper interpretation of probability and statistics concepts. The display of data and the mechanical calculation of formulas will be left to hand-held calculators and statistical software packages available on microcomputer and mainframe systems.
Access to the World Wide Web (WWW)
Access to SPSS and Excel through University Computer Labs:
1. To review the nature of educational statistics, and its role in educational research.
2. To review basic concepts of educational statistics, including data distributions, central tendency, variability, sampling distributions and estimation, hypothesis testing, one-sample and two-sample inference problems, bivariate linear correlation and regression, and introductory nonparametric methods.
3. To master a set of analysis procedures for testing inferences about population means from data obtained from more than two samples (ANOVA).
4. To master a set of analysis for the testing of inferences about population means from data obtained from more than two samples that can be classified at several levels on more than one characteristic (n-way ANOVA).
5. To master a set of analysis procedures for the testing of inferences about population means from data obtained from more than two samples after the influence of one or more correlated variables has been eliminated (ANCOVA).
6. To study special ANOVA methods for specialized research situations (e.g., repeated measures, trend analysis, hierarchical designs).
7. To introduce methods to determine the relationship between sets of variables measuring traits of a group of individuals, to reduce the dimensionality of information from sets of variables, and to test inferences about two ore more population parameters from data obtained from more than two samples (multivariate statistical methods).
Weekly Assignment Guidelines
You are encouraged to work in pairs. However, each student must turn their own copy of the answers to the assignment for each week. Each student must contribute to the homework for each week. You are free to divide the tasks with your partner as you see fit. Feel free to discuss the assignments with other groups in the class but turn in your own work.
By the third week of class, each group is to bring in a batch of data from a real world setting for use in class. Take full advantage of the opportunity to work extensively in class with a particular set of data by choosing data that you need to analyze for other purposes. If you have taken a class from Dr. Lambert in the past, please access data not used in a previous class unless given permission. The batch of data should contain student or teacher outcome variables which are measured on least two different occasions. The data should also contain at least two potential independent variables which are nominal or ordinal. These variables will be used for group comparisons. In addition, the data should contain at least two background demographic variables on the subjects. For example, if the outcome variables are teacher level, then you may choose to include variables such as teacher educational level and years of teaching experience. If the outcome variables are student level, you may choose to include variables such as lunch status and mother’s education level. Please discuss the data you are proposing to use with Dr. Lambert.
Also beginning with the third week of class, each assignment will consist of running analyses in SPSS using data obtained from the website or course CD and completing the applicable answer sheets regarding the results. You may also apply the same type of analyses to your own data for any of the assignments and substitute an answer sheet from the analyses of your data.
Moodle is the preferred method of submitting assignments. There is no reason to print out extra paper. Please bring a memory stick to class so that you can save your work each week. Please bring an electronic copy of each assignment you submit to class as well so that we can refer to them if necessary.
Guidelines for Class projects and presentations
See the course Webpage for guidelines on this component of the course.
Developing Excellent and Competent Professionals
UNC Charlotte is committed to developing knowledgeable, effective, reflective, and responsive professionals who are leaders in their fields. The content, opportunities, and experiences in this course contribute to the development of knowledgeable professionals, provide state of the art information that supports effective practices, facilitate critical analysis and reflective thinking, encourage respect in responding to human needs and differences, and generally increase skills needed by professionals who are effective leaders.
The College of Education at UNC Charlotte is committed to social justice and respect for all individuals, and it seeks to create a culture of inclusion that actively supports all who live, work, and serve in a diverse nation and world. Attaining justice and respect involves all members of our community in recognizing that multi-dimensional diversity contributes to the College’s learning environments, thereby enriching the community and improving opportunities for human understanding. While the term “diversity” is often used to refer to differences, the College’s intention is for inclusiveness, an inclusiveness of individuals who are diverse in ability/disability, age, economic status, ethnicity, gender, language, national origin, race, religion, and sexual orientation. Therefore, the College aspires to become a more diverse community in order to extend its enriching benefits to all participants. An essential feature of our community is an environment that supports exploration, learning, and work free from bias and harassment, thereby improving the growth and development of each member of the community.