Using Data from a Learning Management System to Monitor Student Performance

Size: px

Start display at page:

Download "Using Data from a Learning Management System to Monitor Student Performance"

Kenneth Norton
8 years ago
Views:

1 Georgia Southern University Digital Southern SoTL Commons Conference SoTL Commons Conference Mar 28th, 4:00 PM - 5:30 PM Using Data from a Learning Management System to Monitor Student Performance Beth Dietz-Uhler Miami University, uhlerbd@muohio.edu Follow this and additional works at: Part of the Curriculum and Instruction Commons, Educational Assessment, Evaluation, and Research Commons, Educational Methods Commons, Higher Education Commons, and the Social and Philosophical Foundations of Education Commons Recommended Citation Beth Dietz-Uhler, "Using Data from a Learning Management System to Monitor Student Performance" (March 28, 2013). SoTL Commons Conference. Paper This presentation (open access) is brought to you for free and open access by the Programs and Conferences at Digital Commons@Georgia Southern. It has been accepted for inclusion in SoTL Commons Conference by an authorized administrator of Digital Commons@Georgia Southern. For more information, please contact digitalcommons@georgiasouthern.edu.

2 Using Data from a Learning Management System to Monitor Student Performance Beth Dietz-Uhler Miami University

3 Introduction Recently, there have been governmental demands to increase student success in higher education (e.g., Obama, B. 2009). One such way to increase student success is to increase retention in courses. Among the strategies for doing so, learning analytics offers much promise. Briefly, learning analytics is the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the and the environments in which it occurs (Long & Siemens, 2011, p. 32).

4 There is growing support for the use of learning analytics in higher education, as evidenced by a recent Horizon Report (Johnson, Smith, Willis, Levine, & Haywood, 2011) suggesting that it is an up-and-coming technology that will have a significant impact on education in the next five years. Much of the recent work in learning analytics has focused on big data, often using all possible data that exists for any one student. For example, Purdue University successfully uses Signals to monitor student progress and performance in online classes (LUOJT, 2012). Chico State University is also successfully used big data to improve student success (Whitmer, Fernandes, & Allen, 2012).

5 While these large-scale efforts are commendable, they do not speak to what individual faculty can easily do in their courses. The objective of this presentation is to show how faculty can make use of statistics available in their Learning Management System (LMS) to make predictions about student performance, as well to monitor student progress in their courses.

6 Method O All students (N = 28) from two sections of Introduction to Psychology (online) agreed to allow their performance and behavioral data to be used. O Two types of data: O Behavioral: Data generated from Course Management System (CMS) (e.g., number of times accessed syllabus, number of forum posts read) O Performance: Date generated by student (e.g., number of s sent by student, grades)

7 Questions O What variables predict student performance overall? O What are the early predictors of student performance?

8 Results Overall Performance (Final Grade) Exam 2 Grade Third Quarter Mini Project Performance Exam 1 Grade Fourth Quarter Number of Forum Posts Read Third Quarter Quiz Performance Fourth Quarter Critical Thinking Performance Mini Project Average Grade F (7,20) = , p <.001, R 2 = 99.5

9 Results: Early Predictors (First Quarter Variables) Overall Performance (Final Grade) Exam 21 Grade First Quarter Mini Project Performance First Quarter Critical Thinking Performance F (3,24) = , p <.001, R 2 = 92.6

10 Results: Early Predictors (Time 1-3) Overall Performance (Final Grade) Exam 1 Grade Module 3 Mini Project Performance Module 2 Mini Project Performance Getting Started Quiz F (4,23) = , p <.001, R 2 = 93.1

11 Results: Early Predictors (Time 1-3) Exam 1 Performance Module 3 Mini Project Performance Module 1 Critical Thinking Performance Module 1 Mini Project Performance Quiz 3 F (4,23) = , p <.001, R 2 = 76.0

12 Conclusion O In terms of predicting overall course performance, performance early in the course and later in the course is strongest predictor O The best predictors of early performance are performance variables rather than behavioral ones O The best predictors of early and overall performance are performance variables and not behavioral ones O Limitation: These data include students who generally succeeded (and remained) in the course.

13 References Johnson, L., Smith, R., Willis, H., Levine, A., & Haywood, K. (2011). The 2011 Horizon Report. Long, P., and Siemens, G. (2011). Penetrating the fog: Analytics in learning and education. Educause Review 46(5), LUOJT. Retrieved October 5, 2012 from learning-analytics/ Obama, B. (July 14, 2009). "Remarks by the president on the American Graduation Initiative." Macomb Community College. Warren, MI. Whitmer, J., Fernandes, K., & Allen, W. R. (2012). Analytics in Progress: Technology Use, Student Characteristics, and Student Achievement. Educause Review Online. Retrieved October 6, 2012 from characteristics-and-student-achievement.