Title: Learning analytics in higher education
Author Name: Dasmyne
1. Introduction
This chapter will introduce readers to the use of learning analytics within the field of higher education. Learning analytics (LA) is a relatively new diagnostic tool used to evaluate behavior with regards to teaching and learning. The goal of learning analytics is to determine the effectiveness of a learning experience. Elias (2011) defines LA as “the measurement, collection, analysis, and reporting of data about learners and their contexts for purposes of understanding and optimizing learning and the environments in which it occurs.” Through identified patterns stakeholders can determine areas in which improvement can be implemented for learners, instructors, and organizations. Instructors can refine their teaching methods. Learning analytics can improve learning behavior through the development of learning skills and strategies. It promotes and develops important skills such as communication, critical thinking, and collaboration. Within learning analytics there are four maturity levels that provide guidance for instructional designers in creating and implementing programs.
Four levels of Maturity within Learning Analytics:
Descriptive Analytics: Identifies data that can be collected.
Diagnostic Analytics: Looks for patterns that explain learner behaviors and learning outcomes.
Predictive Analytics: Hypothesizes future trends and outcomes.
Prescriptive Analytics: Provides specific solutions for student success.
In the field of higher education, predictive and prescriptive analytics are most used in prescribing solutions for student success. Without a standardized methodology, learning analytics has been implemented using diverse approaches for various objectives. (Gasevic 2016) This allows for creativity in creating, designing, and implementing programs and solutions that will increase retention and progression.
2. Overview of the Case
Institutions more commonly use learning analytics data to monitor or measure student progress than to predict success or prescribe intervention strategies. The latter activities are indicators of true learning analytics, while the former are conventional best practices of using data and information in traditional ways to inform decisions. (Arroway et al 2016) Georgia State University is one of the top public universities in the state of Georgia. The university is best known for its research, innovation, and degrees awarded to minority students. University stakeholders discovered there was an achievement gap within the school. Stakeholders were interested in improving student outcomes through inclusion rather than exclusion. In 2011 the university made a commitment to improving graduation rates, but not by turning our backs on the low-income, underrepresented, and first-generation students who we have traditionally served. (Renick 2017). By using true learning analytics, Georgia State was able to implement solutions that increased student success and retention. The university monitored the behavior of students through data going back ten years. Through data mining Georgia State was able to use predictive analytics to identify risk factors for students. For example, if data showed freshmen biology students who took afternoon courses were more likely to drop out than those who took morning classes; the university would pay close attention to incoming biology students taking afternoon courses. They would also creative initiatives to steer more of these students towards take morning courses.
3. Solutions Implemented
Through their analysis the university realized students several academic barriers to student success. In response to low retention rates Georgia State University implemented several solutions based on their findings.
Personalized Academic Advisement:
Prior to 2102, academic advisors were meeting with about 1,000 students per year. Those meetings were mostly with students who were on track for graduation. Struggling students often slipped through the cracks. By the time many students spoke with their advisor it was often too late, the damage had been done to their GPA and program eligibility. In 2012 the university revamped their advisement model. They invested in well-structured training for a cross functional team of advising professionals. Advisors not only understand academic policies, but admissions and financial aid too. The school also launched a GPS advising model that uses predictive analytics to track 800 risk factors and identify at-risk students. Once identified, academic advisors continuously work with the students to get them back on track.
Stronger Academic Policies:
Through the analyzed data, the university discovered unintended barriers to academic success for students. Some of the colleges within the school had academic policies or no policies at all that left students spinning the wheel. For example, previously the Robinson College of Business (RCB) had an overall GPA requirement of 2.8 to take upper level major courses. Thousands of students were taking multiple unnecessary courses in other disciplines to meet this requirement. RCB now requires students to have a 2.8 in their Area F, which are the courses specific to the major to take upper level business classes. STEM majors at Georgia State would often be at a standstill with progression in the major. STEM courses build upon each other. For example, in order to take Principles of Physics (PHYS 2211/2211L), a student must pass Calculus I (MATH 2211) with a C or higher. In order to take Calculus I, a student will need to pass Precalculus (MATH 1113) with a C or higher. Students were taking the basic introductory courses multiple times with no success. This delayed major progression and graduation. The university has recently implemented a two attempt Math policy for College Algebra (MATH 1111) and higher. If a student is unsuccessful after taking a higher-level math a second time, they are ineligible to enroll in that course without going through an appeal process.
Flipped Learning:
Georgia State has flipped seats in numerous courses by using adaptive-learning technology such as blended learning. Traditional lecture sections have been replaced with students reviewing lecture material online and applying the concepts in class. This strategy has proved successful in getting students through critical gateway courses into their chosen majors. “Since introducing this model failure rates for introductory course dropped 35 percent. “(CPE 2017)
Other Implementations:
Some of the other implemented solutions include the creation of the Panther Retention Grant and a summer success academy for incoming freshmen. The grant eliminates the financial barrier for seniors who have maxed out on their financial aid. Through analytics the summer success academy targets incoming freshmen who are at risk for dropping out. The data is based on their high school GPA and standardized test scores (ACT and SAT).
4. Outcomes
Many learning analytics advocates believe capturing, archiving, and analyzing student profiles and behaviors will lead to improved institutional decision making, advancements in learning outcomes for at-risk students, greater trust in institutions due to the disclosure of data and significant evolutions in pedagogy, among other things. (Long & Siemens 2011) Student clearinghouse reports that the national completion rate for the fall 2012 cohort of first-time post-secondary students is 58 percent. Black and Hispanic student total completion rate increased considerably, to 48 and 57 percent, respectively, for four-year starters. (2018) This has proven to be true for Georgia State. By predicting and prescribing intervention strategies Georgia State was able to increase its graduation rate from 32 percent in 2003 to more than 54 percent in 2017. “As the number of students who declare a major in STEM has held steady, the number who complete those majors has doubled.” (Murtrie 2018) Through grant funding Georgia State has expanded the GPS model and adapted its advising strategy to increase graduation rates for over 20,000 students. The university continues to make improvements on their graduation rate by closely monitoring student progression through their EAB Navigate system. Academic advisors are mapping students out to degree completion. This gives students a clear vision of what is needed to stay on track and achieve academic success.
5. Implications
Analytics have made it clear that with the correct guidance students can achieve academic success in higher education. By simply making them aware of any problems and working with them on solutions, students are more engaged in their learning experience. Georgia State University has become a model for holistic academic advisement in the nation. Other schools and universities have used the Georgia State advisement model to increase student retention and success at their schools. Although big data and technology play an important role in understanding and fixing the problems, but it was ultimately the people who solved the problems. When there is a culture that prioritizes and supports student success the tools will be effective. If the culture is nonexistent, the tools will be ineffective. “To build a successful learning analytics program at an educational institution, it’s important to engage and inform school leaders, listen to the needs of teachers and students, and educate users on how to consume and act on the data that’s presented. Teachers and administrators will only be committed to data-driven decision making if they can see its value and are educated about how to turn insights gleaned from data into action.”(Miller 2018) Although learning analytics provide many benefits there are also concerns. While emerging learning analytics practices hold some promise to improve higher education, they are morally complicated and raise ethical questions, especially around student privacy. To alleviate some of these valid concerns, institutions should be mindful of how and to whom access to student information is granted.
References
Arroway, Pam, Glenda Morgan, Molly O’Keefe, and Ronald Yanosky. Learning Analytics in Higher Education. Research report. Louisville, CO: ECAR, March 2016.
McMurtrie, Beth. “Georgia State U. Made Its Graduation Rate Jump. How?” CHE, 2018, www.chronicle.com/.
Gasevic D., Dawson, S. and Pardo, A. (2016), “How do we start? State and directions of learning analytics adoption”, International Council for Open and Distance Education
AGB, AGB. “CASE STUDY: GEORGIA STATE UNIVERSITY.” AGB, 27 Oct. 2020, agb.org/.
Long, P., & Siemens, G. (2011). Penetrating the fog: Analytics in learning and education. Educause Review, 46(5), 30–40
Elias, T. (2011). Learning Analytics: Definitions, Processes and Potential.
Miller, Kelsey. “What Is Learning Analytics & How Can It Be Used?” Northeastern University Graduate Programs, 25 Nov. 2020, www.northeastern.edu/graduate/blog/learning-analytics/.
Blog, NSC. “National Six-Year Completion Rate Reaches Highest Level, 58.3 Percent, Since the National Student Clearinghouse Research Center Began Tracking.” Clearinghouse Today Blog, 17 Dec. 2018, www.studentclearinghouse.org/nscblog/national-six-year-completion-rate-reaches-highest-level-58-3-percent-since-the-national-student-clearinghouse-research-center-began-tracking/.
CPE_Communications, Author. “Georgia State University Is Closing Achievement Gaps and Confounding Expectations.” Policy Insight, 12 Apr. 2017, insight.councilonpostsecondaryeducation.org/georgia-state-university-is-closing-achievement-gaps-and-confounding-expectations/.