Language, Education, Technology - Learning Analytics

This topic comes from Week 4 of the MOOC on ?Current/Future of Higher Education?

USING ANALYTICS TO INTERVENE WITH UNDERPERFORMING COLLEGE STUDENTS (INNOVATIVE PRACTICE)?by John Fritz and Eric Kunnen

A talk available at?http://www.educause.edu/eli/events/eli-annual-meeting/2010/using-analytics-intervene-underperforming-college-students-innovative-practice:

• LMSs do have some benefits, e.g. they keep track of students?
• Reference to the ECAR study of 2005 with 5 stages of analytics use by schools

Purdue Experience

• ?Signals? in Blackboard, an example of predictive ?approach of analytics
• Success = academic preparation+performance+EFFORT, hence identifying measures of effort
• CMS and other technologies (real time) => Prediction <= SIS Data (historic), Other data
• Better student performance: more Bs and Cs, fewer Ds and Fs [apparently not much impact on B-students]

UMBC Experience

• D students stay in Blackboard about 30-40%?consistently?less time than C-students
• Students may draw faulty conclusions by drawing connections between their online presence and grades on specific assignments too directly
• According to Fritz?s research, 28% students were surprised what the data showed about their performance
• Overall, it looks like it?s not clear whether and how students use these data and what they do with them, and how they interpret them
• At the institutional level, data help identify most active courses and best practice teachers
• Frequency and duration of clicks are critical to consider not just number of clicks

Grand Rapids Community College: Project Astro Experience

• Bb Building block for analytics available for installation free of charge[??]: Dashboard that shows visuals for tool use overall and by course, instructor vs. student activity, specific student activity, etc.
• http://www.starfishsolutions.com?offers an alert system: possibility to raise a flag for specific students a) manually by the instructor or student counselling office or b) automatically by the system
• More effort is needed to bring multiple systems, sources, and databases together to merge different kinds of data for decision making
• FIRPA concerns about student privacy

PENETRATING THE FOG by Long and Siemens

• A lot of decisions are not data driven
• Learning (course- and department-specific level) vs. Academic (institutional, regional, national/international level) analytics
• Interesting study cited by Morris, Finnegan and Wu about online presence and behavior of?undergraduate?students in online courses
• important caution against taking a deterministic view of analytics as opposed to a more justified probabilistic view

OPEN LEARNING ANALYTICS: AN INTEGRATED & MODULARIZED PLATFORM PROPOSAL TO DESIGN, IMPLEMENT AND EVALUATE AN OPEN PLATFORM TO INTEGRATE HETEROGENEOUS LEARNING ANALYTICS TECHNIQUES by Siemens et al (2011)

• This is a proposal for a ?abroad-based, multi-sourced, contextual and integrated? model of learning analytics (p. 6).
• It pulls together a number of data sources to draw a more holistic picture of student learning and academic effectiveness at the course and institutional level.
• This model also provides recommendations by outsourcing to external sites, e.g. Amazon, etc.

SOME KEY QUESTIONS FOR THE MOOC:?

1. What analytics tools are available for Moodle?
2. When is the proposed platform going to be ready?

Source: http://irshat.edublogs.org/2012/10/30/learning-analytics/

madonna give me all your luvin video roseanne barr president green party day 26 gronkowski new hunger games trailer sasquatch