Education has long been a black box where educators and administrators have a limited understanding of the
methods, activities, and investments that produce quality learning. Learners in turn have only a limited awareness of the opportunities offered by tools and technologies to improve self-regulation in the learning process. Over the past decade, three trends have altered information spaces: quantity of data, type of data, and increased computation power. The quantity of data is evident everywhere, presenting challenges for individuals to cope with information overload. New types of data are produced through, and captured by, social media, mobile devices, and, increasingly, automated data capture through sensors (i.e. the Internet of things). Finally, as Moore’s Law has held reasonably firm, computation power has continued to grow exponentially.
As a result of these three trends, business and organizations increasingly recognize that data is an asset – a resource to be managed and leveraged. “Big data” has been coined to describe the growing quantity, and analytic opportunity, of data. PW Anderson, in 1972, prophetically anticipated the challenge facing organizations today: more is different. As data quantity increases, new approaches for mining and drawing insight from that data are needed. Businesses have responded to the new data reality through development of business intelligence. Governments and many organizations (such as OECD) have responded by opening up their data for researchers to analyze. In the education market, interest in learning analytics is growing as well. Learning analytics is the use of intelligent data, learner-produced data, and analysis models to discover patterns and connections within that data, and to predict and advise on learning.
At a basic level, learning analytics relies on some of the concepts employed in web analysis, such as Google Analytics, as well as those involved in data mining. These analytic approaches focus on learner activity through mining clicks, attention/focus “heat maps”, social network analysis, and recommender systems. Learning analytics extends these basic analytics models by focusing on curriculum mapping, personalization and adaptation, prediction, intervention, and competency determination.
Learners constantly put-off data – sometimes explicitly in the form of a tweet, Facebook update, logging into a learning management system, or creating a blogpost. At other times, unintentionally while in the course of daily affairs (or data that is provided by someone else – such as being tagged in Facebook). These data trails are captured by organizations, waiting for some type of analysis that generates insight of value. Amazon, Google, and Facebook use this data to personalize the experience for site users and to recommend additional resources or social connections.
Each new source of data amplifies the potential value of analytics. Most individuals have a fragmented digital profile with bits and pieces of data captured in LinkedIn, Facebook, Google, Twitter, blogs, and closed organizational systems. When these data sources are brought together, especially with physical world data captured by sensors or mobile devices, they can provide organizations with a profile of an individual (the integration of data sources was the primary intent behind the Information Awareness Office’s concept of Total Information Awareness following 9/11 in the USA). The recent development of semantic linked data holds promise in education. For example, “intelligent” data3 (machine-readable semantic or linked data) can
be combined with learner-produced data, profile information, and curricular data for advanced analysis.
The data trails and profile, in relation to existing curricula, can be analyzed and then used to identify learners who are at risk of dropout or failure. This identification can then be used for automated or human intervention, personalization, and adaptation . Curriculum in schools and higher education is generally pre-planned. Designers create course content, interaction, and support resources well before any learner arrives for a course. This can be described as “efficient learner hypothesis”– the assertion that learners are at roughly
the same stage when they start a course and that they progress at roughly the same pace. Educators generally recognize that this is not they case. Learners enter courses with dramatically varied knowledge levels and progress through content at different paces. As a result, the process used by educational institutions in designing learning is in urgent need of restructuring.
Learning content, for instance, could be computed – a real-time rendering of learning resources and social suggestions based on the profile of a learner, her conceptual understanding of a subject, and her previous experience. Competence (as measured by a degree or certificate) need not even be explicitly pursued. For example, an integrated learning system could track physical and online interactions, analyze how learners are developing skills and competencies, and then compare formal and informal learning with discipline or field of knowledge. This comparison may be possible if a discipline has utilized intelligent/semantic/linked data to define its knowledge goals. Then, the learning system could inform a learner that (you are 64% of the way to a achieving a PhD in psychology, 92% to achieving a masters in science, 100% to achieving a certificate in online learning” and so on). If the learner then decided to pursue a PhD in psychology, the learning system could offer a personalized path forward that adapts constantly to knowledge the learner acquires in the course of work, formal learning, parenting, or the general process of “living life”.
Analytics for pre-defined goals can be approached from a bottom-up or top-down manner. Bottom up involves individual educators using single functionality tools to gain insight into the activities of learners. Fairly simple tools exist for conducting social network or discourse analysis. Additionally, statistics from a learning management system can provide data on login and posting frequency, indicators of learner specific habits. Top-down analytics, in contrast, require the formation of a strategy and vision on the part of an institution. When viewed from an institutional level, analytics can provide valuable information about learner success and failure by leveraging large datasets to reveal interesting correlations. Systems level analytics also gives an organization the ability to plan for automated interventions and develop learner profiles across various data sets. The analytics process, systemically deployed, consists of the stages.
When learning analytics are deployed systemically, several important, and integrated, functionality areas are needed . The analytics engine or module includes the specific analysis being performed on learner data such as social network analysis, predictions of learner failure, and evaluation of the impact of taking different course paths on learner success. The analytics engine draws data from learning management systems and personal learning environments, the social web, organizational learner profiles, and physical world data such as library usage.
The intervention engine is activated when learners exhibit signs of risk for dropout or failure or sub-optimal learning paths and activities. The adaptation and personalization engine uses intelligent learning content and compares learner ongoing activity with the goals or targets of a particular course or learning experience. Finally, dashboards enable end users (learners, educators, and administrators) to query data and visualize how different learners are performing. Many segments of the information industry – music, movies, news – have been dramatically altered by digital information and new options for end users to manage and control information. The financial and accountability pressures facing education, coupled with disruptive advances in technology, present a crisis point for the industry.
Education is adapting and evolving with the changes in today’s technology and information climate. Analytics hold the prospect of enabling changes and restructuring to the processes of teaching, learning, and administration. More importantly, analytics provide a feedback loop to track the impact of change initiatives, in education systems with predefined goals and potentially, as “assessment competences” improve, with open self-referential learning. Today’s recognition of societal complexity and the transformations occurring in the education market are providing important incentives for efforts to construct new platforms for improving the
quality of information that learners and educators use. Learning analytics open up the prospect of providing new resources for leaders, educators, and learners to make a leap in learning productivity happen.
No comments:
Post a Comment