Learning analytics

src: c1.staticflickr.com

Learning analysis is the measurement, collection, analysis, and reporting of data about learners and the context, for the purpose of understanding and optimizing the learning and environment in which it occurs. The related field is the mining of educational data.

Video Learning analytics

Definisi

Definitions and objectives of instructional analytics are contested. One previous definition discussed by the community suggests that "Learning analysis is the use of intelligent data, data produced by learners, and analysis models to find information and social connections to predict and counsel people's learning." But this definition has been criticized by George Siemens and Mike Sharkey.

A more holistic view of definition is provided by the analytical learning framework by Greller and Drachsler (2012). It uses a general morphological analysis (GMA) to divide the domain into six "critical dimensions".

A systematic overview of analytic learning and its key concepts are provided by Chatti et al. (2012) and Chatti et al. (2014) through reference models to study four dimensional analysis, ie data, environment, context (what?), Stakeholders (Who?), Goals (why?), And methods (how?).

It has been shown that there is broad analytic awareness across educational institutions for various stakeholders, but that the way learning analysis is defined and implemented can vary, including:

1. for individual learners to reflect on their attitudes and patterns of behavior in relation to others;
2. as student predictors requiring extra support and attention;
3. to assist teachers and support staff plans that support interventions with individuals and groups;
4. for functional groups such as course teams who want to improve the current course or develop new curriculum offerings; and
5. for institutional administrators to make decisions on things like marketing and recruitment or efficiency and effectiveness measures.

In the briefing paper, Powell and MacNeill went on to point out that some analytical motivations and implementations may conflict with others, for example highlighting potential conflict between analysis for individual learners and organizational stakeholders.

Ga? Evi ?, Dawson, and Siemens argue that the computational aspect of the learning analysis needs to be tied to existing educational research if the field of instructional analysis is to fulfill its promise to understand and optimize learning.

Maps Learning analytics

Distinguishing learning analytics and education data mining

Distinguishing the field of educational data mining (EDM) and analytic learning (LA) has been the concern of some researchers. George Siemens took the position that the mining of educational data includes analytical learning and academic analysis, previously intended for governments, funding agencies, and administrators, not learners and faculty. Baepler and Murdoch define academic analytics as an area that... "combines selected institutional data, statistical analysis, and predictive modeling to create intelligence in which learners, instructors, or administrators can change academic behavior". They continue to try to discriminate the mining of educational data from academic analysis based on whether the process is driven hypotheses or not, although Brooks questions whether this difference exists in the literature. Brooks instead proposes that the better distinction between the EDM and LA communities is the root from which each community originated, with authors in the EDM community dominated by researchers who came from the intelligent teaching paradigm, and studied anaytik researchers becoming more focused on corporate learning systems (eg learning content management system).

Regardless of the differences between LA and EDM communities, both areas have significant overlap both in the investigator's aims and in the methods and techniques used in the investigation. In the MS program offerings in analytical study at Teachers College, Columbia University, students are taught methods of EDM and LA.

src: www.learningservices.is.ed.ac.uk

History

Context of learning analysis

In "The State of Learning Analytics in 2012: Future Reviews and Challenges" Rebecca Ferguson tracks the progress of analysis for learning as development through:

1. An increase in interest in large data for business intelligence
2. The rise of online education is focused around virtual learning environments (VLEs), content management systems (CMS), and management information systems (MIS) for education, which see an increase in digital data on student backgrounds (often held in MIS) and learning data log (from VLEs). This development provides an opportunity to apply business intelligence techniques to educational data
3. Questions related to system optimization to support learning especially given the question of how we can see if a student is involved/understanding if we can not see it?
4. Increase focus on proof of progress and professional standards for accountability systems
5. This focus leads to teacher ownership in analytics - given that they are related to the accountability system
6. Thus, increased emphasis is placed on the pedagogical learning analytic skills
7. This pressure is increasing because of the economic desire to increase engagement in online education for high quality affordable education release

History of techniques and analytical methods of learning

In a discussion of analytic history, Cooper highlighted a number of communities from which analysis of learning attracted techniques, including:

1. Statistics, which is an established means of addressing hypothesis testing.
2. Business intelligence, which has a resemblance to learning analysis, although has historically been targeted to make report production more efficient through allowing data access and summarizing performance indicators.
3. Web analytics, tools like Google analytics reports on web page visits and references to other websites, brands, and key codes on the internet. The more "fine grains" of this technique can be adopted in analytic learning for the exploration of student passes through learning resources (courses, materials, etc.).
4. Operational research, which aims at highlighting design optimization to maximize goals through the use of mathematical models and statistical methods. The technique is involved in studying analytics that seek to create real-world behavior models for practical applications.
5. Artificial intelligence and data mining, machine learning techniques built on data mining and AI methods are able to detect patterns in data. In studying the analysis, such techniques can be used for intelligent guidance systems, the classification of students in a more dynamic way than simple demographic factors, and resources such as a "suggested course" system modeled on collaborative screening techniques.
6. Social network analysis (SNA), which analyzes the relationships between people by exploring implicit (eg interaction in forums) and explicitly (eg "friends" or "followers") bundles them online and offline. SNA is developed from the work of sociologists such as Wellman and Watts, and mathematicians such as Barabasi and Strogatz. The work of these people has given us a good sense of the pattern shown by the network (the small world, the law of force), the attributes of connections (in the early 70s, Granovetter explores connections from the perspective of binding power and impacts on new information), and the social dimension of the network (for example, geography is still important in the world of digital networks). It is primarily used for exploring groups of networks, affecting tissues, engagement and release, and has been used for this purpose in studying the analytic context.
7. Information visualization, which is an important step in many analyzes for sensing around the data provided, and is used in most techniques (including the ones above).

Learning analysis in higher education

The first graduate program focused specifically on analytics learning was created by Ryan S. Baker and launched in the fall semester 2015 at Teachers College, Columbia University. The program description states it

data about learning and learners are being made today on an unprecedented scale. The field of learning analysis (LA) and data mining excavation (EDM) has emerged with the goal of turning this data into new insights that can benefit students, teachers and administrators. As one of the world's leading teaching and research institutions in education, psychology and health, we are proud to offer an innovative graduate curriculum dedicated to improving education through technology and data analysis. "

src: www.northeastern.edu

Analytical method

Methods for learning analysis include:

• Content analysis, especially student-created resources (such as essays).
• Discourse analysis, which aims to capture meaningful data on student interactions that (unlike social network analytics) aims to explore properties of the language used, not just the interaction network, or the number of forum posts, etc.
• Social learning analysis, which aims to explore the role of social interaction in learning, the importance of learning networks, the discourse used to create sensations, etc.
• Disposition analysis, which seeks to capture data on student dispositions for their own learning, and this relationship with their learning. For example, "curious" participants may be more likely to ask questions, and these data can be captured and analyzed for analytic study.

src: mms.businesswire.com

Analytics results

Analytics has been used to:

• Predicted goals, for example to identify "risky" students in terms of drop out or course failure
• Personalize & amp; adaptation, to provide students with customized learning paths, or assessment materials
• Destination intervention, providing educators with information to intervene to support students
• Visualization of information, usually in the form of so-called learning dashboards that provide summary learning data through data visualization tools

src: i.pinimg.com

Software

Most software currently used to study analytics doubles the functionality of web analytics software, but applies it to student interactions with content. Social network analysis tools are usually used to map social connections and discussions. Some examples of learning analysis software tools include:

• BEESTAR INSIGHTS: real-time systems that automatically collect student engagement and attendance, and provide analytics tools and dashboards for students, teachers and management
• LOCO-Analyst: a context-conscious learning tool for analytic learning processes that takes place in a web-based learning environment
• SAM: Student Activity Monitor intended for a private learning environment
• SNAPP: a learning analysis tool that visualizes the interaction network generated from discussion forum posts and replies
• Solutionpath StREAM: A leading UK-based real-time system that utilizes predictive models to determine all aspects of student engagement using structured and unstructured sources for all institutional roles
• Student Success System: a predictive learning analysis tool that predicts student performance and plots students into a risk quadrant based on performance engagement and predictions, and provides indicators to develop an understanding of why learners are not on track through visualization such as interaction networks resulting from involvement social (eg post discussion and replies), performance on appraisal, engagement with content, and other indicators

src: www.adelaide.edu.au

Ethics and privacy

The ethics of data collection, analytics, reporting, and accountability have been raised as a potential concern for analytic study, with concerns arising about:

• Ownership of data
• Communication around the scope and role of learning analytics
• An important role of human feedback and error correction in learning the analysis system
• Share data between systems, organizations, and stakeholders
• Believe in the data client

As Kay, Kom and Oppenheim say, the wide range of data, potentially derived from:

• Recorded activity: student records, attendance, assignments, researcher information (CRIS)
• System interactions: VLE, ​​â € <â €
• Feedback mechanism: survey, customer service
• External systems that offer reliable identification such as shared sector and services and social networking

Thus challenging and diverse legal and ethical situations from country to country, increase the implication to:

• Various data: principles of collection, retention, and exploitation
• Educational missions: fundamental issues of learning management, including social engineering and performance
• Motivation for analytic development: mutuality, a combination of corporate, individual and general goodness
• Customer expectations: effective business practices, social data expectations, cultural considerations from a global customer base.
• Obligation to act: maintenance tasks arising from the knowledge and consequences of student and employee performance management challenges

In some prominent cases such as the inBloom disaster, even full functional systems have been shut down due to a lack of confidence in data collection by governments, stakeholders and civil rights groups. Since then, the learning analysis community has studied extensively the legal conditions in a series of expert workshops on "Ethics & amp; 4 Learning Analytics" which is the use of analytics of trusted learning. Drachsler & amp; Greller released an 8-point checklist named DELICATE based on intensive studies in this area to uncover ethics and privacy discussions surrounding learning analysis.

1. D-etermination: Determine the learning analysis objectives for your institution.
2. E-xplain: Specify the scope for data collection and usage.
3. L-egitimate: Describe how you operate within the legal framework, see important laws.
4. I-nvolve: Talk to stakeholders and provide assurance about the distribution and use of data.
5. C-onsent: Seek approval through clear approval questions.
6. A-nonymise: De-identify as many individuals as possible
7. T-echnical Aspects: Monitor who has access to data, especially in areas with high staff turnover.
8. External partners: Ensure externals provide the highest data security standards

It demonstrates ways to design and provide an analytical adjustment of privacy learning that can benefit all stakeholders. The full DELICATE checklist is publicly available.

src: i.ytimg.com

Open learning analysis

Chatti, Muslim and Schroeder note that the goal of open-ended analytical learning (OLA) is to improve the effectiveness of learning in a lifelong learning environment. The authors refer to OLA as an ongoing analysis process that encompasses diversity across all four dimensions of the instructional reference model of learning.

src: www.disruptordaily.com

See also

• Odds algorithm
• Pattern recognition
• Predictive analysis
• Text analysis

src: gracesystems.nl

Further reading

For the introduction of a common audience, see:

• Educational instruction initiative instruction (2011)
• Educause review on analytic learning (2011)
• UNESCO studied a brief analytics policy (2012)

src: www.e-learningmatters.com

Note

References

External links

• Society for Research Learning Analysis (SoLAR) - research network for learning analysis
• US Department of Education's Report on Analytics Lessons. 2012
• Learn Google Analytics Group with discussions from researchers and individuals interested in the topic.
• International Conference Analysis & amp; Knowledge
• Learn Analytics and Education Data Collection conferences and people
• Next Gen Learning Definition
• Microsoft Education Analytics with information on how to use data to support better educational results.
• Education Data Excavation
• Educate resources to learn about analytics
• Learn infographic analytics

Source of the article : Wikipedia