Terminology  

LEARNING ANALYSIS FOCUS 

… decisions about the use of data approaches in Education must be informed by a scientific understanding of the impact of learning analysis on learners, teachers, institutions, and society.   

 

 

Sciences of Learning & Pedagogical Knowledge

 

A key issue in learning analysis is to actually think through what data you need, and what you want to do with the data:

the sciences of learning together with pedagogical knowledge might be able to help.

Sciences of Learning + Pedagogical Knowledge  = Learning Sciences  
Learning Analysis   OR   Learning Analytics?

 

Learning Analysis: the role of data and data analysis for learning, teaching, and education

 

Three distinct, but overlapping fields that address the role of data and data analysis

  • Educational data mining (EDM)

  • Learning analytics and knowledge (LAK)

  • Big Data

 

 

Educational Data Mining (EDM)

  • Intelligent data mining

  • roots in Artificial Intelligence in Education & Intelligent Tutoring Systems research, as far back as the 1970s

  • applies computational approaches such as data mining, machine learning classification, clustering, Bayesian modelling, relationship mining, discovery with models, statistics, and visualisation to information generated in educational settings to better understand students and the settings in which they learn

 

Learning Analytics and Knowledge (LAK)

  • Emerging research field and design discipline

  • LA is a set of data generation and analysis techniques and tools that may be utilised to gain a deep understanding of profound questions for research, policy and practice, generated by 21st Century learning and skills development

  • LAK facilitates a clear theoretical understanding of what is learning, how we assess it, how we foster it, and how we operationalise it in productive educational practices, teaching and learning environments

 

Big Data in Education

  • Refers to large amounts of data produced very quickly by a high number of diverse sources

  • Data generated by people (e.g., computer logs, an essay) or generated by technology (e.g., sensor readings, photos, videos, GPS signals, etc.)

  • The analysis of large data sets generated in educational context could identify and validate patterns cross institutions, regions and countries, which benefit the different levels of stakeholders in education systems  (predictive analytics)

Is there “big data”,
or rather just “small data” in the  educational sector ???
 
Learner-centric VS Learning-centric analytics        

 

SLATE adopts Zach Stein's distinction between learner-centric and learning-centric analytics (Stein 2012).

 

Learner-centric analytics measures student behaviour in technological environments

  • Learner engagement measured through the number of times a student visits learning materials, logs on an LMS, how long they view a flipped classroom video

  • Give input on design of learning environments, learning material, etc.

 

Learning-centric analytics has to do with conceptual growth and requires examining student artefacts to detect conceptual acquisition (Stein 2012)

  • focus is on "learning", "learning outcomes"

  • have to examine artefacts that students develop to identify if learning has taken place.  

  • one’s understanding of learning, impacts the analytics design