際際滷

際際滷Share a Scribd company logo

Student Success

A
Adriana Wilde

This document discusses using learning analytics to study student success in face-to-face instruction and MOOCs. It first compares the models of traditional instruction and MOOCs, noting both differences and similarities. It then proposes that learning analytics could uncover further variances and commonalities by analyzing data on student characteristics, activities, and performance. Several research questions are posed about predicting outcomes and the effects of course components. Factors influencing student achievement are reviewed for different datasets. The conclusion maintains that learning analytics can further the understanding of the overarching principles of student success that apply across models.

1 of 22
Download to read offline
Student Success on Face-to-Face Instruction and MOOCs What can Learning Analytics uncover? @AdrianaGWilde / Ed Zaluska / Dave Millard (@hoosfoos) 29 June 2015 Web Science Education 2015 #wseducation Web And Internet Science Group
Why is it important to study student success in Higher Education? How is the model of face-to- face instruction different/similar to that in MOOCs? How can learning analytics uncover further differences/similarities? 2 @AdrianaGWilde #websci2015
3 @AdrianaGWilde #websci2015
Student success in Higher Education Stakeholders interest: Personal dimension Institutional Society as a whole! 4 @AdrianaGWilde #websci2015
5 F2F instruction vs MOOCs There are some obvious differences but are these fundamental to the way students learn? What are the commonalities? Koller, D. (2012) What were learning from online education @AdrianaGWilde #websci2015
6 Analysis of data to extract characteristics of students and learnin activity that could be used for understanding of performance, prediction of achievement, and timely interventions. Ferguson (2012) Learning Analytics: drivers, developments and challenges @AdrianaGWilde #websc
Source: www.slideshare.net/sbs/designing-systemic-learning-analytics-at-the-open-university Predictive modelling of student outcomes 7 @AdrianaGWilde #websci2015
8 While all sectors have barriers to capture value from the use of big data [these] are higher for education, because of a lack of data-driven mind-set and available data The challenge! McKinsey Global Institute report (2011) @AdrianaGWilde #websci2015
The Data @AdrianaGWilde #websci2015
The MOOC dataset Enrolments (learner_id,enrolled_at,unenrolled_at) 580787 entries Comments (id,author_id,parent_id,step,text,timestamp,moderated,likes) 145,425 entries Question Response (learner_id,quiz_question,response,submitted_at,correct) 0 entries Step Activity (learner_id,step,first_visited_at,last_completed_at) 468,634 entries 468,518 entries @AdrianaGWilde #websci2015
11 Step Activity Augmenting data in preparation for data mining @AdrianaGWilde #websci2015
Curricular mesh for Civil Engineering in Computing Source: http://escuela.ing.uchile.cl/docencia/ Mallas_Especialidades/COMPUTACION Description as list: http://www.uchile.cl/carreras/4971/i ngenieria-civil-en-computacion @AdrianaGWilde #websci2015
13 Curricular mesh (Plan com炭n) Source: http://escuela.ing.uchile.cl/docencia/Mallas_Especialidades/COMPUTACION @AdrianaGWilde #websci2015
14 The UCh Datasets 221 attributes, including: Socioeconomic data (e.g. parental level of instruction) Previous education attainment (e.g PSU results) Current performance (e.g. marks in intermediate assessments) Students data-trail of their interactions with the mobile version of their VLE, U-Cursos @AdrianaGWilde #websci2015
UCV data 15 @AdrianaGWilde #websci2015
Research Questions @AdrianaGWilde #websci2015
Who elects to do this MOOC? employment 17 Research Questions (on MOOC data) What are the predictors of participants completion? Does the composition of various types of MOOC activities have a measurable effect on participants completion? What type of activity students complete/engage on the best? Does the perceived difficulty of activities have a measurable effect on completion? audios discussions articles exercises videos age education gender confusion boredom happiness frustration difficulty @AdrianaGWilde #websci2015
parents education private/state schooling employment 18 Research Questions (on F2F instruction) What are the predictors of students retention and completion? What type of modules students complete/engage on the best? Does the perceived difficulty of modules have a measurable effect on completion? physics engineering computing chemistry mathematics age PSU scores Initial demographics enrolment rate semester area failure rate difficulty Registration type place of origin gender education @AdrianaGWilde #websci2015
19 What drives student success? @AdrianaGWilde #websci2015
Factor affecting student success In FL data In UCV data In UCh data Disability ? Age 賊 Gender 賊 Previous education 賊 Ethnic group 賊 ? Socio-economic background Subject studied Module assessment strategy 賊 Financial issues 賊 Employer support Personal life events Workload Early engagement Language ability Advice received on course choice 賊 Study calendar/scheduling Peer support and belonging 賊 Family support 賊 20 @AdrianaGWilde #websci2015
Conclusions Student success is important not just to the educational stakeholders but for society as a whole. F2F instruction is different to MOOCs but we hypothesize that there is an overarching common principle manifesting itself in different ways. Learning analytics methods, such as the application of data- mining algorithms on pre-processed data about learners, dealing with the dimensionality of the space to produce understanding diagnosis and predictions. @AdrianaGWilde #websci2015
References and attributions Literature: Pascarella and Terenzini (2005) How college affects students: A third decade of research Jones (2008) Student retention and success: A synthesis of research Lee and Choi (2011) A review of online course dropout research: implications for practice and future research Leony (2014) Rule-based detection of emotions in the Khan Academy Platform. Ferguson (2012) Learning Analytics: drivers, developments and challenges Images: 22 Le Ch辿鱈le Leixlip 5KM Road Race 2014 P. Mooney Mirando estrellas S. Rivas Diagrama CONEST S. Rivas & J. Zambrano Take Away Sign N. Smale. Habits quote: www.matt-potts.com/habits-the-common-denominator-of-both-success-and-failure/ 際際滷s 5,8: www.slideshare.net/sbs/designing-systemic-learning-analytics-at-the-open-university Institutional logos for FutureLearn, and all HEI mentioned Thanks! @AdrianaGWilde @hoosfoos #websci2015

More Related Content

Student Success

  • 1. Student Success on Face-to-Face Instruction and MOOCs What can Learning Analytics uncover? @AdrianaGWilde / Ed Zaluska / Dave Millard (@hoosfoos) 29 June 2015 Web Science Education 2015 #wseducation Web And Internet Science Group
  • 2. Why is it important to study student success in Higher Education? How is the model of face-to- face instruction different/similar to that in MOOCs? How can learning analytics uncover further differences/similarities? 2 @AdrianaGWilde #websci2015
  • 4. Student success in Higher Education Stakeholders interest: Personal dimension Institutional Society as a whole! 4 @AdrianaGWilde #websci2015
  • 5. 5 F2F instruction vs MOOCs There are some obvious differences but are these fundamental to the way students learn? What are the commonalities? Koller, D. (2012) What were learning from online education @AdrianaGWilde #websci2015
  • 6. 6 Analysis of data to extract characteristics of students and learnin activity that could be used for understanding of performance, prediction of achievement, and timely interventions. Ferguson (2012) Learning Analytics: drivers, developments and challenges @AdrianaGWilde #websc
  • 8. 8 While all sectors have barriers to capture value from the use of big data [these] are higher for education, because of a lack of data-driven mind-set and available data The challenge! McKinsey Global Institute report (2011) @AdrianaGWilde #websci2015
  • 10. The MOOC dataset Enrolments (learner_id,enrolled_at,unenrolled_at) 580787 entries Comments (id,author_id,parent_id,step,text,timestamp,moderated,likes) 145,425 entries Question Response (learner_id,quiz_question,response,submitted_at,correct) 0 entries Step Activity (learner_id,step,first_visited_at,last_completed_at) 468,634 entries 468,518 entries @AdrianaGWilde #websci2015
  • 11. 11 Step Activity Augmenting data in preparation for data mining @AdrianaGWilde #websci2015
  • 12. Curricular mesh for Civil Engineering in Computing Source: http://escuela.ing.uchile.cl/docencia/ Mallas_Especialidades/COMPUTACION Description as list: http://www.uchile.cl/carreras/4971/i ngenieria-civil-en-computacion @AdrianaGWilde #websci2015
  • 13. 13 Curricular mesh (Plan com炭n) Source: http://escuela.ing.uchile.cl/docencia/Mallas_Especialidades/COMPUTACION @AdrianaGWilde #websci2015
  • 14. 14 The UCh Datasets 221 attributes, including: Socioeconomic data (e.g. parental level of instruction) Previous education attainment (e.g PSU results) Current performance (e.g. marks in intermediate assessments) Students data-trail of their interactions with the mobile version of their VLE, U-Cursos @AdrianaGWilde #websci2015
  • 17. Who elects to do this MOOC? employment 17 Research Questions (on MOOC data) What are the predictors of participants completion? Does the composition of various types of MOOC activities have a measurable effect on participants completion? What type of activity students complete/engage on the best? Does the perceived difficulty of activities have a measurable effect on completion? audios discussions articles exercises videos age education gender confusion boredom happiness frustration difficulty @AdrianaGWilde #websci2015
  • 18. parents education private/state schooling employment 18 Research Questions (on F2F instruction) What are the predictors of students retention and completion? What type of modules students complete/engage on the best? Does the perceived difficulty of modules have a measurable effect on completion? physics engineering computing chemistry mathematics age PSU scores Initial demographics enrolment rate semester area failure rate difficulty Registration type place of origin gender education @AdrianaGWilde #websci2015
  • 19. 19 What drives student success? @AdrianaGWilde #websci2015
  • 20. Factor affecting student success In FL data In UCV data In UCh data Disability ? Age 賊 Gender 賊 Previous education 賊 Ethnic group 賊 ? Socio-economic background Subject studied Module assessment strategy 賊 Financial issues 賊 Employer support Personal life events Workload Early engagement Language ability Advice received on course choice 賊 Study calendar/scheduling Peer support and belonging 賊 Family support 賊 20 @AdrianaGWilde #websci2015
  • 21. Conclusions Student success is important not just to the educational stakeholders but for society as a whole. F2F instruction is different to MOOCs but we hypothesize that there is an overarching common principle manifesting itself in different ways. Learning analytics methods, such as the application of data- mining algorithms on pre-processed data about learners, dealing with the dimensionality of the space to produce understanding diagnosis and predictions. @AdrianaGWilde #websci2015
  • 22. References and attributions Literature: Pascarella and Terenzini (2005) How college affects students: A third decade of research Jones (2008) Student retention and success: A synthesis of research Lee and Choi (2011) A review of online course dropout research: implications for practice and future research Leony (2014) Rule-based detection of emotions in the Khan Academy Platform. Ferguson (2012) Learning Analytics: drivers, developments and challenges Images: 22 Le Ch辿鱈le Leixlip 5KM Road Race 2014 P. Mooney Mirando estrellas S. Rivas Diagrama CONEST S. Rivas & J. Zambrano Take Away Sign N. Smale. Habits quote: www.matt-potts.com/habits-the-common-denominator-of-both-success-and-failure/ 際際滷s 5,8: www.slideshare.net/sbs/designing-systemic-learning-analytics-at-the-open-university Institutional logos for FutureLearn, and all HEI mentioned Thanks! @AdrianaGWilde @hoosfoos #websci2015