Analysis of Students Interactions on Virtual Learning Environment with QlikView
As Kofi Anna once said “Knowledge is power. Information is liberating. Education is the premise of progress, in every society, in every family.”
We are now living in a world where people can have easy access to information on the internet. Through platform such as YouTube, Udemy and Khan Academy, students and adults can view online courses, videos which enable them to gain a vast amount of knowledge. They can even pursue a university degree online at much lower cost and is available 24/7 at their own pace. However, it also takes commitment and determination to be able to complete and obtain the certification. Using the data from the Open university in United Kingdom (UK), we seek to find out factors that contribute to passing the course and detect signs of student who withdraw from the course in hope to find solutions that reduces the withdrawal rate and failing rate.
Using the QlikView dashboard, let’s answer the following question:
1) What is the passing or failing rate for each module?
2) What is the trend of score for each module assessment type?
3) Will students who are more active in the VLE modules i.e. higher number of clicks score better?
4) What is the academic performance of special needs students (declared disabled)?
5) Common signs and trends of students who withdrew from the modules
i. Trend of interaction with material over the duration of the course
ii. Demographics of students
Data Sources
Using 7 datasets from the Open University from 2013 to 2014 where the modules , student profile and their usage of the Virtual Learning Environment (VLE) have been anonymized, these anonymized Open University Learning Analytics Dataset (OULAD) can be found at https://analyse.kmi.open.ac.uk/open_dataset Kuzilek J., Hlosta M., Zdrahal Z. Open University Learning Analytics dataset Sci. Data 4:170171 doi: 10.1038/sdata.2017.171 (2017).
Data Specification and Definition
The definition of data as per its webpage at https://analyse.kmi.open.ac.uk/open_dataset
Pre-processing data
Datasets have been audited in Microsoft Excel and Qlikview to find existing and possible problems that we could face. Refer to table below on the problems identified and resolved with proposed solutions
The Dashboard
The Dashboard consist one cover page which links to the other 4 sheets.
Assessment Overview — Gives an overview of the % of results, whether the students pass/ fail or withdrawn at the high level and their average scores for assessment. Users can drill down to each module, results or disability for comparison.
Demographics — To analyse on the background of the students and their results. Demographics such as level of deprivation (low IMD band) , education level or region are used.
Activity — To analyse on the student participation in learning in the open university, whether the student have hand in their assignments and learn on the VLE (tracks by the no. of clicks) and also student withdraw in each month from the start of the course. The duration of the module ranges from 7–9 months. Note that there are students who take more than one module.
VLE Usage — A further drill down on the VLE Usage, show the relationship of interaction with VLE and scores. Another table show the total no. of clicks of passing student verse those who fail or withdrawn.
Here’s a video to show the interactions of using the dashboard. Please note that there is no audio.
Analysis & Implications
Answers to the 5 Business Questions:
1 ) What is the passing or failing rate for each module? (assessment overview sheet)
Breakdown of result % of all modules
We can future drill down to on each module to get the passing rate for each module
We can conclude that AAA has the highest passing rate of 72.98% and highest withdrawal is from CCC at 32.58 % and GGG has the highest fail rate at 27.29%.
2) What is the trend of score for each module assessment type? (assessment overview sheet)
Noticed that students score better in TMA and CMA assessment but there is drop exam score in all modules. This could be due to the different level of difficulty in the TMA,CMA assignment as compared to the exam. Teachers should review the difficulty and make them all more inline so that student can have a good estimation on the final exam difficulty.
3) Will students who are more active in the VLE modules i.e. higher number of clicks score better? (VLE usage sheet)
From this chart we can see that each student sum of clicks n their accumulated score from the assessments. Notice the outliners, dots that have exceptionally high scores /sum of clicks which is due to accumulation of score from 2 or more modules. like the example below
So a slider was add for the score
and we can observe that at the score of 35 (below 40 = fail) clicks was less than 3000. Also Score above 70 (pass) have more clicks up to 7000
This can lead us to conclude that students who are more active in the VLE will have higher scores.
4) What is the academic performance of special needs students (declared disabled)? (Assessment overview sheet)
Special needs students have a higher withdrawal rate 32.91% as compared to normal students of 24.29%, a difference of 8.62%. The normal students also have higher passing rate than students with disability. More information can be gathered the type of disabilities and find out how learning can be made easier to reduce the withdrawal rate.
5) Common signs and trends of students who withdrew from the modules
i. Trend of interaction with material over the duration of the course
We can see that the total no. of students who pass have 2 to 3 times more clicks than students who have withdrawn or fail, despise that only 52% of the students pass. We also noticed that student who fail & withdrawn have very similar pattern in their total clicks on the VLE. There is a significant drop in VLE for non-pass students at the 3rd month of study. We can help by sending reminders and encourage students to use the VLE more beginning of the 3rd months to help more student to pass.
Also noticed that the highest no. of withdrawal happen before the module starts at month 0. We should find out the reasons from these withdrawal through surveys and also re-consider if withdrawal before the start of the module should be considered in our studies as the high no. could be due to the refund policy. Eg full refund before starts of the module so student just decided not to take this module. If we start looking from the 1 month onwards at student who withdraw, it is very similar to chart total clicks on VLE per month.
ii. Demographics of students
Conclusion
The students start losing interest after the 2nd month can be seen from the VLE, withdrawal charts.
Low IMD, A level students should be given more help in their studies to help them to pass and stay in the program.
There is limitation cause by students who took more than one module which we need to separate out as they can cause outliners and more reason for withdrawn can be gather to better tackle the reasons.
Sources:
Dashboard Background image: https://www.maxpixel.net/Hand-Classroom-Board-Leave-School-Education-Slate-379214