For my A3 assignment I stepped outside of my conform zone and learned about predictive algorithms in education. Based on my understanding and personal experience with at-risk youth I “created” an app to fill the gap.
My website for the App and background information can be found here: https://misspaslawski.wixsite.com/passapp
Hi Katlyn,
Good to read about something relevant to your own work — the “dilemma”!
There are definitely ethics concerns with something like this, but I think the potential benefits could be great as well. Analyzing student performance in this way could reveal which areas they need to work on more, revealing new and customized pathways to improvement.
Hi Katlyn,
Great app by the way! I must confess that I’m no expert on the subject but you were very concise and clear about the use and functions of your app. I like that you not only provided the research to back up your app but also told us why this issue was important to you. I could see this app working not only in Canada but also in different countries. Obviously it might need some tweaks to adapt to diverse educational settings and social contexts but the main idea would still work.
The issue of deciding whether students should repeat a year or join an alternative program made me think about my own educational system back home. Unlike your school, we don’t have an alternative program for slow learners or repeaters. If students can’t pass their exam, they are automatically repeated. I feel that my education system is doing a disservice to thousands of kids every year. It seems that only the bright and gifted are able to proceed while the rest are held back or kicked out of school. In the Mauritian context, your app would be more useful if they could send early warnings about students who are not performing to a certain level. I’d much rather have a proactive system where interventions could take place to help students improve and be more involved in their own learning journey.
Nonetheless, I think it’s a brilliant idea and combining different factors such as internal, external and psychological factors enables us to look at a student’s performance from different angles.
Hi Pascaline,
Thanks for your comments. It sounds like you are thinking of an early warning system kind of detection for students. These kind of predictive algorithms do currently exist in school. I shared some thoughts about it with Matt below. https://blogs.ubc.ca/etec523/2020/07/31/a3-predictive-algorithm-for-student-success-app/#comment-2432
Hi Katlyn,
Well done with a streamlined, visually appealing presentation on a very interesting topic. I agree with previous comments about ethics and was happy to see that you touched on the topics of informed consent and bias in algorithms. It’s becoming clear that algorithms will reflect the biases of the maker and that it takes a lot of work in for that not to happen. Demographic information can be informative but it’s important that it not be used to stereotype or pigeon-hole students.
Hi Megan,
I agree completely! I can see a potential for algorithms and a lot of work needed to get them to the right place as you mentioned. In my experience I have seen great bias towards students in these situations regardless of a computer algorithm. I would hope that with an algorithm humans may be able to push aside variables that they usually weigh on heavily for a more just decision.
I really see a lot of value in an app such as this Katlyn and I can see it being used in other educational situations in addition to students retaking courses. One struggle I have had in the past in Alberta was with streaming students from grade 9 to grade 10, as some courses could have 3 different options in grade 10 and parents can be quite upset if they think that their child is being placed in the wrong stream, however something like this could really assist the teacher with that placement. It is quite unfortunate when a student is suggested to go into one stream, but the parents argue and the student gets placed in a higher stream, only to fail that course. In a situation like this, your app could be used to support the teacher’s decision. I like that you touched on some of the concerns, I could see some of the concerns parents could have, but I do think there could be ways to have anonymity within the app, such as a name not being associated with the student, but a number, just to take out that concern.
Hi Micheal,
You bring up an interesting aspect that I had yet to consider. I have also faced the issue of student placement in streamed courses and disagreements with parents. Not related to the PASS app but my school found course selection in grade 10 (specially Math) easier when we transitioned grade 9 math to being year long. This allowed more time for students to gain and understanding of where they were at and allowed for parents to get a better sense as well which resulted in less discrepancies at course selection time.
This is an excellent forecast, presented from a professional perspective and your own personal experience which makes it very powerful. I agree that the future of artificial intelligence is to augment and not replace human decision making and I think you made the argument very clear in your site.
While the concern about a machine making decisions about someone’s future is a very valid one, and one we should be aware of, I think that the system you proposed has the potential to be more objective and provide information to support decision making (not making decisions itself). Perhaps a panel/review committee would be assigned to consider and evaluate the suggestions given by this program.
You made excellent use of contrast and varying colours which made me remember its emphasis during the mobile usability OER, however one piece of constructive criticism is that I found the colours used in the site a little bit challenging to read with some of the light colours on white and the purples on red, I’m red-green colourblind so perhaps that was it.
I felt your focus on the fact that too often interventions come too late was significant, and I wonder if by using techniques such as data mining this system could even serve as a kind of “early warning” system and provide suggestions for interventions that could happen within the first course being failed?
Thank you for a thought provoking presentation.
Hi Matt,
I like the idea of indicating a review panel, that is a more clear way to stat that the algorithm doesn’t make the final decision.
Thanks for your feedback on the colours, that is definitely something I should have taken into consideration!
What you are suggestion with the early warning system does exist in some school districts although it is unclear where. Many publications are from Computer Science Journals conducting studies on accuracy of the predictive algorithms as an early warning system. I would like to know how prevalent the early warning system algorithms are being used.
Hi Katlyn,
Data mining is a topic that recently caught my interest too so I am glad to see how you incorporate that into your project. As someone who also works with at-risk students, I can see the potential of an app like P.A.S.S. It is especially helpful for schools that are lacking resources, whether it’s funding or shortage of administrators/teachers to spend significant amount of time to have one on one time with the identified students.
Like others have mentioned, ethics is a sensitive topics. Another thing that came to my mind is how reliable is the collected student data. Some of the external and psychological variables are not set in stone or somehow the data that was provided to school didn’t reflect the actual situation. Something came to my mind was engage time as students (or simply put, it’s natural human tendency) tend to report something they think would represent themselves better. Those things can impact the algorithm but are hard to verify. Overall, I like this P.A.S.S. app a lot and can see the potential to use it even in the post-secondary education setting. Thanks for sharing this project with us!
Hi Pei,
Thanks for your feedback. I completely agree that student data may be difficult to collect and sometimes incorrect. Even now I often find parent contact information to be out of date as it is usually inputted into the system when the student enters high school and is only updated if the parent calls in to inform the school. In order for the PASS algorithm to work there would need to be a development in student data.
HI Katlyn, this is new territory for me and am quite fascinated with AI and machine learning myself.
Your app is something to consider into production as I think it would indeed evolve into something useful not just for students, but I’m thinking also revolutionizing other assessments for companies and HR in general. I do see the downsides of student data collection, bias and ethics that will hinder some adoption. However, all technologies today still need the human interaction to balance the quantitative analysis that we see from Algorithms. We would still need counsellors and teachers to provide face to face experiences for students and perhaps use the app as supplementary to see how progress is made.
I have worked with EDM at my institution and mostly for staff development. I can see your app working for those with lower skills that can upgrade through PD, and recommendations put forth. Great job on presentation and color scheme.
Hi Luke,
Thanks for your comments. I glad you can see this application being put forth into other avenues, I think the future of algorithms has so much space to develop!
Hi Katlyn,
I agree this could be very useful but also wondered about the ethics of an algorithm when looking at students backgrounds/race/sex etc. Although that data is completely relevant, wether students would want to be polled against the statistics is another issue. That being said administrators are faced with making difficult decisions that they hope will help the students as best they can moving forward. Anything that can help that decision making problem could be seen very helpful. I also like how you added its not a one size fits all program and there will always be considerations in place.
I like the design and flow of the webpage and thought the content fit really nicely. I liked your font, colours and pictures.
Great job!!
Hi Lindsay,
Absolutely! Ethics is a huge component of algorithms and has led to serious issues, especially in the criminal justice system. I listened to a podcast yesterday from The Daily about a facial recognition algorithm incorrectly identify an innocent man as the criminal, however the it raised the issue of the officers who arrested him based off of the facial recognition algorithm solely. Goes to show that algorithms need to be improved, but humans also need to be the final deciding factors based on information providing and must be aware of the biases.
Thanks 🙂
Wow, Katlyn. This forecast has given me a ton to think about. Thank you for sharing this issue and shedding light on the problem of students falling through the cracks in the school system. I also have experience in working in the alternative school system. As you know, a lot of the success comes from building the strong connection between teacher and student. They need to know that you are there to support them and that you know they can succeed.
I hope to discuss this with you more. Can I ask what would happen if a student is unsuccessful in a core course, such a Math and the P.A.S.S system says they would be unsuccessful a second time. What are the recommendations for that student? Is this discussed with the school based team? Are the recommendations a different method of delivery of the course?
Hi Kristin,
Thanks for your great questions! That helps me consider how I can add more detail to my site to make it more clear 🙂
I envision a scenario where a student is not successful in a course such as Math and the PASS app providing a recommendation to admin indicating if they would be a) likely successful in re-taking the course or b) unsuccessful in re-taking the course in regular programming. Then the admin, along with teachers, student services and parents would discuss options which could include being placed into the alternative program for Math.
Does that answer your questions?
Hi Katlyn,
That does answer my questions, thanks! I assume one option would be to recommend online or an alternative face-to-face program. I see this a lot with administration telling the student that since they were unsuccessful face-to-face, they must now take it online. This is usually the incorrect recommendation because many students are unsuccessful because of a lack of motivation, not a lack of ability. If the student cannot be bothered to come to a class at a scheduled time, how are they expected to sign into their computer on (what they consider) their own time? To be an online student there needs to be a reliable support system in place for vulnerable students. More often than not, it is better to have a student go to an alternative program where they can be held accountable. I think this app would help clear up the confusion about online and alternative options.
I love a good acronym, and naming is everything! Well done on that front. I really like how fleshed out your app is and it’s totally something I could see being useful. The principal at my school says that our role as educators is to help students reach their full potential. And as you said on your site, one size doesn’t fit all, and that’s true of what to do with a student who doesn’t succeed in a course and with what success – the reaching of one’s potential – looks like.