Data, Information, and Decision Making
Text translated in by Simon Massicote-Côté
What are artificial intelligence (AI), analytics, and other digital technologies with regards to your data? That is the question that Ph.D. candidate Helene-Sarah Becotte decided to answer during her webinar called “Comprendre l’intelligence artificielle, l’analytique et d’autres technologies” on February 2, 2021.
Helene-Sarah Becotte, Ph.D. candidate in applied mathematics, is known for her promotion of the use of mathematics on her blog, helenebecotte.com. Her objective is to allow internet users to “optimize their lives” through the prism of sciences and mathematics. The blogger, a consultant and one of the founders of the startup GRISDD, started by creating videos on a variety of topics and linking them to mathematical concepts. She also offers guidance to businesses, notably in decision making, by identifying and realizing digital intelligence projects. She calls it the mathematization of decisions.
What I loved about the webinar was the approach and the dynamism of Helene-Sarah Becotte who made me understand, in under an hour, the ever-evolving world of data. The analogy that she makes with the flour, the bread, and the toast, to define the correlation between data, the treatment of information, and decision making played a large role in helping me understand. I then decided to create a picture of that analogy, applying it to the context of education. The following diagram illustrates the data-management mechanism in a digital learning environment.
Source: Florence S Muratet via Canva
We can see in this picture that the flour represents raw data, the bread represents the treatment of the information, and the toast illustrates the contextualization of the data.
One day, a teacher was asking me about the use of the data that appeared on the teaching platform they were using for their course.
“There is data that appears on my Moodle interface, I don’t know what to do with it, it might be useful, but for what?”
The teacher asking this question is referring to the raw data (the flour). They correspond to the raw material. When we treat the data to obtain information (the bread), we obtain a value that does not allow us to orient ourselves toward a decision. In the image above, the percentage of students who have consulted the contents offered on a DLE is informative, but we do not know what to do with that information.
Digital learning environments are filled with raw data that feed tables with dates, numbers, names, or even diagrams, etc.
To get a clearer picture of that information data, the toast is a good representation of the process of contextualizing data that will then lead to decision making. What Helene-Sarah Becotte calls “the intelligence of data”, in the analogy, would be the percentage of students who have consulted the contents of the DLE in comparison with the previous weeks, for example.
Source: Florence S Muratet par Canva
We then come to the phenomenon of making pedagogical decisions based on data, which is a growing trend in the world of education. It is the process through which educators examine their evaluation data to identify the strengths and weaknesses of the students and use the results to adapt their practice (Mertler, 2007; Mertler & Zachel, 2006). This can certainly help teachers make more precise pedagogical decisions to favour academic success, for example.
That being said, I am a bit puzzled when it concerns the widespread use of this type of method from an institutional point of view because we could uncover some ethical concerns about the use of the data of the learners. Can a school use the private data of the students in order to identify risks? Do the learners have a word to say in the treatment of their private information? I have always found it difficult to understand why a school would want to know the job of the learner’s parents or their age. We reside in Quebec, but, having lived abroad, I have seen school forms that included questions about the ethnical background of the parents or even their religion. Add one tablespoon of data analysis and a pinch of predictive analysis from the AI and you have a potential recipe for disaster. The article from the RIRE on the possibilities and the concerns of the use of AI in education, published by Maryliz Racine, tackles that issue by making reference to some studies from the Conseil Superieur de L’Education (CSE):
“Since the AI develops from the data we feed it, there is a possibility that it will drift off, the CSE notes. It appears that the conception of the algorithms, incomplete data, or inexact input could lead to some problems in education. Those three elements can lead to a “cataloguing” of the students, a uniformization of educational paths, a standardization of the evaluation of learnings, just as much as it could lead to a reinforcement of conscious and unconscious biases, by the nature of the data used alone. This would then potentially have the effect of increasing the existing inequalities and the presence of discrimination or prejudice.” [Our translation.]
In sum, understanding data analysis made me aware of its potential for pedagogy: better management of learning, a better adaption of the contents, and a multitude of adjustments. In this regard, the second part of the Savoir project of the Centre de Transfert pour la réussite éducative du Québec (CTREQ) and the RIRE is enlightening about the way to approach data analysis in the context of education and offers support tools for decision making.
To learn more about Helene-Sarah Becotte’s presentation, please watch her webinar.