SAMR and AI Chatbots

In case you didn't already know, the SAMR model, developed by Dr. Ruben Puentedura in 2010, is a popular framework for integrating #technology effectively in #education.

Original SAMR model (2010) from http://hippasus.com/blog

Consisting of four levels: Substitution, Augmentation, Modification, and Redefinition, the SAMR model is intended to help #teachers create more engaging and meaningful educational activities as well as foster critical thinking, creativity, and collaboration among #students.

However, I would argue that interpretations and applications of the model often fail to recognise the role that students play in technology use and attempts to integrate technology in any #education practice. You can see how my own #research shows the power of student involvement in tech integration here. There are also many in academic circles who argue that the SAMR model has little academic rigour and favour #TPACK but I will put that argument aside for now.

The significance here, is that the SAMR model has offered an easy to understand, easy to use and easy to follow process for technology integration for many in education:

The SAMR model can be easily adapted and interpreted in multiple ways, helping teachers reflect as technology is used to achieve specific outcomes. As a planning tool, it enables teachers to design, develop, and infuse digital learning experiences that use technology.

(https://research.com/education/how-to-use-samr-model-in-designing-instruction#2)

However, one of the major issues still faced by #teachers at all levels and in all sectors of education focuses on #assessments. Their #design, the capacity of #students to plagiarise (although best not get me started on what that actually means) and most importantly, what can be done to address the pressures of rapid and very disruptive technology development.‍‍‍‍‍‍‍

So, as we are now light years (7 months) into our #aieducation journey dominated by #chatGPT, I thought it was time to offer my own interpretation of the SAMR model focusing on assessment, based on lived experience in my role as Leader of Digital Technologies across an R-12 school and through my involvement in the network of #aieducators that has blossomed in that period. This is what I think the SAMR model looks like when we consider it through the lens of #AI #chatbots:

SAMR Chatbots model (Jackson, N, 2023)

While recognising that others have already attempted reinterpretations of the SAMR model incorporating AI such as here, I don't recall seeing one that focuses on Chatbots or on assessment specifically.

This model challenges early parts of the structure, suggesting that SUBSTITUTION actually comes from students not teachers. This new model adds another phase to the existing model, REALISATION. I feel this important to recognise. Students at the early stages of technology integration or where little/no technology integration that addresses the potential use of chatbots, have the capacity to SUBSTITUTE their own work with that produced by a chatbot.

The remainder of the model, provides the processes by which assessments can be changed and improved. The improvement incorporates both the essence of technology integration and improvements in the quality of assessment. This culminates in the REDEFINITION stage where very few educators are at right now. However, that is exactly what we should be aspiring and collaborating to achieve. This is the stage at which the power of #aitechnology gets realised in transforming assessment and education. ‍‍

Having ridden the #aieducation wave thus far, I have witnessed first hand and can see from the posts, tweets, through conversations, webinars, podcasts, keynotes, workshops with other educators in a plethora of settings, that this model represents a decent interpretation of what is happening and what needs to happen to address assessment issues. However, I welcome feedback to further improve or amend the model.

As Sam Altman says in the podcast with Lex Friedman, the #Openai approach is:

iterative deployment and iterative discovery

I think that's a pretty fine approach so I will emulate that with this model.