This month's ATAIN presentation.
Deciding, not just adopting: how early-adopter
educators navigate generative AI
Karl McGuirk, University of Auckland
Abstract: When generative AI arrived, there was no shared
playbook. Educators made decisions about it largely on their own, in the gaps
between policy, disciplinary norms and fast-changing tools. This session shares
early findings from interviews with early-adopter educators across all
faculties of a large research-intensive university, and unpacks a preliminary
multi-stage framework for understanding how those decisions unfold.
Early analysis suggests the stages work less as
a fixed sequence than as a shared vocabulary, that disciplinary context shapes
decisions as much as technical confidence does, and that educators in different
areas were independently building similar local guardrails. We’ll explore what
this means for how networks like ATAIN, and the institutions we work in, can
better support educators to make context-sensitive decisions together.
Notes taken below:
Began with asking participants to think about 'the last Gen AI decision made - around teaching or practice'. Was it a linear or messy process. Did you go through - awareness/exposure, initial sense-making, evaluation and feasibility, design and implementation, and reflection and adaptation (the 5 stage framework)..
Limitations - working patterns, not conclusions, one instituion and early adopters. Was an insider (shared language, trust, context) and outsider (a 'researcher' not colleague)
A Master of Education project. 13 semi-structured interview with reflective thematic analysis. wanted to find out how early adopters use of Gen AI aligned (or not) to the 'use of Gen AI process'.
Shared how participants had experiences which align with the five stage framework.
Framework was based on 3 theoretical lenses at practitioner level (reflection (Schon), thinking in an don action, beyond technical rationality) through disciplinary integrit (TPACK, wht context, the psycnological other), and instituional systems (diffusion of innovation (Rogers, 2003), what is visible, tirable and compatible, decentralised diffusion, bottom up reinvention overcoming top down policy).
The model was presented to the participants and their perspectives included - acceptance, extension and/or reshaping/contesting. Summarised the various ways participants interpreted the framework and the additions/alternatives proffered.
Tensions also presented and discussed. Some regarded policy as an roadblock, others as a challenge to overcome. Localise guardrails were generally ineffective.
Tentative patterns coming through include:
The influence of disciplinary & professional knowledge on pedagogical decision-making
- educators in design, IT and engineering reframe the model's trailability phase using interative, non-linear workflows from their field.
- Educators in humanities, law and philosophy deploy disciplinary risk-aversion to constrain the model's stages and defend authentic human cogntion.
- Those with administrative or faculty leadership roles extend the individual decision-making models into formal governance in review cycles.
Explanatory power - technical confidence vs disciplinary vs contextual knowledge.
- technical condistions did not predict uncritical acceptance or uniform adoption
- disciplinary epiistemologies dictate procedural reshaping
- contextual constrains bounded feasability.
Therefore, working to reframe through 'from staircase to foxtrot'. Presented a model that reflects the perspectives of early adopters. Therefore the stages are a vocabulary, not a sequence.
Shared the micro, meso and macro levels on how the framework may be applied to inform practice going forward.
Q & A followed.
No comments:
Post a Comment