doctoral research in progress.
pedagogical friction in the age of generative ai.
a mixed-methods collective instrumental case study examining how k–12 educators and recent learners navigate the friction-reducing affordances of generative ai.
extending ong's media ecology.
walter ong's developmental account of how communication technologies restructure consciousness provides the theoretical ground for this research. the dissertation proposes two extensions to ong's framework to account for shifts he did not anticipate — the algorithmic curation of human-created content, and the algorithmic generation of symbolic content itself.
- 01 ong
primary orality
pre-literate cultures in which knowledge is preserved through memory, repetition, and embodied performance.
- 02 ong
literacy
writing and print externalize memory, enabling analytical detachment, individual authorship, and abstract reasoning.
- 03 ong
secondary orality
broadcast media retrieve participatory and communal qualities of oral culture within a literate, scripted framework.
- 04 proposed extension
algorithmic secondary orality
humans continue to create symbolic content, but algorithms increasingly determine what content reaches which consciousness. the shared symbolic environment fragments into individualized streams.
- 05 proposed extension
tertiary algorithmicity
algorithmic systems both curate and generate symbolic content, rendering human authorship optional at scale. the condition under which contemporary students increasingly learn.
the national structural context.
analysis of nces school pulse panel data reveals a critical pedagogical infrastructure gap underpinning this media transition:
- device access parity: 88% of high-poverty schools and 89% of low-poverty schools report 1:1 device programs.
- policy gap: higher-poverty schools are 9% less likely to have a written ai policy (24% vs 33%).
- literacy gap: higher-poverty schools are 10% less likely to teach ai topics to students (39% vs 49%).
the pedagogical friction framework.
if the media environment increasingly defaults toward the bypass of cognitive labor, and if learning science establishes that cognitive labor is constitutive of durable understanding, then the educational question is one of design. pedagogical friction names the intentional preservation of resistance necessary for learning, across four dimensions.
structural conditions
(predictors)
- school neighborhood poverty level
- written ai policy (presence/absence)
- ai literacy instruction available
- professional development access
noetic friction
cognitive struggle
rhetorical friction
real audiences
existential friction
owning claims
infrastructural friction
policy conditions
cross-cutting equity construct:
productive vs. exclusionary friction
(english learner paradox; discrit)
outcomes
(by rq)
- teacher practice
rq1 - institutional conditions
rq2 - student experience
srq - policy implications
rq3
noetic friction
the head.
the internal cognitive resistance required to transform external information into internalized understanding — synthesis, argumentation, the struggle to revise one's own thinking.
rhetorical friction
the room.
the social and dialogic struggle of discussing and defending ideas against unpredictable human interlocutors whose responses cannot be scripted or controlled.
existential friction
the world.
the experience of being personally accountable for claims made in real space — the vulnerability of presenting one's own thinking and bearing intellectual risk.
infrastructural friction
the system.
the structural and policy conditions — institutional values, assessment norms, professional development priorities — that make the other three dimensions possible to sustain.
the framework distinguishes productive friction — resistance that builds capacity — from exclusionary friction — arbitrary barriers that prevent participation. preserving the first while removing the second requires situated professional judgment that no blanket policy can fully specify.
mixed methods design.
embedded within a collective instrumental case study, this pragmatist approach integrates phenomenological practitioner/student accounts with structural quantitative datasets to produce actionable meta-inferences.
qualitative strand (dominant)
participants (purposeful sampling)
- › 4 k-12 teachers (var. content/poverty)
- › 1-2 building administrators
- › 2 district leaders
- › 4 university students (high school cohorts of '22-'23)
data collection methods
- › semi-structured interviews (temporal & current)
- › interactive card sort protocol
- › institutional document analysis
quantitative strand (contextual)
secondary data
- › nces school pulse panel (dec 2024 data)
- › rand american educator panel (aep)
primary data collection
- › original teacher survey (n=50-100)
- › scales measuring noetic, rhetorical, existential, and infrastructural practices
study phases.
phase 1: concurrent data collection
interviews, card sort, survey distribution, and secondary data compilation occur concurrently.
phase 2: independent analysis
thematic coding via framework lens (a priori + emergent). temporal coding applied exclusively to university student transcripts. descriptive stats for quan data.
phase 3: integration
joint displays construct meta-inferences comparing structural patterns to phenomenological accounts.
phase 4: ai comparison (saq)
structured discourse comparison: applying identical interview protocols to 3 major llms.
research questions.
three primary questions operationalize the framework for k–12 contexts. a secondary question captures the learner perspective on the media transition the framework theorizes. a supplementary analytical question tests the framework's claims about ai-generated discourse.
primary research questions.
- rq1
how do k–12 teachers understand and navigate the friction-reducing affordances of generative ai in academic work?
- rq2
what institutional conditions — policy, assessment design, professional development, leadership disposition — enable or constrain friction-preserving pedagogy?
- rq3
how can the pedagogical friction framework inform ai policy development in k–12 contexts?
secondary research question.
- srq
how do current university students who were in grades 10–12 when generative ai became publicly available describe changes in their experiences of cognitive struggle, authorial ownership, and learning processes across the pre-genai and post-genai transition?
supplementary analytical question.
- saq
when agentic ai systems are prompted to respond as educators to the same interview protocol and survey instrument, how do their outputs differ structurally from human practitioner responses, and what do these differences reveal about the pedagogical reasoning the framework identifies as necessary for friction-preserving practice?
examples of research instruments applied for this research.
examples of some custom-built tools support data collection and analysis for this study. access is provided for methodological transparency; participant data is not publicly hosted.
virtual interviewer
semi-structured interview protocol organized around the four friction dimensions, with practitioner and student versions. deployed via github pages to support asynchronous participation under irb-approved consent procedures.
instrument 02.analysis dashboard
qualitative coding environment for transcript analysis. supports thematic coding against the pedagogical friction framework, with a temporal layer for student retrospective accounts.
research collaboration & inquiries.
for questions about the framework, requests to cite in-progress work, or speaking and consulting inquiries related to this research, use the form below. replies typically come within five business days.