Educational researcher · cognition, learning, and expertise
Research profile
I study how people build understanding, retain knowledge, and develop the capacity for independent judgement, particularly when cognitive technologies become part of the work through which they learn.
My academic home is educational science, with a focus on professional and workplace learning. I draw on cognitive science, cognitive psychology, linguistics, human–computer interaction, and organisational research to examine the mechanisms beneath learning in practice: attention, active generation, explanation, memory integration, metacognitive monitoring, and transfer.
A recurring focus in my work is the difference between performing a task successfully (now) and developing the knowledge needed to judge well (later). This distinction becomes especially important when epistemic technologies are employed to retrieve, organise, classify, summarise, or explain information on a learner’s behalf.
Research interests
I am interested in the learning consequences of generating an explanation, interpretation, or solution rather than receiving one. This includes the generation effect, self-explanation, retrieval practice, productive failure, error-based learning, and the conditions under which effort becomes educationally productive rather than merely burdensome.
My work asks how information is encoded and integrated into durable relational structures: situation models, causal and temporal representations, and forms of knowledge that can later be reconstructed rather than simply recognised. I am particularly interested in the difference between processing fluency, surface recall, and transferable understanding.
Summaries, interfaces, prompts, and linguistic forms influence what becomes salient, what is integrated, and what disappears from view. I study when technological structuring removes extraneous load and when it removes the very operations (comparison, inference, organisation, revision) through which learning occurs.
How do learners and professionals judge what they know, identify uncertainty, calibrate confidence, assess sources, and recognise when an apparently coherent output has not been understood? I am interested in monitoring and control processes in both ordinary learning and human–AI decision-making.
Cognition is distributed across people, documents, procedures, interfaces, and machines. Without passing general judgement on whether offloading is inherently good or bad, I am interested in which operations can be delegated without weakening encoding, conceptual organisation, pattern recognition, transfer, or the development of independent judgement.
I examine how participation in work, supervision, feedback, access to progressively demanding tasks, and exposure to ordinary as well as exceptional cases support the formation and reproduction of expertise. I am especially interested in transitions from novice performance to flexible, accountable, and adaptive judgement.
Current and developing work
Current PhD research project
Examining how AI and other epistemic technologies redistribute interpretive labour: the work through which evidence, categories, uncertainty, and explanation become an accountable representation of a case. The central concern is how different arrangements of task, technology, supervision, and participation shape professional learning and the reproduction of expertise.
Approaches: Cognitive ethnography, Cognitive Task Analysis, comparative workplace studies, case walkthroughs, behavioural tasks, and mixed-methods analysis.
Completed Master’s research project
My master thesis in linguistics developed a theory-motivated framework for identifying sources of interpretive load in human- and large-language-model-generated discourse. The broader research question was how differently organised (and sourced) language changes the work required to construct a coherent situation model, and what remains available to the reader afterwards.
Approaches: discourse comprehension, cognitive linguistics, situation-model theory, corpus analysis, NLP, and quantitative modelling.
Collaborative research · BI Norwegian Business School
Together with Jan Ketil Arnulf and Janicke Lilletvedt Skar this project develops AI- and NLP-enabled methods for identifying theoretically defined governance rationales in annual and sustainability reports. It asks whether language can reveal dynamic changes in how organisations justify governance arrangements beyond conventional structural indicators.
Approaches: theory-led annotation, LLM-assisted classification, longitudinal corpus analysis, validation, and reproducible computational workflows.
Selected work
Teaching
My teaching is guided by the ideas I am researching, i.e. that helping someone complete the present task is not the same as helping them become capable of solving the next one. I therefore try to make reasoning visible, elicit predictions before explanations are given, provide feedback on process as well as outcome, and create opportunities for retrieval, revision, and transfer.
I have taught statistics, advanced statistics, research methods, academic writing, language, and professional courses across university and organisational settings.
Academic and professional background
Behavioural and cognitive: task and manipulation design, operationalisation, construct validity, memory and transfer measures, self-explanation and generation paradigms, confidence and metacognitive judgements.
Qualitative and workplace-based: cognitive ethnography, semi-structured and critical-incident interviews, Cognitive Task Analysis, observation, stimulated recall, workflow reconstruction, and comparative case analysis.
Computational and quantitative: Python, R, NLP, LLM-assisted classification, corpus analysis, annotation design, statistical modelling, visualisation, Git, and reproducible workflows.
Collaboration
I am highly collaborative and especially enjoy projects in which different disciplines get to refine the research question. I would be glad to hear from researchers, laboratories, and organisations working on learning, memory, expertise, and technology-mediated cognition.
I am by no means presenting, or wanting to present myself as an independently trained EEG specialist or applied (neuro)scientist. I am interested in co-developing neurocognitive studies with collaborators who bring neurophysiological expertise. I can contribute the educational problem, cognitive theory, experimental task design, behavioural measures, and computational analysis (and likely more, which we often discover together).
For research conversations, collaboration ideas, workshops, or guest teaching: jovanak@uio.no