Learning desk

MultiAgent EDU StackGather good sources. Teach what matters.
T5TauricResearch/TradingAgentsT5A Man Who Invented Modern AI (Before Everyone Else) – Jürgen Schmidhuber [video]T5GPT-4 finished training four years ago todayT5AI Settles a 25 Year-Old Problem We Left BehindT5What it was like working on LLMs and security at Meta (2022-2026)T5Ask HN: How do you go from writing code to deploying with agents?T5What Happened: OpenAI and HuggingFaceT5Apple says Mac users in China can connect to Alibaba's Qwen AI serviceT5Show HN: Try Benzi – A coding harness/agent beating Claude Code itself on SonnetT5The AI Apocalypse Is HereT3Auto mode is now the default in Claude Code for Pro, Max, and Team plansT5Show HN: Tura – Build agent that uses 80% less token and delivers better resultsT5TauricResearch/TradingAgentsT5A Man Who Invented Modern AI (Before Everyone Else) – Jürgen Schmidhuber [video]T5GPT-4 finished training four years ago todayT5AI Settles a 25 Year-Old Problem We Left BehindT5What it was like working on LLMs and security at Meta (2022-2026)T5Ask HN: How do you go from writing code to deploying with agents?T5What Happened: OpenAI and HuggingFaceT5Apple says Mac users in China can connect to Alibaba's Qwen AI serviceT5Show HN: Try Benzi – A coding harness/agent beating Claude Code itself on SonnetT5The AI Apocalypse Is HereT3Auto mode is now the default in Claude Code for Pro, Max, and Team plansT5Show HN: Tura – Build agent that uses 80% less token and delivers better results
← Dispatches

Effects of Partial vs. Full Cognitive Offloading with Generative AI on Learning Retention

Primary research

#657

T1new
Topic
unassigned (set during synthesis)
First seen
2026-07-23 07:15:51
Last seen
2026-07-23 07:15:51

Source raw items (1)

  • Semantic Scholar2026-07-23 07:15:18
    Effects of Partial vs. Full Cognitive Offloading with Generative AI on Learning Retention

    Even though the effects of Generative AI (GenAI) on learning retention are still not adequately understood, its use in higher education is increasing. Recent studies show that GenAI has the potential to improve immediate performance while undermining retention following the withdrawal of the tool. This study examines whether retention varies based on the degree of cognitive offloading in GenAI-supported learning. The study used a two-group quasi-experimental design, framed by cognitive load theory and the desirable difficulties principle, involving 63 students studying for their master's degrees at a Saudi public university. The full cognitive offloading group (FCO) used ChatGPT (GPT-4o), which provided direct responses. The group delegated substantial cognitive processing to GenAI. The partial cognitive offloading group (PCO) employed a custom GPT designed to provide structured guidance and scaffolding rather than answers, leaving the cognitive processes with the student. A researcher-developed cognitive test was used for measuring learning retention and was mapped to five levels of the revised Bloom's taxonomy. It was administered at three time points: pre-test, immediate post-test, and delayed post-test five weeks later. Data were analysed using a 2 x 3 mixed-design ANOVA, with follow-up MANOVA across the five levels of the revised Bloom’s taxonomy. Results indicate that PCO students outperformed FCO students at both post-tests, with the advantage extending across all five levels of Bloom’s taxonomy, and remaining stable across the five-week retention interval. The findings suggest that GenAI is more effective at supporting retention when used to scaffold the active reasoning of learners as opposed to substituting for their cognitive processing.