Solving the global STEM educational crisis using Cognitive Load Optimization and Artificial Intelligence

Solving the global STEM educational crisis using Cognitive Load Optimization and Artificial Intelligence–A preliminary comparative analysis Maj

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Solving the global STEM educational crisis using Cognitive Load Optimization and Artificial Intelligence–A preliminary comparative analysis 2024 Eurasia Journal of Mathematics, Science and Technology Education 10.29333/ejmste/14448 Maj2024-solving-the-global-stem-educational-crisis-u Maj2024-solving-the-global-stem-educational-crisis-u.txt, the project's staged copy b17cf5d3bfd3722af3ec666e86098cea90f97312 not-redistributable education

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them method Quotation withheld because this record is not cleared for redistribution. cognitive load in STEM instructional materials Quotation withheld because this record is not cleared for redistribution. review review Quotation withheld because this record is not cleared for redistribution. mechanisms short-term (working) memory has limited capacity, so high-intrinsic-load material overloads it and blocks schema construction in long-term memory Quotation withheld because this record is not cleared for redistribution. mechanisms cognitive gaps in the material raise cognitive load, students fill the gaps with their own incorrect conclusions, and the resulting misconceptions then handicap further learning Quotation withheld because this record is not cleared for redistribution. mechanisms because existing learning theories are qualitative and open to interpretation, material quality tracks the individual educator's skill, which produces wide variation in pedagogical quality and pass rates Quotation withheld because this record is not cleared for redistribution. CLO is time consuming to implement, which the author offers economies of scale to offset. what this limits: the practicality claim: HOWEVER, CLO IS TIME CONSUMING TO IMPLEMENT Quotation withheld because this record is not cleared for redistribution. The effect of ChatGPT on learning outcomes was not determined by this study. what this limits: any learning-outcome claim: FURTHER WORK IS NEEDED TO DETERMINE AI CHATGPT'S IMPACT ON LEARNING OUTCOMES Quotation withheld because this record is not cleared for redistribution. The new data in the paper come only from analysis of printed materials, not from a new controlled study. what this limits: the evidential basis of the paper: NEW DATA PRESENTED IN THIS PAPER STEMS FROM AN ANALYSIS OF PRINTED MATERIALS ONLY - no new controlled study Quotation withheld because this record is not cleared for redistribution. CLO is a detailed methodology that is time consuming to implement duplicate of: limitations[1] · what this limits: the same scope as limitations[1] - implementation cost Quotation withheld because this record is not cleared for redistribution. CLO is new and its detailed procedures and metrics must be learned before use, which costs time what this limits: the adoption claim specifically: CLO IS A NEW LEARNING THEORY WITH A DETAILED METHODOLOGY (PROCEDURES, METRICS ETC.) THAT IS TIME CONSUMING TO USE Quotation withheld because this record is not cleared for redistribution. The effect of ChatGPT on learning outcomes is not established by this study duplicate of: limitations[2] · what this limits: any learning-outcome claim: FURTHER WORK IS NEEDED TO DETERMINE AI CHATGPT'S IMPACT ON LEARNING OUTCOMES Quotation withheld because this record is not cleared for redistribution. Only text-based ChatGPT is considered what this limits: the modality covered: THIS PAPER IS RESTRICTED TO TEXT-BASED CHATGPT Quotation withheld because this record is not cleared for redistribution. The evidence offered for the method is one cohort given three lectures, and the cohort is described as students whose first language was not English. what this limits: the claim that this methodology addresses a global STEM educational crisis: the reported experiment is one cohort of 33 non-native-English-speaking students given three lectures in remote online mode Quotation withheld because this record is not cleared for redistribution. The AI half of the proposal is repeatedly marked as not yet established, in the same table that presents the framework. what this limits: the artificial-intelligence component of this proposal: the paper states ChatGPT output is not necessarily coherent and contiguous and that further work is needed on its role in instructional design and assessment Quotation withheld because this record is not cleared for redistribution. Its effect on the outcome the paper cares about is explicitly still open. what this limits: any expectation that adding ChatGPT improves learning outcomes here: the paper states further work is needed to determine its impact on them Quotation withheld because this record is not cleared for redistribution. The comparison class is characterised rather than measured: existing theories are described as qualitative and subjective, and the description is the ground for replacing them. what this limits: the contrast between this quantitative theory and existing ones: the existing theories are characterised as qualitative, subjective and dependent on the educator's skill, rather than compared on a measured outcome Quotation withheld because this record is not cleared for redistribution. The internalisation mechanism the framework depends on is supported only as 'some evidence', with further work stated as needed. what this limits: the claim that CLO schema diagrams are internalised into long-term memory: the paper reports only 'some evidence' for it and hedges the downstream scaffolding with 'may' Quotation withheld because this record is not cleared for redistribution. the design of instructional material is recast as a combinatorial optimization problem — identifying, ranking and sequentially ordering interdependent elements to minimise a measured cost function (cognitive load) Quotation withheld because this record is not cleared for redistribution. cognitive load optimization cognitive load optimization Quotation withheld because this record is not cleared for redistribution. science of learning science of learning Quotation withheld because this record is not cleared for redistribution. SOLO taxonomy SOLO taxonomy Quotation withheld because this record is not cleared for redistribution. cognitive load optimization - claimed as a QUANTITATIVE science-of-learning theory, defined by modelling relational knowledge as schemas optimised for the lowest cognitive load Quotation withheld because this record is not cleared for redistribution. learning theory - the paper draws a line ChatGPT is placed on the far side of: a useful pedagogical tool, but not a learning theory Quotation withheld because this record is not cleared for redistribution. existing learning theories - characterised as qualitative soft-science principles with subjective guidelines open to interpretation, which is the stated ground for a quantitative replacement Quotation withheld because this record is not cleared for redistribution. extends → Cognitive Load Theory (CLT); de Jong, 2010 Quotation withheld because this record is not cleared for redistribution. extends → Cognitive Load Theory (CLT), de Jong 2010 Quotation withheld because this record is not cleared for redistribution. contradicts → Constructivist learning theory Quotation withheld because this record is not cleared for redistribution. builds-on → SOLO taxonomy (Biggs & Collis, 1982, 1989) Quotation withheld because this record is not cleared for redistribution. cognitive load optimization —is-a→ science of learning cognitive load optimization science of learning strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. descriptive statistics descriptive statistics Quotation withheld because this record is not cleared for redistribution. mean mean Quotation withheld because this record is not cleared for redistribution. standard deviation standard deviation Quotation withheld because this record is not cleared for redistribution. ChatGPT 3.5's output on hubs versus switches imposed a high cognitive load with cognitive gaps, risking student misconceptions. Quotation withheld because this record is not cleared for redistribution. ChatGPT's explanation of the two sine equations used degrees in one and radians in the other, a potential source of confusion. Quotation withheld because this record is not cleared for redistribution. Every STEM curriculum item examined across school, college and university level was non-sequential and contained cognitive gaps, i.e. carried a high cognitive load. Quotation withheld because this record is not cleared for redistribution. 64 business students who had passed three STEM IT units all scored zero on ten simple relational-knowledge questions. Quotation withheld because this record is not cleared for redistribution. CLO consistently achieved high pass and retention rates (typically 100%) together with very positive student satisfaction. Quotation withheld because this record is not cleared for redistribution. Cognitive Load Theory lacks a reliable quantitative metric for intrinsic cognitive load. Quotation withheld because this record is not cleared for redistribution. Ninety-nine percent of the students in the comparison preferred CLO-based learning over the standard method. Quotation withheld because this record is not cleared for redistribution. ChatGPT works across STEM disciplines and levels but is an educational technology rather than a learning theory, and gives no assurance of pedagogical quality. Quotation withheld because this record is not cleared for redistribution. Cognitive Load Optimization (CLO) Cognitive Load Optimization (CLO) Quotation withheld because this record is not cleared for redistribution. Cognitive Load Theory (CLT) Cognitive Load Theory (CLT) Quotation withheld because this record is not cleared for redistribution. intrinsic cognitive load (ICL) intrinsic cognitive load (ICL) Quotation withheld because this record is not cleared for redistribution. schema schema Quotation withheld because this record is not cleared for redistribution. relational knowledge relational knowledge Quotation withheld because this record is not cleared for redistribution. misconceptions misconceptions Quotation withheld because this record is not cleared for redistribution. SOLO taxonomy SOLO taxonomy Quotation withheld because this record is not cleared for redistribution. pass and retention rates pass and retention rates Quotation withheld because this record is not cleared for redistribution. 99% — share of students preferring CLO-based learning to the standard method Quotation withheld because this record is not cleared for redistribution. 64 ... zero — business students tested on ten relational-knowledge questions after passing three STEM IT units, and their score Quotation withheld because this record is not cleared for redistribution. 100% — pass and retention rates for engineering mathematics taught by the CLO method in remote access mode Quotation withheld because this record is not cleared for redistribution. mean of 4.09 with a standard deviation of 0.739 — rating of the Relative Benefit factor by the Chinese graduate-student cohort taught with CLO Quotation withheld because this record is not cleared for redistribution. 210 — business studies graduate students whose behavioural intention to adopt CLO was evaluated with TAM and UTAUT WHOSE RESULT THIS IS: students’ behavioral intention to adopt/use CLO (Nuangjamnong, 2022). - THE RESULT OR THE NUMBER IS THE CITED WORK'S. Read, not swept - the trailing-citation shape is about 55% precise and the main clause is what decides it. Second reading pass over the 25 candidates left after 2026-08-25 11:39: 9 marked, 16 refused. 'The behavioral…' - a cited study's cohort and result. Quotation withheld because this record is not cleared for redistribution. 29% ... 34% — respondents in a national online-learning report complaining of lack of engagement, and of inadequate academic interaction Quotation withheld because this record is not cleared for redistribution. one million — additional STEM graduates called for by the US President's Council of Advisors on Science and Technology Quotation withheld because this record is not cleared for redistribution. 33 — non-native-English students taught three remote lectures with CLO materials in the reported experiment Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.29333/ejmste/14448.