AI Brief: The ascent of individual variance in education—the increasing importance of individual differences

This is another IQs Corner AI Brief.  

Prepared by Dr. Kevin McGrew with major assist from Google NotebookLM.  (Click here for brief explanation of how IQs Corner creates AI Briefs from article PDFs).  

For the first time I’m also experimenting with the Google NotebookLM feature of creating an AI generated infographic (Beta) from the research article—click on the image to enlarge for easy viewing.

The Ascent of Individual Variance in Global Education

 

The article "The growing role of individual differences: A cross-National Study of achievement variance reallocation from grade 4 to 8," published in the journal Intelligence (Eriksson et al., 2026; click here to acquire open access PDF copy), explores how the determinants of student achievement shift as children transition from late childhood (approximately age 10) to early adolescence (approximately age 14). The researchers, led by Kimmo Eriksson, sought to determine whether environmental factors, such as the quality of a national school system, become more influential over time through compounding advantages, or if individual learning characteristics grow in importance as academic material becomes more complex.

 

Theoretical Framework

 

The study tested three competing theoretical perspectives on achievement development between Grade 4 and Grade 8:

• Skills-Beget-Skills: Suggests early academic advantages create cascading benefits, predicting that high-quality national systems should lead to compounding advantages and an increase in the proportion of variance attributable to countries.
• Opportunity-to-Learn (OTL): Emphasizes exposure to content and predicts that variance at the school and class levels should increase as curricula become more specialized and students are sorted into different tracks.
• Individual Differences + Institutional Response: The authors’ integrated framework proposes that developmental processes create new individual-level variance, while educational systems respond by sorting students into different classes (tracking/streaming), thereby reallocating that variance to the class level.

Methodology

 

The researchers utilized data from the Trends in International Mathematics and Science Study (TIMSS) across three cohorts (2011–2015, 2015–2019, and 2019–2023). Their analysis involved dozens of countries and two primary methods:

1. Systematic Variance Decomposition: A four-level partition of achievement variance across countries, schools within countries, classes within schools, and individual students.
2. Cross-National Analysis: A formal model examining the relationship between individual characteristics (proxied by within-country relative standing) and educational system quality (proxied by country mean achievement).

Key Findings

 

The results across all cohorts and both subjects (mathematics and science) consistently supported the Individual Differences + Institutional Response hypothesis (H3) and directly contradicted the Skills-Beget-Skills hypothesis.

• Decrease in Country Influence: The proportion of achievement variance attributable to the country level decreased substantially (by 4–11 percentage points) as students moved from Grade 4 to Grade 8.
• Increase in Class-Level Importance: The proportion of variance at the class level increased substantially (by 3–7 percentage points). The class level was unique in benefiting from both the creation of new variance (through differentiated instruction) and the movement of variance (through ability-based sorting).
• Compensatory Advantage: The cross-national analysis revealed that the "slope" relating individual characteristics to system quality was shallower in Grade 8 than in Grade 4. This means that while students in weaker systems need higher individual characteristics to reach a certain achievement level (e.g., 500 points), this compensatory requirement is smaller in Grade 8, indicating that individual traits are increasingly pulling students ahead regardless of their national system's quality.

Conclusions and Implications

 

The authors conclude that stable individual characteristics affecting learning capacity—such as cognitive abilities, motivation, and self-regulation—become more influential as students mature. These traits are further magnified through interaction with educational environments, such as the "Matthew effect," where high-performing students elicit more challenging opportunities and resources. For educational practice, these findings suggest that pedagogical strategies may need to accommodate a wider range of learning profiles as students progress through school. Furthermore, the study cautions researchers that interventions targeting specific early skills may experience "fadeout" if they do not address the underlying learning capacities that become increasingly determinative during adolescence.

 

Research alert: #ExecutiveFunctions, #Metacognition, #Self-Regulation, and #Self-RegulatedLearning – What are We Talking About? A Review and Introduction of the #EMERGE Model

Quick email FYI research alert post.  

 

This is a much needed and interesting attempt to deal with the “jingle-jangle” fallacy amoung the cognition-related constructs of executive functions, metacognition, self-regulation, and self-regulated learning—via the proposed EMERGE model.  

 

I’m setting this aside for focused reflective reading.  Good news…it is open access and thus downloadable 👍

 

Executive Functions, Metacognition, Self-Regulation, and Self-Regulated Learning – What are We Talking About? A Review and Introduction of the EMERGE Model | Educational Psychology Review | Springer Nature Link 
https://link.springer.com/article/10.1007/s10648-026-10157-0

Abstract 

As students progress through school, they are expected to increasingly regulate their attention, behaviour, and learning. While some meet these demands with ease, others face ongoing challenges that can hinder their academic success. Research has identified four key concepts in this area: executive functions (EF), metacognition (MC), self-regulation (SR) and self-regulated learning (SRL). Although these constructs are conceptually related, they have often been examined in isolation due to disciplinary and methodological divides, resulting in fragmented accounts that obscure their dynamic interplay. This review addresses this issue by providing a comparative overview of EF, MC, SR and SRL in terms of their definitions, how they are operationalised and the research designs used. Based on this synthesis, we introduce the EMERGE model, which positions these constructs along a continuum ranging from more biologically grounded mechanics (e.g., EF) to more culturally shaped pragmatics (e.g., strategy knowledge in SRL). The model highlights both shared mechanisms and distinct functions and conceptualises SR in learning situa-tions as an integrative construct. Building on this framework, we propose two guid-ing hypotheses: the stage-setting hypothesis, which emphasises long-term develop-mental interplay; and the compensatory hypothesis, which focuses on short-term interactions that predict learning outcomes. Together, these perspectives highlight the need for longitudinal, experimental, and hybrid designs to capture developmen-tal and dynamic processes. The EMERGE model thus aims to bridge fragmented research traditions, improve diagnostics, and inform interventions that effectively support students in meeting the growing demands of self-regulated and adaptive learning.

Click on image to enlarge for easy viewing/reading

The McGrew (2022) Cognitive-Affective-Motivation Model of Learning (CAMML) article update - “something is happening here, what it is ain’t exactly clear” - #cognitive #intelligence #affective #motivation #intelligence #CHC #CAMMl #schoolpsychology #schoolpsychologists

In 2022 I published an invited big-picture “thought piece” on a proposed CAMML (cognitive-affective-motivation model of learning) in the Canadian Journal of School Psychology   The title wasThe Cognitive-Affective-Motivation Model of Learning (CAMML): Standing on the Shoulders of Giants.

I had hoped that by challenging existing narrow assessment practices in school psychology (SP), and proposing a more whole-child assessment model approach (where cognitive testing would be more limited and selective…not the knee jerk practice of most all referred kids for learning problems being administered a complete intelligence test battery), it would gain traction in some SP circles. From the informal and formal professional media sources I monitor, it has not..at least not yet.

The article was deliberately provocative and challenged the field of SP (especially trainers and leaders) to consider new assessment ideas and paradigms.  I fully recognized that the inertia of tradition and the constraints imposed by vested interest groups makes drastic paradigm changes in education difficult.  But as an invited tought piece one has more degree’s of freedom 😉. 

Recognizing how difficult it is to change established assessment practices, and recognizing the “ivory tower” orientation of the article, I stated: 

Integrating CAMML aptitude-trait complexes, which emphasize that motivation and SRL constructs are the focal personal investment learning mechanisms, in contemporary SP practice is an aspirational goal. The constraints of regulatory frameworks and the understandable skepticism of disability-specific advocacy groups will make such a paradigm-shift difficult. However, embracing the model of CAMML aptitude complexes may be what SP and education need to better address the complex nuances of individual differences in student learning. Snow's concept of aptitude, if embraced in reborn form as the CAMML framework, could reduce the unbalanced emphasis on intelligence testing in SPs assessment practices. However, the greatest impediment to change may be the inertia of tradition in SP”

Several weeks ago I completed a Google Scholar search to ascertain how frequently this article had been cited.  I was curious as I had seen no references to the article in traditional SP or assessment-related sources.  It is clear that the CAMML model (or any parts of it) have not yet resonated in SP or closely related education fields.  Perhaps it never will. Or………

To my surprise the search revealed 28 citations, most (but not all) outside of SP or related assessment publication outlets (except for another article I authored in 2023 and, of course, the motivation special issue introduction to that specific journal volume).  Here is a link to the results of this search.  The graph below shows a slow but increasing annual rate of reference citations.  Hmmmm….

Click on images to enlarge

Most references provide links to PDF articles if you want to skim the wide variety of non-SP contexts where the CAMML article has been cited.  It is a very interesting mix of professional topics and outlets. In my 45+ years of scholarship, I’ve never had a journal publication recognized almost exclusively outside of the intended professional audience.  Perhaps this is good…perhaps not.  I find it fascinating. Perhaps the diversity of professional outlet citations might foreshadow more wide-ranging (yet more gradual) future impact.  Below is the abstract and keywords from the article.  The CAMML article can be downloaded from my professional web page here. Below are colorized versions of the two figures from the article.

Abstract: The Cognitive-Affective-Motivation Model of Learning (CAMML) is a proposed framework for integrating contemporary motivation, affective (Big 5 personality) and cognitive (CHC theory) constructs in the practice of school psychologists (SPs). The central tenet of this article is that SPs need to integrate motivation alongside affective and cognitive constructs vis-à-vis an updated trilogy-of-the-mind (cognitive, conative, affective) model of intellectual functioning. CAMML builds on Richard Snow's seminal research on academic aptitudes—which are not synonymous with cognitive abilities. Learning aptitude complexes are academic domain-specific cognitive abilities and personal investment mechanisms (motivation and self-regulation) that collectively produce a student's readiness to learn in a specific domain. CAMML incorporates the “crossing the Rubicon” commitment pathway model of motivated self-regulated learning. It is recommended SPs take a fresh look at motivation theory, constructs, and research, embedded in the CAMML aptitude framework, by going back-to-the-future guided by the wisdom of giants from the field of cognition, intelligence, and educational psychology.

Keywords:  motivation, self-regulated learning, aptitudes, domain-specific, aptitude complexes, crossing the Rubicon, taxonomies, individual differences, readiness, CHC theory, Big 5, Gf-Gc theory

Click on images to enlarge

A #metaanalysis of #assessment of self-regulated learning (#SRL) - #selfregulatedlearning #learning #motivation #CAMML #EDPSY #schoolpsychologists #schoolpsychology #conative

Self-regulated learning (SRL) strategies are an important component of models of school learning.  Below is a new meta-analysis of SRL assessment methods.  Overall effect sizes are not large.  More R&D is needed to develop applied practical SRL measurement tools.  SRL is a one of the major components of the 2022 Cognitive-Affective-Motivation Model of Learning; CAMML; click here to access article),

Multimethod assessment of self-regulated learning in primary, secondary, and tertiary education – A meta-analysis.  Learning and Individual Differences (open access—click here to access).

Author links open overlay panel

Julia Ruhl, Franziska Perels , Laura Dorrenbacher-Ulrich

Abstract

Self-regulated learning (SRL) can be measured in several ways, which can be broadly classified into online and offline instruments. Although both online and offline measurements have advantages and disadvantages, the over-dependence of SRL research on offline measurements has been criticised considerably. Currently, efforts are being made to use multimethod SRL assessments. We examined 20 articles with 351 effect sizes that assessed SRL with at least two instruments on at least two SRL components. Most effect sizes were not statistically significant but descriptively higher than others. Combinations of two online instruments showed the highest effect size (r = 0.24). Overall correlations between instruments were highest for university students (r = 0.21). Additionally, results for cognition showed the highest effect size measured with behavioural traces (r = 0.28), and for metacognition measured with microanalysis (r = 0.35). The component of motivation was best measured using self-report questionnaires (r = 0.29).

Educational relevance statement

Self-regulated learning is an important predictor of academical success. It is therefore necessary to measure it as precise and comprehensive as possible. Knowing which instruments are best suited for each age group, SRL component, or reliably predict a specific achievement variable can help educators pick the best instrument for their needs.

Leaving no child behind—Beyond cognitive and achievement abilities - #CAMML source “fugitive/grey” working paper now available. Enjoy - #NCLB #learning #EDSPY #motivation #affective #cognitive #intelligence #conative #noncognitive #schoolpsychology #schoolpsychologists

I’ve recently made several posts regarding the importance of conative (i.e., motivation; self-regulated learning strategies; etc.) learner characteristics and how they should be integrated with cognitive abilities (as per the CHC theory of cognitive abilities) to better understand the interplay between learner characteristics and school learning.  These posts have mentioned (and I provided a link) to my recent 2022 article where I articulate a Cognitive-Affective-Motivation Model of Learning; CAMML; click here to access).

In the article I mention that the 2022 CAMML model had its roots in early work I completed as one of the first set of Principal Investigators during the first five years of the University of Minnesota’s National Center on Educational Outcomes (NCEO).  As a result of those posts I’ve had several requests for the original working paper which is best characterized as being “fugitive” or “grey” literature.

The brief back story is that the original 2004 document was a “working paper” (6-15-04; Increasing the Chance of No Child Being Left Behind: Beyond Cognitive and Achievement Abilities, by Kevin McGrew, David Johnson, Anna Casio, Jeffrey Evans) that was written with the aid of discretionary funds from the then Department of Education’s Office of Special Education (OSEP) during the influence of NCLB.  The working draft was submitted but curiously never saw the light of day.

With this post I’m now making the complete 2004 “working paper” (with writing, spelling, and grammar blemish’s in their full glory) available as a PDF.  Click here to access.  Although dated 20 years, IMHO the lengthy paper provides a good accounting of the relevant literature up to 2004, much of which is still relevant.  Below are images of the TOC pages which should give you an hint of the treasure trove of information and literature reviewed.  Enjoy.  Hopefully this MIA paper may help others pursue research and theoretical study in this important area.

Click on images to enlarge for easy reading

Research Byte: Is trying hard enough? Causal analysis of the effort-IQ relationship suggests not - #intelligence #IQ #motivation #volition #CAMML #conative #noncognitive

Is Trying Harder Enough? Causal Analysis of the Effort-IQ Relationship Suggests Not.  Timothy Bates. Intelligence and Cognitive Abilities (open access—click here to locate article to read or download)

Abstract

Claims that effort increases cognitive scores are now under great doubt. What is needed is randomized controlled trials optimized for testing causal influence and avoiding confounding of self-evaluation of performance with feelings of good effort. Here we report three large studies using unconfounded measures of effort and instrumental analysis to isolate any causal effect of effort on cognitive score. An initial study (N = 393) validated an appropriate effort measure, demonstrating excellent external and convergent validity (β = .61). Study 2 (N = 500, preregistered) randomly allocated subjects to a performance incentive, using an instrumental variable analysis to detect causal effects of effort. The incentive successfully manipulated effort (𝛽 = .18, p = .001). However, the causal effect of effort on scores was near-zero and non-significant (𝛽 = .04, p = .886). Study 3 (N=1,237) replicated this null result with preregistered analysis and an externally developed measure of effort: incentive again raised reported effort (𝛽 = .17, p <.001), but effort had no significant causal effect on cognitive score (β2 = .27 [-0.07, 0.62]), p = .15). Alongside evidence of research fraud and confounding in earlier studies, the present evidence for the absence of any causal effects of effort on cognitive scores, effort research should shift its focus to goal setting – where effort is useful – rather than raising basic ability, which it appears unable to do.

Select quote from discussion: “The present results suggest a potential ‘central dogma of cognition’: that volitional effort can direct cognitive resources but cannot fundamentally alter or bypass the efficacy of the underlying cognitive systems themselves”

These findings are consistent with my proposed cognitive-affective-motivation-model-of-learning (CAMML), grounded extensively on Richard Snows concept of aptitude trait complexes, where motivational constructs are seen as driving and directing the use of cognitive abilities (via personal investment mechanisms), but not directly having a causal effect on cognitive abilities.  See first of two figures below.  Note lack of causal arrows from conative and affective domain constructs to CHC cognitive abilities.  Paper can be accessed by clicking here.

Click on images to enlarge for easier viewing

Let’s hear it for #conative (#noncognitive) variables in understanding learning—#CAMML #aptitude #traitcomplexes #cognitive #affective #motivation #schoolpsychology

 Variation in the intensity and consistency of attention during learning: The role of conative factors

Click here to view online

Abstract

The present study examined whether conative factors (e.g., self-efficacy, self-set goal difficulty, and task-specific motivation) are reliable predictors of learning and memory abilities and whether any observed relationships could be explained by two related, yet distinct aspects of attention. Specifically, the present study examined whether the relationship between conative factors and overall learning performance is explained by attentional intensity (the amount of attention allocated to a task) and attentional consistency (the consistency with which attention is allocated to said task). In two studies (N’s > 160), participants completed a paired associate’s (PA) cued recall task while pupil diameter was simultaneously recorded to provide an index of the intensity of attention. Measures of working memory, general episodic long-term memory, task-specific motivation, and memory self-efficacy were also included. Study 2 adopted a similar procedure but embedded thought probes into the encoding phase of each list to provide an index of the consistency of attention. Study 2 also added measures of self-set goal difficulty and effective strategy use. Results suggested that all conative factors were related to intensity and consistency in challenging learning contexts. Furthermore, intensity, consistency, and the variance shared between self-efficacy and self-set goal difficulty (r = 0.86) each explained substantial unique variance in learning when controlling for the influence of other important predictors. Overall, results suggest conative factors are important for understanding individual differences in learning and memory abilities, and part of the reason why these factors are associated with improved learning outcomes is due to intensity and consistency.

Comment:  I’ve always believed that conative (non-cognitive) individual difference variables should receive just as much attention as cognitive variables in understanding learning.  In fact, in an invited article, I recently proposed the CAMML (cognitive-affective-motivation model of learning) “crossing the rubicon” model of learning that integrates conative (motivation and self-regulated learning), affective (Big 5 personality) and cognitive (CHC) variables in an overarching framework (building on Richard Snow’s concept of aptitude-trait complexes).  Click here to download or read the CAMML article.  Below are the two key figures for understanding the CAMML model.

Click on each image to enlarge for viewing

The McGrew Model of Achievement Competence Model (MACM)--Standing on the shoulders of giants: CJSP article supplementary materials

The Model of Achievement Competence Motivation (MACM) has been  under development since the early 2000's by Dr. Kevin S. McGrew.   The work is (has) been formally presented in an invited article--"The Model of Achievement Competence Motivation (MACM)--Standing on the shoulders of giants" (McGrew, in press, 2021), for a forthcoming special issue on motivation in the Canadian Journal of School Psychology). 

Due to the page length constraints of the journal, significant background and explanatory information could not be presented in the article.  Thus, I have "off-loaded" this material for supplementary viewing via on-line PPT slide shows and downloadable PDF files.

Five MACM PPT modules have been posted at SlideShare and can be viewed and downloaded from that site.  For those who would prefer to directly download PDF versions of the PPT modules from one page...here it is.  Below are the titles of the five modules and associated download links.  In addition, the paper includes, in a table footnote, definitions for 16 self-regulatory constructs from a recent article by Sitzman and Ely (2011).  That PDF file is also available from download below.

Enjoy.

The Model of Achievement Competence Motivation (MACM)

• Introduction and overview (A)
  
• Overview of the model (B)
• The motivation domains defined (C)
• The volition and self-regulation domain (D)
• Crossing the Rubicon Commitment Pathway Model to Learning (E)
  
• 16 self-regulatory constructs (from article definition table)
  

The Model of Achievement Competence Motivation (MACM) Part E: Crossing the Rubicon Commitment Pathway Model to Learning

 

The Model of Achievement Competence Motivation (MACM) is a series of slide modules.  By clicking on the link you can view the slides at SlideShare.  This is the fifth and final (Part E) in the series.  This one is brief...only 11 slides.  Crossing the Rubicon Commitment Pathway Model to Learning.  There will be a total of five modules.  The modules will serve as supplemental materials to "The Model of Achievement Competence Motivation (MACM)--Standing on the shoulders of giants" (McGrew, in press, 2021 - in a forthcoming special issue on motivation in the Canadian Journal of School Psychology)

You should be able to access the prior modules (A-C) from the link above.

Click here for prior "beyond IQ" labeled posts at this blog.

The Model of Achievement Competence Motivation (MACM): Part D-Volition and Self-Regulated Learning Domains

 The Model of Achievement Competence Motivation (MACM) is a series of slide modules.  By clicking on the link you can view the slides at SlideShare.  This is the fourth (Part D) in the series--Volition and Self-regulated Learning Domains described..  There will be a total of five modules.  The modules will serve as supplemental materials to "The Model of Achievement Competence Motivation (MACM)--Standing on the shoulders of giants" (McGrew, in press, 2021 - in a forthcoming special issue on motivation in the Canadian Journal of School Psychology)

You should be able to access the prior modules (A-C) from the link above.

Click here for prior "beyond IQ" labeled posts at this blog.

The Model of Achievement Competence Motivation (MACM) Part B: An overview of the MACM model

The Model of Achievement Competence Motivation (MACM) is a series of slide modules.  By clicking on the link you can view the slides at SlideShare.  This is the second (Part B) in the series--An overview of the model.  There will be a total of five modules.  The modules will serve as supplemental materials to "The Model of Achievement Competence Motivation (MACM)--Standing on the shoulders of giants" (McGrew, in press, 2021 - in a forthcoming special issue on motivation in the Canadian Journal of School Psychology)

Click here for first of the series (Part A:  Introduction and Background)

Click here for prior "beyond IQ" labeled posts at this blog.

The Model of Achievement Competence Motivation (MACM): Part A - Introduction to module series

The Model of Achievement Competence Motivation (MACM) is a series of slide modules.  By clicking on the link you can view the slides at SlideShare.  This is the first (Part A) in the series. The modules will serve as supplemental materials to "The Model of Achievement Competence Motivation (MACM)--Standing on the shoulders of giants" (McGrew, in press, 2021 - in a forthcoming special issue on motivation in the Canadian Journal of School Psychology)

Click here for prior "beyond IQ" labeled posts at this blog.

School readiness = self regulation learning competence?

Excellent review article on the relationship between the development of self-regulated learning strategies/competence and school readiness. Having worked in the schools as a school psychologist for 12 years, I would like to have a buck for every time a kindergarten teacher described children who where struggling in terms of self-regulation--although they did not use that term. A must ready for anyone working with preschool and early elementary students and stuff.

Click on images to enlarge. For more on self-regulated learning as per the Model of Academic Competence and Motivation (MACM), click here.










- Posted using BlogPress from my iPad

A cybernetic model for improving self-control

Click on images to enlarge







- Posted using BlogPress from my iPad

Excellent article on self-control in children

This is an excellent overview article re: the construct of self-control. It also compares and contrasts self-control research with the self-regulatory learning strategy research literature. Click on images to enlarge.












- Posted using BlogPress from my iPad

The MACM Commitment to Pathway Learning Model

[Click on images to enlarge]

I pleased to make available the first MindHub (TM) Pub.  This material first appeared as a blog post.  I received a number of requests for printed or electronic copies of the post, so I decided to make if available as a PDF...a MindHub Pub.  The title, which is also the download link, is "The Motivation and Academic Competence (MACM) Pathway to Commitment to Learning Model:  Crossing the Rubicon to Learning Action." It can also be accessed at the MindHub (TM).