Research Alert: Deary on Construct validity and intelligence differences: The murder board, the magic number, and the breadcrumb trail

Quick FYI email research alert post:  This is open access article.👍

 

Construct validity and intelligence differences: The murder board, the magic number, and the breadcrumb trail - ScienceDirect 
https://www.sciencedirect.com/science/article/pii/S0160289626000279?ref=cra_js_challenge&fr=RR-1

 

Abstract

In this piece there is a resume of some of the most-cited (though maybe less-than-closely-read) articles on construct validity, and a rediscovery that they articulate it in various ways, including destroying it. That moves on to considering: whether ideas about construct validity are different from/part of the philosophy of science; whether scientists generally and those, more specifically, working in intelligence, acquaint themselves with these ideas and guide their practice from having done so (and whether it matters if they don't); and whether the relationship between writers about construct validity/philosophy of science and empirical researchers is bite or bark (aka fire or smoke). Then, with attention to the construct/idea/field of individual differences in intelligence, there is an as-honest-as-introspection-affords account of what intelligence is like in this researcher's head: thus, the murder board, the magic number, and the breadcrumb trail. Conclusion: whereas intelligence test scores continue to accrue their impressive empirical regularities and its bio-, psycho-, and social- origins prove hard to divine, and whereas there will continue to be writings about construct validity, maybe most researchers in the field will continue to have their bespoke and more-or-less well-researched and -applied guardrails against bad science.

 

Select quotes:

 

“This preamble is by way of introducing that I thought it might be useful to try to put down how I represent—after more than 40 years of working on it—the construct validity of intelligence in my head. It's an idiographic approach, but it is intended to be an everyman account of the working- day utility that construct validity provides, including any personal wonkiness, vagueness, and inaccuracies.”

 

“The best metaphor I have been able to think of for how intelligence- differences-research-is-represented-in-my-brain is the murder board that one sees in TV detective programmes—the rather corny board with photos and post-it notes and other scraps of paper and writing, linked with bits of red tape secured with pins. The metaphor is just a way of representing the kinds of evidence that, to date, have accrued about and afford interpretations of cognitive test scores. The missing centre of the murder board is the culprit, though it might have suspects. With intelligence's murder board, we might have a holding note at the centre of the board with ‘intelligence' written on it. But, don't be fooled, that does not mean it is a thing, like the DNA molecule. Rather, it is a holding term for the load of stuff that surrounds it. For now, let's assume that that note called intelligence means the total score from a diverse battery of cognitive tests that were administered in a standardised way. It's that score, without assumptions about it (some will be thinking that there are always assumptions). The stuff that surrounds the ‘intelligence test score' note at the centre of the board records the empirical regularities found in relation to those scores. We might form a circle round it with notes that summarise phenotypic information; listing just a few examples, we would note the discovery and replication that cognitive tests tend to correlate positively (Carroll, 1993), that different batteries of cognitive tests afford a general cognitive component/factor that ranks people similarly (Johnson et al., 2008), that intelligence has high test- retest reliability and high longitudinal stability (Deary, 2014), that there are well-replicated ageing patterns (which tell us that there is a useful distinction between fluid and crystallised intelligence) (T. A. Salthouse, 2009), and so forth. It's been said many times before, but it's worth stating that intelligence (cognitive test scores) arguably enjoys more and more-robust empirical regularities than maybe anything else in psychology. At the centre of the board, then, we have haloes of well-replicated empirical regularities that surround test scores on diverse types of cognitive capability. Intelligence is a holding term for this (the suspect).”

 

 

Pardon typos and spelling errors-Message may be sent from iPhone and I've always had spelling problems :)

Research Alert: Comparative Analysis of #Dyslexia Laws by State: A Review and Case Study

Quick FYI email-based post.  Sorry…this is not an open access paper😕

 

Comparative Analysis of Dyslexia Laws by State: A Review and Case Study | Learning Disabilities: A Multidisciplinary Journal 
https://js.sagamorepub.com/index.php/ldmj/article/view/13093

 

Abstract

The current study provides an updated review of enacted dyslexia laws that includes specific implementation requirements for each state. Except Hawaii, all states and the District of Columbia now have a dyslexia law, and 41 states include a codified definition of dyslexia. Thirty-four state definitions specify that dyslexia is neurobiological in origin. There was considerable variability in key implementation requirements across states. Forty-four states mandate screening for dyslexia, and 37 states include intervention requirements. Regarding educator training, requirements for in-service training were more common than pre-service training. Given the ambiguity for implementation of dyslexia law requirements, a survey to gauge “dyslexia preparedness” was given to school psychologists in a specific state. Survey results suggested school psychologists generally perceived a lack of preparedness and insufficient training as major challenges for implementing dyslexia mandates. Implications for policy guidelines and broader school-level implementation of dyslexia mandates are discussed. 

Pardon typos and spelling errors-Message may be sent from iPhone and I've always had spelling problems :)

AI Alert: Aligning AI agents with developmentally appropriate CHC (select) cognitive domains—a thought provoking AI approach for school-age children and youth

 The source article is an arXiv open-access article available here.  I have no idea if it has undergone peer-review and is a pre-print or if it is working draft paper.  

Click on inserted images to enlarge for easier viewing.

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

ChildAgentEval, a psychometrically grounded benchmark designed to evaluate how well artificial intelligence aligns with the cognitive development of children and youth. Inspired by the Wechsler Intelligence Scale for Children-IV four-factor CHC model (the study does not use actual WISC-IV items as per test security standards), this framework assesses multimodal AI agents across ten interactive subtests covering the CHC domains like working memory (Gwm), verbal abstraction and vocabulary (Gc), fluid-visual spatial reasoning (Gf/Gv), and processing speed (Gs).

The study reveals that standard prompting—simply asking an AI to "act like a child"—does not authentically replicate pediatric developmental cognition, as models often maintain adult-level reasoning. The authors introduced a skill-guided distillation strategy that applies data-driven filters to simulate age-appropriate cognitive constraints and limitations. While AI agents can successfully adapt their vocabulary, the research highlights significant challenges in mimicking human-like bottlenecks in perceptual processing and memory retention. 

ChildAgentEval is the first psychometrically grounded interactive benchmark designed to measure cognitive age alignment in multimodal large language model (MLLM) agents. The study addresses a critical gap in current AI development: while state-of-the-art agents excel at complex reasoning, they often do not scaffold learning within a child's Zone of Proximal Development, often using adult-level abstractions that exceed a young user's cognitive grasp. The authors advocate for a shift in AI development from maximizing raw capability to ensuring developmental appropriateness for young users (e.g., school-age children and youth)—a feature that is much needed if AI is to be used in with school-age children and youth. 

The ChildAgentEval Framework

Inspired by the Wechsler Intelligence Scale for Children (WISC-IV), the benchmark evaluates agents across ten interactive subtests mapped to four primary CHC cognitive factors:

• Crystallized Intelligence (Gc): Verbal abstraction and vocabulary.
• Fluid and Visual-Spatial Reasoning (Gf/Gv): Rule induction and spatial problem-solving.
• Working Memory (Gwm): Information retention and manipulation.
• Processing Speed (Gs): Quickness of visual scanning and timed execution.

Unlike static evaluations, ChildAgentEval utilizes a Playwright-driven browser environment where agents must perform physical actions like clicking and typing to solve tasks. The framework also implements clinical protocols such as reversal and discontinuation rules to ensure developmental validity.

Skill-Guided Distillation

A core contribution of the research is a data-driven skill distillation strategy that moves beyond simple "act like a child" prompts. By analyzing a multi-source corpus of real child and adolescent interactions (ages 6–17), the researchers developed cognitive profile vectors. These vectors are converted into executable constraints via five cognitive filter modules injected into the agent's prompt, memory, and reasoning layers:

1. Vocabulary Abstraction Filter: Limits academic concepts and controls syntactic complexity.
2. Working Memory Mask: Physically simulates shorter memory spans by injecting noise or restricting cross-page information.
3. Reasoning Budget Controller: Restricts the depth of multi-step logic.
4. Visual Reliance Module: Reproduces cognitive biases, such as being misled by physical arrangement illusions.
5. Social Perspective Filter: Restricts reasoning to age-appropriate viewpoints, such as first-person vs. institutional perspectives.

Key Findings and Experimental Results

The study evaluated several proprietary models (e.g., GPT-5.4, Gemini-3.1-Pro) and open-weight models (e.g., Qwen3.5-27B). The experiments revealed three major insights:

• Standard Prompting Fails: Merely asking an agent to "act younger" does not reliably change its underlying cognitive behavior; most models continue to maximize correctness regardless of the requested age.
• Skill Guidance Enables Alignment: In high-performing proprietary models, the distillation method successfully induced monotonic score trajectories, where performance expanded naturally as the target age increased—the results produced performances associated with developmental growth curves of cognitive abilities.
• Uneven Domain Alignment: While agents easily adapted their linguistic style (Gc), they struggled to authentically simulate human-like limits in working memory and perceptual reasoning. This "domain dissociation" suggests that MLLM architectures currently lack the structural developmental bottlenecks found in biological cognition.

Conclusion

The authors conclude that for sensitive applications like educational tutoring, technical correctness must be secondary to developmental appropriateness. ChildAgentEval establishes a new paradigm for AI safety and alignment, shifting the focus from maximizing raw capability to authentic cognitive simulation.

Breaking legal and research news: SCOTUS (US Supreme Court) decision re multiple IQ scores in intellectual disability (ID) death penalty case (Hamm v Smith) - A DIG decision upholding the lower court decision

Today SCOTUS ruled on the Hamm v Smith multiple IQ score Atkins intellectual disability (ID) case that was argued before SCOTUS last December. This important decision, with all concurring and dissenting opinions, is available here.  Interested parties should read all the opinions. 

All documents and the history of the case can be found in a prior IQs Corner post.  As noted in the prior lengthy post, APA submitted an Amicus brief that, based on the reading of some of the justices opinions, suggests it (together with a AAIDD Amicus Brief) played an important role in the decision.

Briefly, when SCOTUS dismisses the writ of certiorari as improvidently granted it means the justices recognized that they made a mistake in agreeing to hear the case and decided to dismiss it without issuing a ruling on the merits. In legal shorthand, this is often referred to as a "DIG" (Dismissed as Improvidently Granted). This typically happens after the Court has accepted the case, reviewed the briefs, and sometimes even heard oral arguments. 

In practice, this means that the lower court decision—which is consistent with APA’s Amicus brief—that argued that multiple IQ scores should be viewed holistically  augmented by clinical judgement—stands. This means Mr. Smith will not be executed and, more importantly, the ruling does not upset the status of prior Atkins case law.  Given the recent direction and flavor of SCOTUS, there was a fear that SCOTUS might render a formal decision that would set back prior SCOTUS Atkins ID decisions.  So, this is a win for maintaining the principles of clinical judgement, the opinions and standards established by the established medical communities (both APA’s, AAIDD), and the need for a holistic approach to the interpretation of multiple IQ scores together with adaptive behavior.

A reading of the concurring opinions indicates that APA’s brief, as well as several key APA and AAIDD ID-related publications referenced in their decisions, were influential in the decision.  Several of the dissenting opinions reveal some troubling thinking by several justices. 

As a potential conflict of interest (COI) notice, I (Dr. Kevin McGrew), together with Dr. Joel Schneider and Dr. Cecil Reynolds (as noted on page three the APA amicus brief), were consultants to APA in the drafting of that brief.

It should be noted that as a result of the work of APA Amicus brief working committee, a peer-reviewed paper outlining a psychometrically sound approach (developed by Dr. Joel Schneider) for integrating multiple IQ scores was started during the deliberations and was recently published (Schneider, Reynolds, McGrew & Salekin, 2026) after the oral argument's.  This issue will likely to be revisited in future state cases, with Schneider et al. (2026) now elevated to a scientifically sound multiple IQ composite score method.

Research Alert: Sex Differences in Cognitive Processing Speed (Gs) Across the Lifespan: A Meta-Analysis - ProQuest

This is another quick email-based post.  

 

This is for a doctoral thesis that is not open access, although you can read some of the first pages and the abstract at the provided link.  If interested you can order and pay for a copy or, if at a university, you can request via ProQuest.  Given that Dr. Mathew Reynolds is the primary advisor, a scholar whom I hold in high esteem for his research skills, I am looking forward to seeing this in the future, most likely as a journal pub.

 

File under Gs as per the CHC taxonomy

 

Sex Differences in Cognitive Processing Speed Across the Lifespan: A Meta-Analysis - ProQuest 
https://www.proquest.com/openview/6be586738a6bfd359b6c8be220de7071/1?pq-origsite=gscholar&cbl=18750&diss=y

Research Alert: Cognitive Networks for Knowledge Modeling: A Gentle Introduction for Data‐ and Cognitive Scientists

Quick FYI email-based Research Alert.  Is open access👍

 

 

Cognitive Networks for Knowledge Modeling: A Gentle Introduction for Data- and Cognitive Scientists - Haim - 2026 - WIREs Cognitive Science - Wiley Online Library 

https://wires.onlinelibrary.wiley.com/doi/10.1002/wcs.70026

 

ABSTRACT

In this paper, we introduce the reader to the field of cognitive network science, that is, the application of network science methods to study human cognition and knowledge structures. Cognitive networks are representations of associative knowledge between concepts in a cognitive system apt at acquiring, storing, processing and producing language, that is, the mental lexicon. In a cognitive network, nodes represent concepts with links expressing relations, such as semantic, syntactic, phonological and visual connections, for example, “canine” and “dog” (nodes) linked by “being synonyms” (link). Hence, cognitive networks represent associative knowledge in mathematical, measurable and quantifiable ways. Can such structure be used to gain insights over cognitive phenomena? We explore this research question by reviewing recent, pioneering key applications and limitations of cognitive networks across visual, auditory, and semantic language processing tasks, either in healthy or clinical populations. We also review applications of cognitive networks modeling language acquisition, reconstructing text content and assessing creativity or personality traits in individuals. Our paper also gently introduces the reader to mathematical notations, definitions and measures about single-layer and multiplex networks as well as hypergraphs. Last but not least, across phonological, semantic and syntactic networks, we guide the reader through relevant psychological frameworks, datasets and software packages that might all aid current and future cognitive network scientists.

This article is categorized under:

• Psychology > Memory
• Psychology > Theory and Methods
• Linguistics > Cognitive

Graphical Abstract

Cognitive network science helps organize associative knowledge—that is, the connections between concepts. These connections play a key role in cognitive processes such as language understanding and context interpretation, even though they are not obvious in language use. For example, we do not see syntactic links, as depicted in the figure, between words in a written or spoken text. Further information can be highlighted visually in cognitive representations, such as the emotional valence of words (here indicated with the colors blue for positive, red for negative, gray for neutral and purple for a connection between positive and negative concepts). Giving structure to knowledge via cognitive networks represents a new frontier. This gentle primer offers a clear overview and introduces tools for cognitive scientists and psychologists interested in exploring cognitive representations.

Click on image to enlarge for easy viewing

 

Research Alert: Technological Devices for Developing Working Memory in Children with ADHD: A Systematic Review

Email phone-based quick FYI post.  Open access article👍

 

Technological Devices for Developing Working Memory in Children with ADHD: A Systematic Review 
https://www.mdpi.com/2673-5318/7/3/104

 

Abstract

Attention-Deficit/Hyperactivity Disorder (ADHD) is frequently associated with working-memory (WM) weaknesses that affect learning and everyday functioning. This systematic review examined the extent to which technology-delivered interventions improve WM in children and adolescents with ADHD, with primary emphasis on standardized objective WM outcomes and secondary consideration of rating-based or in-app measures. Following PRISMA 2020 guidelines, we searched PubMed, PsycINFO, Web of Science, Scopus, ERIC, and IEEE Xplore and identified 22 eligible studies spanning PC-based training, mobile interventions, AI-adaptive programs, wearables, and neurofeedback. Across modalities, the clearest near-transfer signal emerged from process-specific PC-based interventions and some AI-adaptive programs evaluated with standardized objective WM tests. Mobile and neurofeedback approaches appeared promising in some studies, but the evidence was more heterogeneous and was more often supported by ratings, in-app composites, or less rigorous designs. Overall, current evidence suggests that technology-assisted WM interventions are most promising when they are process-specific, adaptive, and delivered at a sufficient dose, although conclusions remain constrained by heterogeneity in study design, outcome type, and methodological rigor.

 

Keywords: 

attention deficit; hyperactivity disorder; working memory; computers; mobile applications; video games; neurofeedback; executive functions; cognitive training; AI; wearable technologies

Pardon typos and spelling errors-Message may be sent from iPhone and I've always had spelling problems :)

*****************************************
Kevin S. McGrew, PhD
Educational & School Psychologist
Director
Institute for Applied Psychometrics (IAP)
https://www.themindhub.com
******************************************

AI Brief: What does Grok say about Richard Snow’s aptitude trait complexes for learning?

 The following is a Grok AI generated summary when prompted (by Dr. Kevin McGrew) to explain Richard Snow’s aptitude complexes on March 7, 2026).

(Grok refers to a generative artificial intelligence developed by xAI, an Elon Musk company, as well as a foundational term in science fiction for deep understanding. Several editorial changes were made to insure accuracy, and select links added, by McGrew—to the Grok summary).

 

Richard E. Snow’s aptitude complexes (introduced in the 1980s–1990s, with key publications in 1987, 1992, and especially the 1996 chapter with Lyn Corno and Douglas Jackson) represent a major shift in how educational psychologists think about aptitude—moving far beyond the traditional narrow view of aptitude as “just cognitive ability” (e.g., IQ or a single test score).

 

Core Definition

Snow redefined aptitude as:

“the degree of readiness to learn and to perform well in a particular situation or in a fixed domain.”

It is a predisposition or propensity—a latent quality that enables more advanced performance under specific conditions. Aptitudes are situational and domain-specific: what makes someone ready to succeed in math class may be different from what works in a history seminar or a hands-on lab.  

Aptitude Complexes (the key innovation)

Snow argued that single constructs (like “fluid reasoning” or “achievement motivation”) are insufficient. Instead, success in learning comes from aptitude complexes—dynamic constellations or critical combinations of variables that work together as a coordinated system.

These complexes draw from the classic “trilogy of the mind”:

•  Cognition — abilities and processes for analyzing, interpreting, and solving (e.g., reasoning, knowledge, strategies, cognitive style, CHC abilities).

•  Affect — emotions, anxiety, self-concept, emotion regulation, personality traits.

•  Conation — motivation, volition, goal-setting, effort, persistence, will (the “want to” and “stick with it” aspects).

An aptitude complex is not just a list of traits—it is how these elements assemble and coordinate in real time within a specific task and context. They are amalgams of cognitive, conative, and affective characteristics.

 

The Two Pathways That Build Aptitude Complexes

Snow (and later Corno et al., 2002) described aptitudes developing through two parallel, interacting pathways (sometimes called the commitment pathway and the performance/action pathway):

1.  Commitment Pathway (motivational/affective/volitional)

     •  Assembles motivational resources that energize effort.

     •  Affective and volitional processes modulate how the work proceeds (e.g., regulating anxiety, sustaining intention).

     •  Outcome: Propensity (how likely the person is to engage and persist).

2.  Performance (Action) Pathway (cognitive)

     •  Assembles and deploys cognitive resources (abilities, strategies, knowledge) to do the task.

     •  Outcome: Ability/accuracy in execution.

When the two pathways coordinate effectively in a given “problem space” (the task + the broader situation/context), successful performance occurs. Over repeated experiences, these coordinated processes coalesce into stable aptitude complexes. The learner’s perception of the situation’s constraints and affordances is crucial—aptitudes are dynamic and transactional (person ↔ environment). 

Aptitudes are both input (what the student brings to learning) and product (they develop and change through education).

Snow Academic Aptitude Model (SAAM)

Snow synthesized decades of research into a provisional taxonomy (called the Snow Academic Aptitude Model or SAAM in an unpublished working white 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]) which is now available for download here. It treats academic aptitudes as a “multivariate repertoire” of readiness for general and domain-specific learning. It explicitly includes affect (personality and social-emotional skills) and conative related skills and traits (e.g., achievement motivation, goal orientation, freedom from debilitating anxiety, positive self-concept, self-efficacy, interests) alongside cognitive abilities. The model emphasizes that “there is more to knowing about an individual’s propensity for academic learning than simply knowing their IQ scores!” 

 

Why This Mattered (and Still Matters)

Snow’s work grew out of his famous Aptitude-Treatment Interaction (ATI) research with Lee Cronbach. He showed that simple aptitude scores rarely predict learning outcomes well because real-world performance depends on these richer, situation-specific complexes.

This framework directly inspired later theories (including Kevin McGrew’s 2022 CAMML model, which updates Snow’s complexes with contemporary CHC cognitive theory, Big Five personality, and self-regulated learning mechanisms while retaining the commitment/action pathway structure).

In short, Snow’s aptitude complexes gave us a whole-person, dynamic, domain-specific way to understand why two students with the same IQ can have dramatically different learning outcomes: their aptitude complexes (the coordinated cognitive + affective + conative “package” or “amalgam”) fit the demands of the task and context differently.

Snow’s legacy is summarized in the 2002 book Remaking the Concept of Aptitude (Corno et al.), which remains the definitive extension of his ideas.