Almaatouq et al. view the purpose of research is to map variable-to-variable relationships (e.g., the effect of X on Y). They also view theory as this mapping of variable-to-variable relationships rather than an explanation of why the relationships occur. However, it is theory as explanation that allows us to reconcile disparate findings and that should guide application.

Almaatouq et al. propose an “integrative approach” to increase the generalisability and commensurability of experiments. Yet their metascientific approach has one glaring omission (and misinterpretation of) – the role of sample diversity in generalisability. In this commentary, we challenge false notions of subsumed duality between contexts, population, and diversity, and propose modifications to their design space to accommodate sample diversity.

The integrative experiment design proposal currently only relates to group results, but downplays individual differences between participants, which may nevertheless be substantial enough to constitute a relevant dimension in the design space. Excluding the individual participant in the integrative design will not solve all problems mentioned in the target article, because averaging results may obscure the underlying mechanisms.

Almaatouq et al. argue that an “integrative experiment design” approach can help generating cumulative empirical and theoretical knowledge. Here, we discuss the novelty of their approach and scrutinize its promises and pitfalls. We argue that setting up a “design space” may turn out to be theoretically uninformative, inefficient, and even impossible. Designing truly diagnostic experiments provides a better alternative.

Almaatouq et al. propose an integrative experiment design space combined with large samples for scientific advancement. We argue recent innovative designs combining closed-loop experiment designs and Bayesian optimisation allow for integrative experiments at an individual level during a single session, circumventing the necessity for large samples. This method can be applied across disciplines, including developmental and clinical research.

Almaatouq et al. propose a novel integrative approach to experiments. We provide three examples of how unaddressed measurement issues threaten the feasibility of the approach and its promise of promoting commensurability and knowledge integration.

Almaatouq et al.'s prescription for more integrative experimental designs is welcome but does not address an equally important problem: Lack of adequate theories. We highlight two features theories ought to satisfy: “Well-specified” and “grounded.” We discuss the importance of these features, some positive exemplars, and the complementarity between the target article's prescriptions and improved theorizing.

We disagree with Almaatouq et al. that no realistic alternative exists to the “one-at-a-time” paradigm. Seventy years ago, Egon Brunswik introduced representative design, which offers a clear path to commensurability and generality. Almaatouq et al.'s integrative design cannot guarantee the external validity and generalizability of results which is sorely needed, while representative design tackles the problem head on.

Integrative experimentation will improve on the status quo in empirical behavioral science. However, the results integrative experiments produce will remain conditional on the various assumptions used to produce them. Without a theory of interpretability, it remains unclear how viable it is to address the crud factor without sacrificing explainability.

To succeed, we posit that research cartography will require high-throughput natural description to identify unknown unknowns in a particular design space. High-throughput natural description, the systematic collection and annotation of representative corpora of real-world stimuli, faces logistical challenges, but these can be overcome by solutions that are deployed in the later stages of integrative experiment design.

This commentary argues against the indictment of current experimental practices such as piecemeal testing, and the proposed integrated experiment design (IED) approach, which we see as yet another attempt at automating scientific thinking. We identify a number of undesirable features of IED that lead us to believe that its broad application will hinder scientific progress.

Integrative experiment design is a needed improvement over ad hoc experiments, but the specific proposed method has limitations. We urge a further break with tradition through the use of an enormous untapped resource: Decades of causal discovery artificial intelligence (AI) literature on optimizing the design of systematic experimentation.

The target article argues researchers should be more ambitious, designing studies that systematically and comprehensively explore the space of possible experiments in one fell swoop. We argue that while “systematic” is rarely achievable, “comprehensive” is often enough. Critically, the recent popularization of massive online experiments shows that comprehensive studies are achievable for most cognitive and behavioral research questions.

Generalizations strengthen in traditional sciences, but in psychology (and social and behavioral sciences, more generally) they decay. This is usually viewed as a problem requiring solution. It could be viewed instead as a law-like phenomenon. Generalization decay cannot be squelched because human behavior is metastable and all behavioral data collected thus far have resulted from a thin sliver of human time.

We comment on the limits of relying on prior literature when constructing the design space for an integrative experiment; the adaptive nature of social and behavioral phenomena and the implications for the use of theory and modeling when constructing the design space; and on the challenges of measuring random errors and lab-related biases in measurement without replication.

We expect that consensus meetings, where researchers come together to discuss their theoretical viewpoints, prioritize the factors they agree are important to study, standardize their measures, and determine a smallest effect size of interest, will prove to be a more efficient solution to the lack of coordination and integration of claims in science than integrative experiments.

Dominant paradigms in science foster integration of research findings, but at what cost? Forcing convergence requires centralizing decision-making authority, and risks reducing the diversity of methods and contributors, both of which are essential for the breakthrough ideas that advance science.

The target article ignores the crucial role of correlational methods in the behavioral and social sciences. Yet such methods are often mandated by the greater complexity of the phenomena investigated. This necessity is especially conspicuous in psychological research where its position in the hierarchy of the sciences implies the need for both experimental and correlational investigations, each featuring distinct assets.

The authors rightly critique existing social sciences approaches. However, they are too quick to dismiss the criticism that their proposed paradigm is atheoretical. Social and cognitive theories are indeed incommensurate, often due to the lack of a unifying framework. Without proper integration with theoretical frameworks, their proposal may merely produce a resource-intensive veneer of thoroughness without substantive improvements to understanding.

We affirm the utility of integrative modeling, according to which it is advantageous to move beyond “one-at-a-time binary paradigms” through studies that position themselves within realistic multidimensional design spaces. We extend the integrative modeling approach to a target domain with which we are familiar, the consequences of bilingualism on mind and brain, often referred to as the “bilingual advantage.” In doing so, we highlight work from our group consistent with integrative modeling.