Don A. Affognon- Department of Educational Psychology, University of Nevada, Las Vegas, NV, United States
Challenging the dominant view of learning as abstract content internalization, Situated Learning Theory (SLT) reframed learning as a socially mediated, context-bound process and gained influence by aligning with late-twentieth-century commitments to authenticity, participation, and identity. This paper examines the inferential structure of Situated Learning Theory as a theoretical framework, arguing that despite its descriptive and sociocultural appeal, it cannot sustain empirical scrutiny, with references to curricular, policy, professional development, and workplace contexts serving only to illustrate the downstream consequences of adopting a theory without explicit inferential constraints. Key terms such as authentic context and community of practice remain analytically vague and difficult to operationalize. The analysis applies formal tools from Signal Detection Theory, Bayesian model comparison, and information theory to identify three criteria that any viable learning theory must satisfy: (1) a rule for distinguishing learning from non-learning, (2) a mechanism for penalizing predictive error through likelihood updating, and (3) a basis for comparing competing explanations by parsimony and explanatory yield. These standards reveal that SLT offers no testable boundary conditions, no criterion for identifying failure, and no mechanism for theoretical revision. Like the Ptolemaic system that expanded epicycles to preserve its assumptions, SLT accommodates all outcomes but predicts none. This critique does not dispute the interpretive or descriptive contributions of sociocultural research, but evaluates SLT’s scientific standing in contexts where theories are expected to support discrimination, prediction, and revision. As a result, the paper positions its critique not as a terminal verdict, but as an invitation to recast SLT into a form that can be tested, refined, and applied in instructional contexts—preserving its sociocultural insights while restoring empirical accountability. This approach redefines theory advancement in education as the capacity to generate, test, and refine explanations of learning across contexts.
1 Introduction
This paper situates the emergence of Situated Learning Theory (SLT) within late-twentieth-century shifts in educational thought. Earlier cognitive and instructional theories typically treated learning as the internalization of abstract content detached from its social and material contexts, prompting a new generation of scholars to recast learning as active participation in socially meaningful activity. Brown et al. (1989) argued that cognition arises through engagement in authentic activity. Rogoff (1990) emphasized the role of culturally mediated collaboration, highlighting how shared routines and educators’ interpersonal guidance shape learning. This focus on guided participation was further elaborated in Lave and Wenger’s (1991) concept of legitimate peripheral participation, which described how learners move from observation to full engagement within shared practices. Taken together, these accounts converged on the idea that learning is not accumulated in isolation but co-constructed through structured participation in varied and important sociocultural settings. These foundational accounts established the sociocultural rationale that later crystallized into SLT.
The shift toward socially grounded models of learning coincided with broader changes in educational discourse, where concerns about authenticity, participation, and identity displaced earlier emphases on abstract transfer of learning. As this reorientation continued, SLT spread across diverse settings—teacher education, workplace training, and instructional design—precisely because it reflected several emerging priorities (Barab et al., 2012; Hodkinson et al., 2008). The appeal of key terms such as community of practice, authentic activity, and situated participation resonated with calls for context-sensitive, experientially rich approaches to learning. Empirical studies across settings—from workplace training (Billett, 2001) to everyday cognition (Lave, 1988) and classroom collaboration (Gutiérrez and Rogoff, 2003)—appeared to substantiate this participatory view. Collectively, these studies shifted attention from the transmission of content to the design of environments that foster participation, identity formation, and contextual responsiveness. As SLT gained ground within these discourses, its vocabulary became prevalent in educational theory and practice. In this paper, SLT is examined as a theoretical framework; references to instructional design, workplace learning, and educational policy serve only to illustrate how a theory’s conceptual commitments propagate into applied contexts once adopted. Yet SLT’s core concepts remained ill-defined and inconsistently applied in research and implementation.
Despite its growing influence, SLT’s conceptual vocabulary began to drift from empirical specification. For example, while community of practice appeared frequently in educational research, its meaning varied widely across studies (Engeström, 2001; Wenger, 1998). Efforts to define authentic context added further ambiguity, as instructional designers often invoked the term without specifying measurable features (Herrington and Oliver, 2000). As a result, researchers operationalized its key constructs in incompatible ways. In applied settings, participation was interpreted variously as attendance, informal engagement, or broadly defined involvement, with little alignment to observable learning outcomes (Pike et al., 2011; West and Williams, 2017; Zhao and Kuh, 2004). As these terms spread beyond their ethnographic roots, their meanings expanded to cover nearly any setting in which people interacted, regardless of instructional structure, learning objectives, or measurable change in competence (Handley et al., 2006; Hughes et al., 2013; Roberts, 2006). These inconsistencies weakened SLT’s inferential power by blurring the distinction between descriptive language and testable propositions.
SLT’s linguistic flexibility further enhanced its policy appeal, where broad resonance often outweighed explanatory precision. Once adopted in institutional documents, phrases like authentic activity and learning in context gained public prominence even as the concepts lost analytic clarity (Gulikers et al., 2004; Herrington and Oliver, 2000). Curricular guidelines and professional development initiatives began to use these terms to signal innovation, but rarely defined what “counts” as authentic or how context shapes learning (Darling-Hammond et al., 2009; Grossman et al., 2009; Zeichner, 2010). Reduced to rhetorical slogans, such phrases gave SLT the appearance of depth while obscuring how the theory might be empirically tested or updated. By the turn of the century, SLT’s visibility in policy discourse exceeded its capacity for empirical evaluation, which is a disjunction that this paper addresses.
2 Theory without constraints
SLT gained momentum amid growing dissatisfaction with instructional models that treated knowledge as decontextualized content that transferred from educator to learner without regard to origin or use. In response, new theoretical frameworks redefined learning as active participation in socially meaningful activity. Seminal texts framed cognition as a product of communal routines and guided interaction, introducing constructs such as “authentic activity” and “legitimate peripheral participation” to formalize this view (Brown et al., 1989; Rogoff, 1990; Lave and Wenger, 1991). These terms recast learning not as individual acquisition but as participation structured by practice, where development emerges through guided involvement in shared routines. As this vocabulary spread in educational discourse, it acquired double status: a conceptual orientation and a rhetorical signal of experiential relevance. In time, SLT terminology began to feature prominently in workplace training, professional development, and instructional design, despite differences in how instruction was planned, competence assessed, and outcomes defined across these domains [Hodkinson et al., 2008; Billett, 2001; see also Barab et al. (2012)]. These domains are not examined as independent objects of analysis here; they illustrate how SLT’s unconstrained concepts travel across institutional settings without acquiring additional inferential structure. Such diffusion broadened the theory’s institutional presence, but at the cost of leaving its explanatory commitments unspecified.
As the theory’s institutional appeal and standing grew, its key terms (e.g., “practice,” “participation,” and “context”) spread across policy documents, curricular frameworks, and workforce development programs, without theoretical grounding or empirical scaffolding (Gulikers et al., 2004; Darling-Hammond et al., 2009; Grossman et al., 2009; Hager and Hodkinson, 2009). These phrases appeared in course descriptions and training manuals, adopted more to convey alignment with learner-centered values than to model learning mechanisms. Programs that produced learning gains were interpreted as exemplars of situated alignment; those that failed to produce learning were explained away as insufficiently contextualized (Tummons, 2018). Across these settings, SLT functioned less as a predictive or explanatory model and more as a signaling device, endorsing participation and relevance while withholding criteria for conceptual change or outcome evaluation.
This lack of definitional constraint made operationalization difficult, since attempts to specify “situated” constructs produced no stable metrics for learning and widened the scope for interpretive adjustment instead. With no known inferential constraints, such as specifying conditions under which transfer should not occur or articulating mechanisms that distinguish surface imitation from conceptual generalization, the theory absorbed outcomes without explanation (Tuomi-Gröhn and Engeström, 2003). It provided a vocabulary for articulating instructional goals, without supplying a method for tracing conceptual development. More significantly, SLT advanced no mechanisms, adjusted no predictions, and responded to no empirical tests with conceptual revision when evidence was contrary to the theory’s claims (Fenwick, 2008).
2.1 From description to explanation
SLT presents learning as participation in communal routines, guided apprenticeship, and immersion in social practice. Its vocabulary implies process, but the structure of the theory does not accommodate explanation of a process. Often credited as its originators, Lave and Wenger (1991) defined learning as an engagement toward greater participation in communities of practice. But their account offers no specification of the conditions under which this progression might stall, reverse, or fail to produce substantive change. The settings they describe are extremely detailed, yet the theory ignores the inferential structures that allow observations to differentiate amongst alternatives. Said another way, SLT and the research and practice that follow from it do not clearly suggest when the phenomena happen, why or how they happen, or otherwise provide concrete claims that can be refuted with observed, empirical data.
Many of SLT’s advocates avoid or dismiss explanatory request by redefining critiques as epistemologically misplaced. They claim that prediction is inappropriate in learning research because learning arises within varied systems shaped by culture, identity, and historical contingency (Fenwick et al., 2012; Hodkinson et al., 2008). Proponents of SLT also suggest that the theory’s lack of causal structure reflects a deliberate epistemic stance, not a methodological flaw. The issue here is not epistemological orientation, but whether SLT, when advanced as a theory of learning, supplies the inferential commitments required for explanation, comparison, and revision. Recasting the absence of causal, predictive, or testable structure as a deliberate epistemic stance rather than a theoretical deficiency does not resolve the problem; it shifts attention away from inferential adequacy and into a discourse that evades evaluation.
As Sloman and Lagnado (2015) argue, theoretical explanation depends on a structure that narrows possible outcomes and reduces uncertainty by identifying experimentally manipulable variables. The conceptualizing around SLT has not provided such structure. It portrays participation as a process but offers no account of what specific outcomes should follow such participation or what observations could falsify its claims. According to Glymour (2001), explanation also requires taking risks by generating claims that could be proven false. SLT’s accounts are structured in ways that minimize evidentiary vulnerability—for example, by interpreting all forms of participation as indicative of learning, regardless of measurable outcome. By sidestepping constraint and omitting testable claims, the theory detaches from the logic of inference and functions as an interpretive narrative, not an explanatory model.
When learning is treated as entirely context-bound, theories lose the capacity to generalize, compare alternatives, or allow empirical test, which weakens their ability to select hypotheses, model counterfactuals, or extend predictions across settings (Giere, 1988; Strevens, 2020). Such insulation carries consequences. SLT preserves its claims by exempting them from revision. It offers no error signals, invites no counterexamples, and produces no rival models, so its structure remains static. Phrases such as “legitimate peripheral participation” and “community of practice” categorize events without any indication of what shifts, what develops, or what does not emerge. Because the theory does not account for instructional failure or stalled understanding, it cannot distinguish genuine development from the mere continuation of unproductive activity. As a result, descriptions accumulate but none of its principles shows what mechanisms support transfer, sustain acquisition, or explain variation in learning outcomes. Without these, the theory resists the pressure to evolve in response to new evidence (Campbell and Stanley, 1963; Glymour, 2001; Sloman and Lagnado, 2015).
2.2 No discrimination rule
Scientific theories must distinguish meaningful patterns from background noise (Borsboom et al., 2004; Meehl, 1990a). This requires specifying when learning has occurred, when it has not, and how confidently those judgments can be made. One framework that formalizes such distinctions is Signal Detection Theory (SDT), which models how decisions are made under uncertainty. In simple terms, SDT separates “signal” (real learning effects) from “noise” (random variability or superficial gains) by using two key concepts: sensitivity (d′), which reflects how well one can tell the difference, and decision threshold (β), which captures how willing one is to say a signal is present (Macmillan and Creelman, 2005). In education, these ideas help differentiate genuine conceptual change from rote rehearsal or test familiarity (Rouder and Haaf, 2018; Starns et al., 2012). Situated Learning Theory offers no comparable criteria. It does not define what counts as a “hit” versus a “miss,” or how to guard against false positives in interpreting student behavior. With no baseline for performance, no model of measurement error, and no account of variability across contexts, the theory lacks the inferential machinery needed to discriminate signal from noise.
This deficit compromises SLT’s empirical utility. SLT and the research and practice that follow from it do not specify when learning should occur, how it should unfold, or what observations would count as empirical refutation. This is precisely the kind of ambiguity that SDT was designed to overcome. The d′ index in SDT accounts for error or “noise” and probability, making this limitation clear by quantifying discriminability as the scaled or relative distance between the signal (e.g., genuine learning) and noise (e.g., non-learning) distributions (Macmillan and Creelman, 2005): d′ = (μ_signal − μ_noise)/σ, where σ represents the shared standard deviation. SLT offers no such variables. It identifies no average level of learning, no null condition (i.e., a defined baseline or counterfactual against which learning is judged to be absent), and no variance to assess separability. As a result, d′ is undefined in SLT, as the theory does not provide statistical basis for distinguishing success from failure. Contemporary models of inference address this challenge through signal-based formulations, such as what SDT attempts to accomplish, that define thresholds, estimate uncertainty, and enable diagnostic decisions (Kellen and Klauer, 2015; Rouder and Haaf, 2018). SLT does not possess these components and in doing so does not offer a mechanism to identify a miss, structure to quantify a false alarm, or a standard by which progress can be evaluated. By not providing a limit to what counts as learning, SLT is ineligible for empirical refinement.
As indicated earlier, error detection is more than a diagnostic or analytic feature; it is a precondition for model comparison. For a theory to be valid, it must be shown to predict or explain better than other alternative explanations or predictions. A theory that cannot say when it fails cannot justify its preference over alternative theories. In formal inference, model comparison depends not only on coherence but also on constraint. The Minimum Description Length (MDL) principle rewards models that explain data using fewer assumptions, fewer free parameters, and minimal redundancy (Cover and Thomas, 2006; Grünwald, 2007). Bayesian model comparison provides a principled way to evaluate competing theories by favoring those that make accurate predictions while remaining parsimonious—that is, simple enough to avoid overfitting but precise enough to account for observed data (Wagenmakers et al., 2010). This means that a theory’s explanatory reach must be earned by demonstrating predictive adequacy and inferential constraint, not just through compelling narratives or broad interpretive claims. SLT does not meet this standard, failing to define a compression metric—a way to quantify how well a theory reduces data into a compact, generalizable representation—nor does it detect redundancy or specify boundaries for generalization. A strong and valid social theory identifies a core process of decision-making or learning that remains coherent across contexts, enabling adaptation without conceptual instability. This consistency constrains the theory’s scope, supports generalization, and reduces the risk of redundancy, ensuring that claims about behavior are explanatory rather than interchangeable. Without such structure, theories proliferate without cumulative insight, making it difficult to distinguish novelty from noise or prediction from description. Participation, practice, and community are invoked across various domains, such as apprenticeships, design studios, reading groups, without saying how these cases are comparable. The theory’s scope expands with each example, but its likelihood remains static. It accumulates cases without increasing explanatory power (Pitt et al., 2002), which violates the MDL principle and erodes predictive accountability.
For example, consider a nursing student who performs with apparent fluency in a high-fidelity simulation lab designed to replicate real-world emergencies. The student correctly identifies symptoms, administers treatments, and communicates effectively—the hallmarks of situated competence. Yet when placed in an actual hospital with a deteriorating patient, the same student fails to recognize cues or transfer procedures. The simulation provides procedural cues embedded in controlled routines, but SLT offers no abstraction principle to help learners extract transferable knowledge from those routines. Because the theory treats context as constitutive of learning, it lacks a mechanism for projecting competence beyond its original setting. In practice, this outcome is not rare. Empirical studies have reported such breakdowns across simulation-to-clinic contexts (Issenberg et al., 2005; McKenna and Glendon, 1985). SLT treats each occurrence not as falsification, but as a narrative problem, citing inadequate contextualization or insufficient participation. The nursing student performing poorly with a real patient is not evidence that their learning was improperly designed or facilitated, but instead that they did not participate correctly or were not able to transfer from one context to another. This would be the determination rather than considering this evidence that SLT-informed instruction may not facilitate learning. In this case, the failure to perform in a real clinical setting is not treated as theoretical disconfirmation, but as a contextual mismatch, insulating the model from revision rather than exposing its limits (Popper, 1959; Meehl, 1990a,b).
More broadly, a theory that expands without constraint becomes insulated from correction. If such expansion is mistaken for inclusivity, the only outcome is rhetorical substitution with the appearance of flexibility replacing the benefits of testability (Borsboom et al., 2004; Giere, 1988). But advancing educational theory requires more than interpretive breadth. It requires tools that signal interpretive overreach, detects overfit or complexity of explanations, and invite replacement when better models are accessible. SLT theorizing and research do not make such provisions. These include formal constraints that allow rival theories to be compared, revised, or ruled out on empirical grounds. Without evaluative standards, it cannot be weighed against rival explanations, does not limit excess complexity in explanations, or is unable to diagnose inferential error (Longino, 1990). Information-theoretic approaches offer one such standard, rewarding models that minimize explanatory complexity while maximizing predictive clarity (Cover and Thomas, 2006; Myung et al., 2006). In declining to define what failure looks like, SLT exempted itself from the commitments that enable testability and revealed a structural limitation in its capacity to support core inferential operations. As Table 1 shows, the theory falls short of the inferential principles upheld by signal detection, Bayesian, and information-theoretic models.
Table 1. Inferential flaws of Situated Learning Theory across three mathematical frameworks.
3 The missing links
3.1 No transfer mechanism
What SLT lacks in inferential precision, it also lacks in functional scope. A theory of learning must do more than describe localized engagement. It must also explain how such engagement supports performance across various settings (Greeno et al., 1996). The process through which engagement produces learning outcomes must be explained through social and cognitive forces that operate across settings, independent of local variation. SLT’s core explanations overlook the functional structure required to anticipate when prior learning will generalize. Authoritative theories of learning, whether behaviorist, cognitive, or connectionist, embed mechanisms that relate prior knowledge to new contexts through structural similarity, schematic abstraction, or procedural continuity (Anderson, 2013; Barnett and Ceci, 2002; Singley and Anderson, 1989). SLT does not have an equivalent function, be it structural, computational, or causal, that connects participation in context A to performance in context B. Nor can it differentiate whether successful performance in context B stems from situated engagement or from pre-existing knowledge. The issue here is not the diversity of applied contexts, but the lack of a mechanism that links participation in one context to performance in another. By treating context as constitutive of knowledge rather than as a medium for abstraction (i.e., a way to demonstrate how a concept might manifest), the conditions that support generalization are absent from SLT. Without a principle to model how learning transfers across domains, for example, SLT lacks the functional component that educational research and instructional design require to separate context-specific success from transferable competence.
The lack of a learning transfer mechanism carries significant consequences. A viable theory of learning must model when and how knowledge generalizes across settings. SLT offers no such model. Instead, in SLT research and practice, transfer of knowing is affirmed if it occurs and reinterpreted to mean something else if it does not. The theory adapts its descriptions to fit observed outcomes, even when those observations are contrary to SLT’s explanations, rather than specifying conditions under which transfer should succeed or fail (Lobato, 2006). SLT researchers aim to identify how their findings could still fit with SLT rather than identify boundary conditions for when transfer may succeed or may fail. This violates a fundamental principle of predictive utility: explaining variation by anticipating it, not by retrofitting the narratives around it (Lobato, 2006; Rittle-Johnson and Alibali, 1999). Across domains such as workforce training, clinical education, and civic learning, SLT-based applications have produced inconsistent results, precisely because the theory does not define what counts as successful transfer (Tuomi-Gröhn and Engeström, 2003). Its focus on authentic participation prevents the identification of generalizable mechanisms. As a result, SLT cannot state under what conditions learning transfers (Perkins and Salomon, 1994). This limitation is especially clear in the earlier example of the nursing student who demonstrated competence in simulation but faltered in a live clinical setting. SLT provides no account of how procedural learning transfers across environments. To avoid the question of what underlying mechanisms enable or inhibit generalization, such discrepancies are only reframed as contextual misalignment (i.e., practicing in the wrong context) rather than inferential failure (i.e., learning to generalize performance in one context to performance in another). This interpretive maneuver relinquishes diagnostic responsibility, leaving educators without tools to anticipate or remediate breakdowns. Cognitive theories such as Cognitive Load Theory (CLT) and ACT-R retain this responsibility by modeling transfer breakdowns in terms of schema generalization failure, cue overload, and working-memory interference, each specifying how previously acquired knowledge fails to support performance in new contexts (Sweller, 1988; Anderson et al., 2004; van Merriënboer and Sweller, 2005). By defining the conditions under which learning succeeds or fails, these models provide the predictive precision and empirical testability that SLT does not supply.
3.2 No test of generalization
Scientific theories do a lot more than describe particular cases. They impose constraints that identify generalizable patterns amid contextual variation (Borsboom et al., 2004; Giere, 1988; Popper, 1968). In learning research, this requires specifying what generalizes, under what conditions, and with what reliability. SLT has not been systematically evaluated under these constraints (Greeno, 1998). Its conceptual apparatus is anchored to specific environments rather than to functional, generalized regularities. This interpretive elasticity allows SLT to span domains from classrooms to clinics to communities, but it limits systematic testing and refinement by anchoring explanations to specific contexts (Meehl, 1990a; Shavelson and Towne, 2002).
This limitation extends to instructional design. Educational interventions must operate across groups, settings, and tasks—not reproduce effects under unique conditions (Campbell and Stanley, 1963; Cronbach, 1975). Yet SLT sets no boundary conditions, probabilistic expectations, or model predictions that can be compared or falsified (Shadish et al., 2001). It embeds no means for distinguishing between competing explanations or for identifying when transfer should fail. In treating each case as self-contained, the theory circumvents the inferential responsibilities of generalization. In effect, SLT operates as a descriptive convention lacking the constraints that make prediction, abstraction, and empirical correction possible [Anderson and Lebiere, 1998; see also Wagenmakers et al. (2012)].
Taken together, the missing links outlined in this section limit the applicability of SLT in settings that require transfer beyond the initial learning context. In medical education, for example, situated simulations are often praised for their fidelity, yet learners may struggle in clinical settings when transfer mechanisms are underspecified (Issenberg et al., 2005; McKenna and Glendon, 1985). In civic education, SLT-based curricula emphasize participatory authenticity but fail to prepare students for novel policy challenges that extend beyond classroom enactments (Westheimer and Kahne, 2004). Even in vocational training, studies show that mastering context-specific tasks does not ensure adaptation when tools, supervisors, or institutional norms change (Baartman and De Bruijn, 2011). These inconsistencies reflect the lack of a formal test of generalization within the theory. Without a principle for generalization, context-bound instances are treated as post hoc narratives. What seems to succeed in one setting constitutes no basis for instructional decision-making in another. Nor can empirical variation be modeled, anticipated, or explained if the theory resists abstraction (Shulman, 1986).
4 The affective insulation of SLT
It is noteworthy that SLT’s limited predictive power is accompanied by strong affective appeal. Its primary assumptions (i.e., authentic participation, community immersion, and context-dependent knowledge) are in close alignment with contemporary ideals of inclusion, equity, and learner-centered practice (Bang and Vossoughi, 2016; Gutiérrez and Rogoff, 2003). These associations carry moral resonance, making the theory more difficult to scrutinize on empirical grounds. Whenever a theory also functions as an emblem of justice, any critics of its explanatory or predictive flaws are seen as ideological opponents (Biesta, 2007; Giroux, 2024). This insulation by sentiment shifts the burden of argument, as detractors must now defend their counterclaims as well as their values. What typically ensues is an asymmetry of critique in which empirical evaluation yields to political alignment. Even when SLT fails to predict, explain, or generalize, its terminology evokes institutional commitments that few researchers or educators are eager to oppose. Whether intended or not, this moral freight undermines the norms that lend theories their legitimacy (Cronbach and Meehl, 1955; Shavelson and Towne, 2002).
What is more troubling is that affective insulation deepens as the theory is codified in professional discourse as a signaling device. Educational programs, funding applications, and reform agendas frequently invoke SLT as shorthand for progressive pedagogy (Davies and Bansel, 2007). This invocation performs a political function: It signals alignment with reformist values rather than with scientific criteria (Luke, 2018). In this way, SLT serves as rhetorical vehicle for framing practices as socially responsive, even when these practices lack operational definitions, transfer criteria, or testable mechanisms (Reeves, 2006; Smagorinsky, 2011). As a result, the lack of formalism is reinterpreted as contextual sensitivity, even at the prohibitive cost of conflating moral alignment and theoretical adequacy. Making matters worse, research and practice tend to coalesce around symbolic alignment with reformist values, sidelining the accumulation of knowledge grounded in empirical evidence and logical inference (Shulman and Wilson, 2004; Willingham, 2012).
Such epistemic distortion corrodes the evaluative infrastructure of educational research. Theories are meant to accrue credibility by withstanding empirical scrutiny and adapting in light of disconfirming evidence (Lakatos and Lakatos, 1999; Longino, 2002). But whenever normative associations are allowed to stand in for explanatory or predictive performance, the distinction between theory and advocacy disappears. As Feyerabend (1975) has warned, the substitution of ideological alignment for evidentiary accountability turns allegiance into a proxy for argument. Within this interpretive context, affective appeal and discursive conformity displace the inferential criteria that ordinarily govern theory evaluation (Giere, 1999; Slavin, 2002).
5 Better theories, sharper tools
5.1 Which theories deliver
Educational theories that claim scientific status must impose constraints on what can be observed, predicted, or rejected. They specify constructs more precisely, link them to identifiable mechanisms, and generate predictions subject to empirical test (Wagenmakers et al., 2012). In so doing, they narrow possible outcomes, making error detectable and correction possible. The strongest frameworks in educational psychology achieve explanatory value through internal structure and subsequent refinement, not external description. The principle has long been established: Popper (1968) has argued that progress depends on theories structured enough to be wrong and specific enough to be corrected. Earlier, Cronbach and Meehl (1955) had made the same case for construct validity, showing that theories without constraint permit endless reinterpretation without resolution. This concern for constraint carried forward into the educational domain, when Shavelson and Towne (2002) extended the rationale to educational research, urging scholars to adopt models that admit error and yield consequences. In short, a theory without constraint does not support inquiry; it sustains affiliation without generating empirically testable consequences (Slavin, 2002).
CLT exemplifies how theoretical constraint supports empirical testing. It differentiates intrinsic, extraneous, and germane cognitive load, each defined by function and linked to observable effects (Sweller et al., 2011). For example, CLT predicts that split-attention formats in worked examples (e.g., diagrams placed far from their labels) will increase extraneous load and reduce learning efficiency—a claim supported by studies across mathematics, science, and medical training (Mayer, 2004; Sweller, 1988). ACT-R treats learning as the activation and strengthening of production rules over time, governed by timing functions and memory parameters (Anderson and Lebiere, 1998). This framework predicts, for example, that reaction time in procedural learning tasks (e.g., keyboard navigation or dosage calculations) will decrease logarithmically as rules become automated, a pattern it has reproduced in several domains. Dual-process models distinguish between fast, intuitive responses and slower, rule-based reasoning, and predict that learners under time pressure or cognitive load will rely more heavily on intuitive but error-prone heuristics (Evans and Stanovich, 2013). These outcomes (i.e., response time curves, transfer across domains, load-dependent performance) are tractable and replicable because the models impose mechanisms that are defined and delimited. SLT lacks such boundary conditions. Its reliance on contextual participation precludes predictions about error trajectories, transfer dynamics, or task complexity. The explanatory strength of these cognitive models derives from their ability to compress complexity into testable form.
5.2 The virtues of precision
CLT, ACT-R, and dual-process models succeed because their assumptions are explicit and their claims are testable. Each specifies what it asserts, what would support those assertions, and how that support should be observed (Shavelson and Towne, 2002). Their constructs are operationally defined and embedded in measurable systems. This precision exposes every one of these models to disconfirmation and increases their scientific value. The models also build revision into their core, because testability enables refinement and a theory that cannot be falsified cannot improve (Meehl, 1990a). In other words, their credibility rests on predictive success and error correction rather than on sentiment or alignment (Jaynes and Bretthorst, 2003; Wagenmakers et al., 2012). This evidentiary clarity carries practical value as well. Transparency allows critique and supports application: Instructional strategies derived from CLT or ACT-R can be tested and improved in practice.
Formal inference evaluates the consequences that transparency makes possible. Theories that aim to support explanation and guide design must indeed meet criteria that reward constraint, enable prediction, and discourage conceptual inflation. CLT satisfies these conditions across multiple frameworks. Under SDT it establishes thresholds for separating learning from non-learning: When instructional designs exceed working memory limits, performance declines in predictable ways (Macmillan and Creelman, 2005; Sweller et al., 2011). This allows researchers to estimate false alarms, misses, and hits in instructional outcomes. Within Bayesian model comparison, CLT supports prior expectations, such as the advantage of worked examples over unguided discovery in high-complexity tasks, and produces likelihoods that can easily be updated as data accumulate (van Merriënboer and Sweller, 2005; Wagenmakers et al., 2012). Further, competing strategies can be ranked by predictive strength. From an information-theoretic perspective, CLT reduces uncertainty by minimizing extraneous load, narrowing the design space, and reducing the explanatory burden [Cover and Thomas, 2006; see also Paas et al. (2003)]. These capacities (i.e., discriminability, revisability, and inferential economy) make the theory useful and testable. Table 2 contrasts SLT and CLT across these inferential dimensions.
Table 2. Comparative performance of SLT and CLT across three mathematical frameworks.
6 Concluding remarks
Throughout the history of science, some frameworks have preserved their credibility by expanding to accommodate inconsistencies rather than by enhancing predictive precision. The Ptolemaic system of epicycles is a case in point. To preserve geocentric assumptions, it relied on circular orbits layered atop circular orbits to account for the irregular motion of planets (Kuhn, 1997). Each new observation required a new adjustment, but none prompted a revision of the model’s core premises. In statistical terms, the system lacked priors that could be disconfirmed and likelihoods that could be recalibrated (Jaynes and Bretthorst, 2003). It did not detect error; it absorbed it instead. Rather than conclude that epicycles were not accurate means to describe planetary motion, the model retained its central premise and introduced caveats to sustain it.
Situated Learning Theory (SLT) operates under a similar logic. Instead of generating constrained predictions and revising its claims in response to empirical challenge, it assimilates contradictory findings by invoking additional context. A theory of learning that expands to explain all possible outcomes lacks the conditions required for error detection and improvement. In this sense, SLT adopts the same inferential posture as models that predated the scientific method: internally coherent, impervious to correction, and insulated from empirical testing (Meehl, 1990b). Insulation of this kind rarely persists under sustained empirical and conceptual pressure. Under mounting pressure from empirical failure and conceptual critique, the Ptolemaic system eventually disintegrated, and its collapse cleared the ground for a new standard: explanatory power through falsifiability.
That standard now confronts SLT, and this paper applies it not to discard the theory’s contributions, but to reassert the inferential obligations that a viable theory must meet. In so doing, the paper reaffirms a major requirement of scientific explanation: a theory must define what changes, why it changes, and how such change can be detected. Without this inferential core, theories such as SLT promote instructional designs that operate without evaluative constraints, assessments that conflate participation with progress, and learning environments that may not account for transfer breakdowns. These limitations exert broader systemic effects when embedded in large-scale pedagogical reforms, such as when educational innovations are adopted without mechanisms for monitoring their effectiveness, diagnosing failure, or supporting generalization across environments (Barnett and Ceci, 2002; Kirschner et al., 2006; Mayer, 2004; Penuel and Gallagher, 2017; Shepard, 2000).
SLT is often defended on the grounds that interpretive flexibility allows it to accommodate diverse learning contexts without the rigidity of formal models. Advocates may contend that constraining such a theory risks reducing its explanatory reach. Yet flexibility without testable constraints does not safeguard relevance; it displaces explanatory accountability. The most durable theories in the learning sciences—whether Cognitive Load Theory, ACT-R, or schema theory—have preserved domain generality precisely because they specify the conditions under which their predictions can fail (Anderson, 2013; Sweller et al., 2011). Such constraint has not diminished their scope; it has enhanced their resilience by making them subject to empirical correction. Without analogous inferential boundaries, SLT remains insulated from decisive evidence, leaving its interpretations consistent with every possible outcome. This is not breadth, but immunity from falsification, a condition that halts rather than advances the growth of knowledge (Meehl, 1990a,b; Popper, 1959).
A straightforward test of these inferential standards could be conducted using an experimental design comparing SLT-informed instructional interventions to those guided by a model with explicit predictive constraints. For instance, two groups of novice participants could engage in a simulated workplace training: one designed according to SLT’s emphasis on authentic participation and community engagement, the other structured by a model that specifies transfer conditions such as structural fidelity and variability. Performance could be measured not only on tasks embedded in the original context but also on structurally equivalent tasks in novel settings. If SLT offers no explicit mechanism for predicting the likelihood or pattern of transfer, its outcomes will remain indistinguishable from baseline variance, whereas the constrained model’s predictions can be directly confirmed or disconfirmed. Such a design would produce the kind of discriminating evidence needed to assess whether SLT can generate, and withstand, decisive empirical tests.
The broader implication is that educational theory thrives when its claims can be stated in terms that permit decisive tests. The inferential framework outlined here—discrimination rules, predictive updating, and comparative efficiency—offers a transferable template for evaluating not only SLT but any account of learning that aspires to guide practice. Future research could operationalize these criteria to design empirical studies that expose where theories succeed or fail in supporting generalization, from classroom-based interventions to technology-enhanced learning environments. Such a program would transform critiques into engines of theoretical progress, replacing conceptual insulation with cycles of prediction, testing, and revision. The standards proposed here link explanatory ambition to empirical discipline, equipping the field to move beyond description toward cumulative, testable understanding.
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
DA: Writing – original draft, Writing – review & editing.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
Anderson, J. R., and Lebiere, C. (1998). The atomic components of thought. Cambridge, MA: Lawrence Erlbaum Associates.
Anderson, J. R., Bothell, D., Byrne, M. D., Douglass, S., Lebiere, C., and Qin, Y. (2004). An integrated theory of the mind. Psychol. Rev. 111, 1036–1060. doi: 10.1037/0033-295X.111.4.1036
Baartman, L. K. J., and De Bruijn, E. (2011). Integrating knowledge, skills and attitudes: conceptualising learning processes towards vocational competence. Educ. Res. Rev. 6, 125–134. doi: 10.1016/j.edurev.2011.03.001
Bang, M., and Vossoughi, S. (2016). Participatory design research and educational justice: studying learning and relations within social change making. Cogn. Instr. 34, 173–193. doi: 10.1080/07370008.2016.1181879
Barab, S. A., Duffy, T. M., Land, S., and Jonassen, D. (2012). “From practice fields to communities of practice” in Theoretical foundations of learning environments. eds. D. Jonassen and S. Land. 2nd ed (New York, NY: Routledge), 29–65.
Barnett, S. M., and Ceci, S. J. (2002). When and where do we apply what we learn?: a taxonomy for far transfer. Psychol. Bull. 128, 612–637. doi: 10.1037/0033-2909.128.4.612,
Biesta, G. (2007). Why “what works” won’t work: evidence-based practice and the democratic deficit in educational research. Educ. Theory 57, 1–22. doi: 10.1111/j.1741-5446.2006.00241.x
Billett, S. (2001). Learning through work: workplace affordances and individual engagement. J. Workplace Learn. 13, 209–214. doi: 10.1108/EUM0000000005548
Borsboom, D., Mellenbergh, G. J., and Van Heerden, J. (2004). The concept of validity. Psychol. Rev. 111, 1061–1071. doi: 10.1037/0033-295x.111.4.1061,
Brown, J. S., Collins, A., and Duguid, P. (1989). Situated cognition and the culture of learning. Educ. Res. 18, 32–42. doi: 10.3102/0013189X018001032
Campbell, D. T., and Stanley, J. (1963). Experimental and quasi-experimental designs for research. 1st Edn. Chicago, IL: Cengage.
Cover, T. M., and Thomas, J. A. (2006). Elements of information theory. 2nd Edn. Hoboken, NJ: Wiley-Interscience.
Cronbach, L. J. (1975). Beyond the two disciplines of scientific psychology. Am. Psychol. 30, 116–127. doi: 10.1037/h0076829
Cronbach, L. J., and Meehl, P. E. (1955). Construct validity in psychological tests. Psychol. Bull. 52, 281–302. doi: 10.1037/h0040957
Darling-Hammond, L., Wei, R. C., Andree, A., Richardson, N., and Orphelios, S. (2009). Professional learning in the learning profession: a status report on Teacher Development in the United States and abroad. Dallas, TX: National Staff Development Council.
Davies, B., and Bansel, P. (2007). Neoliberalism and education. Int. J. Qual. Stud. Educ. 20, 247–259. doi: 10.1080/09518390701281751
Engeström, Y. (2001). Expansive learning at work: toward an activity theoretical reconceptualization. J. Educ. Work. 14, 133–156. doi: 10.1080/13639080020028747
Evans, J. S. B. T., and Stanovich, K. E. (2013). Dual-process theories of higher cognition: advancing the debate. Perspect. Psychol. Sci. 8, 223–241. doi: 10.1177/1745691612460685,
Fenwick, T. (2008). Understanding relations of individual—collective learning in work: a review of research. Manage. Learn. 39, 227–243. doi: 10.1177/1350507608090875
Fenwick, T. J., Edwards, R., and Sawchuk, P. (2012). Emerging approaches in educational research: tracing the socio-material. 1st Edn. London, UK: Routledge.
Feyerabend, P. (1975). Against method: outline of an anarchistic theory of knowledge. London: New Left Books.
Giroux, H. A. (2024). Critical pedagogy, cultural politics, and the discourse of experience. 1st Edn. London, UK: Bloomsbury Publishing Plc.
Glymour, C. N. (2001). The mind’s arrows: Bayes nets and graphical causal models in psychology. Cambridge, MA: MIT Press.
Greeno, J. G. (1998).The situativity of knowing, learning, and research. American Psychologist, 53, 5–26. doi: 10.1037/0003-066X.53.1.5
Greeno, J. G., Collins, A. M., and Resnick, L. B. (1996). “Cognition and learning” in Handbook of educational psychology. eds. D. C. Berliner and R. C. Calfee. 1st ed. (Mahwah, NJ: Macmillan Library Reference), 15–46.
Grossman, P., Hammerness, K., and McDonald, M. (2009). Redefining teaching, re-imagining teacher education. Teach. Teach. 15, 273–289. doi: 10.1080/13540600902875340
Gulikers, J. T. M., Bastiaens, T. J., and Kirschner, P. A. (2004). A five-dimensional framework for authentic assessment. Educ. Technol. Res. Dev. 52, 67–86. doi: 10.1007/BF02504676
Gutiérrez, K. D., and Rogoff, B. (2003). Cultural ways of learning: individual traits or repertoires of practice. Educ. Res. 32, 19–25. doi: 10.3102/0013189x032005019
Hager, P., and Hodkinson, P. (2009). Moving beyond the metaphor of transfer of learning. British Educational Research Journal, 35, 619–638. doi: 10.1080/01411920802642371
Handley, K., Sturdy, A., Fincham, R., and Clark, T. (2006). Within and beyond communities of practice: making sense of learning through participation, identity and practice. J. Manage. Stud. 43, 641–653. doi: 10.1111/j.1467-6486.2006.00605.x
Herrington, J., and Oliver, R. (2000). An instructional design framework for authentic learning environments. Educ. Technol. Res. Dev. 48, 23–48. doi: 10.1007/BF02319856
Hodkinson, P., Biesta, G., and James, D. (2008). Understanding learning culturally: overcoming the dualism between social and individual views of learning. Vocat. Learn. 1, 27–47. doi: 10.1007/s12186-007-9001-y
Hughes, J., Jewson, N., and Unwin, L. (2013). Communities of practice. 1st Edn. New York, NY: Routledge.
Issenberg, S. B., Mcgaghie, W. C., Petrusa, E. R., Lee Gordon, D., and Scalese, R. J. (2005). Features and uses of high-fidelity medical simulations that lead to effective learning: a BEME systematic review. Med. Teach. 27, 10–28. doi: 10.1080/01421590500046924,
Jaynes, E. T., and Bretthorst, G. L. (2003). Probability theory: the logic of science. Cambridge, UK: Cambridge University Press.
Kellen, D., and Klauer, K. C. (2015). Signal detection and threshold modeling of confidence-rating ROCs: a critical test with minimal assumptions. Psychol. Rev. 122, 542–557. doi: 10.1037/a0039251,
Kirschner, P. A., Sweller, J., and Clark, R. E. (2006). Why minimal guidance during instruction does not work: an analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educ. Psychol. 41, 75–86. doi: 10.1207/s15326985ep4102_1
Kuhn, T. S. (1997). The Copernican revolution: planetary astronomy in the development of western thought. Chicago, IL: Harvard University Press.
Lakatos, I., and Lakatos, I. (1999). The methodology of scientific research programmes. Cambridge, UK: Cambridge University Press.
Lave, J. (1988). Cognition in practice: Mind, mathematics and culture in everyday life. Cambridge University Press.
Lave, J., and Wenger, E. (1991). Situated learning: legitimate peripheral participation. 1st Edn. Cambridge, UK: Cambridge University Press.
Lobato, J. (2006). Alternative perspectives on the transfer of learning: history, issues, and challenges for future research. J. Learn. Sci. 15, 431–449. doi: 10.1207/s15327809jls1504_1
Longino, H. E. (1990). Science as social knowledge: values and objectivity in scientific inquiry. Princeton, NJ: Princeton University Press.
Macmillan, N. A., and Creelman, C. D. (2005). Detection theory: a user’s guide. 2nd Edn. Mahwah, NJ: Lawrence Erlbaum Associates.
Mayer, R. E. (2004). Should there be a three-strikes rule against pure discovery learning? Am. Psychol. 59, 14–19. doi: 10.1037/0003-066X.59.1.14,
McKenna, S. P., and Glendon, A. I. (1985). Occupational first aid training: decay in cardiopulmonary resuscitation (CPR) skills. J. Occup. Psychol. 58, 109–117. doi: 10.1111/j.2044-8325.1985.tb00186.x
Meehl, P. E. (1990a). Appraising and amending theories: the strategy of Lakatosian defense and two principles that warrant it. Psychol. Inq. 1, 108–141. doi: 10.1207/s15327965pli0102_1
Meehl, P. E. (1990b). Why summaries of research on psychological theories are often uninterpretable. Psychol. Rep. 66, 195–244. doi: 10.2466/pr0.1990.66.1.195
Myung, J. I., Navarro, D. J., and Pitt, M. A. (2006). Model selection by normalized maximum likelihood. J. Math. Psychol. 50, 167–179. doi: 10.1016/j.jmp.2005.06.008
Paas, F., Renkl, A., and Sweller, J. (2003). Cognitive load theory and instructional design: recent developments. Educ. Psychol. 38, 1–4. doi: 10.1207/S15326985EP3801_1
Penuel, W. R., and Gallagher, D. J. (2017). Creating research–practice partnerships in education. Harvard Education Press.
Perkins, D. N., and Salomon, G. (1994). “Transfer of learning” in The international encyclopedia of education. eds. T. Husén and T. N. Postlethwaite. 2nd ed (Cambridge, UK: Pergamon Press), 11:6452–6457.
Pike, G. R., Kuh, G. D., and McCormick, A. C. (2011). An investigation of the contingent relationships between learning community participation and student engagement. Res. High. Educ. 52, 300–322. doi: 10.1007/s11162-010-9192-1
Pitt, M. A., Myung, I. J., and Zhang, S. (2002). Toward a method of selecting among computational models of cognition. Psychol. Rev. 109, 472–491. doi: 10.1037/0033-295x.109.3.472,
Rittle-Johnson, B., and Alibali, M. W. (1999). Conceptual and procedural knowledge of mathematics: does one lead to the other? J. Educ. Psychol. 91, 175–189. doi: 10.1037/0022-0663.91.1.175
Roberts, J. (2006). Limits to communities of practice. J. Manage. Stud. 43, 623–639. doi: 10.1111/j.1467-6486.2006.00618.x
Rogoff, B. (1990). Apprenticeship in thinking: cognitive development in social context. New York, NY: Oxford University Press.
Rouder, J. N., and Haaf, J. M. (2018). A psychometrics of individual differences in experimental tasks. New York, NY: Center for Open Science.
Shadish, W. R., Cook, T. D., and Campbell, D. T. (2001). Experimental and quasi-experimental designs for generalized causal inference. Boston, MA: Houghton Mifflin.
Shavelson, R. J., and Towne, L. (2002). Scientific research in education. Washington, DC: National Academies Press.
Shepard, L. A. (2000). The role of assessment in a learning culture. Educ. Res. 29, 4–14. doi: 10.3102/0013189x029007004
Shulman, L. S. (1986). Those who understand: knowledge growth in teaching. Educ. Res. 15, 4–14. doi: 10.3102/0013189X015002004
Shulman, L. S., and Wilson, S. M. (2004). The wisdom of practice: essays on teaching, learning, and learning to teach. 1st Edn. San Francisco, CA: Jossey-Bass.
Singley, M. K., and Anderson, J. R. (1989). The transfer of cognitive skill. Cambridge, MA: Harvard University press.
Slavin, R. E. (2002). Evidence-based education policies: transforming educational practice and research. Educ. Res. 31, 15–21. doi: 10.3102/0013189x031007015
Sloman, S. A., and Lagnado, D. (2015). Causality in thought. Annu. Rev. Psychol. 66, 223–247. doi: 10.1146/annurev-psych-010814-015135,
Starns, J. J., Ratcliff, R., and McKoon, G. (2012). Evaluating the unequal-variance and dual-process explanations of zROC slopes with response time data and the diffusion model. Cogn. Psychol. 64, 1–34. doi: 10.1016/j.cogpsych.2011.10.002,
Strevens, M. (2020). The knowledge machine: How irrationality created modern science. 1st Edn. New York, NY: Liveright Publishing Corporation.
Sweller, J. (1988). Cognitive load during problem solving: effects on learning. Cogn. Sci. 12, 257–285. doi: 10.1207/s15516709cog1202_4
Tummons, J. (2018). Learning architectures in higher education: Beyond communities of practice. London, UK: Bloomsbury Academic.
Tuomi-Gröhn, T., and Engeström, Y. (2003). Between school and work: New perspectives on transfer and boundary-crossing. 1st Edn. Amsterdam, Netherlands: Emerald Publishing Limited.
van Merriënboer, J. J. G., and Sweller, J. (2005). Cognitive load theory and complex learning: recent developments and future directions. Educ. Psychol. Rev. 17, 147–177. doi: 10.1007/s10648-005-3951-0
Wagenmakers, E.-J., Lodewyckx, T., Kuriyal, H., and Grasman, R. (2010). Bayesian hypothesis testing for psychologists: a tutorial on the savage–dickey method. Cogn. Psychol. 60, 158–189. doi: 10.1016/j.cogpsych.2009.12.001,
Wagenmakers, E.-J., Wetzels, R., Borsboom, D., van der Maas, H. L. J., and Kievit, R. A. (2012). An agenda for purely confirmatory research. Perspect. Psychol. Sci. 7, 632–638. doi: 10.1177/1745691612463078,
Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. Cambridge University Press.
West, R. E., and Williams, G. S. (2017). “I don’t think that word means what you think it means”: a proposed framework for defining learning communities. Educ. Technol. Res. Dev. 65, 1569–1582. doi: 10.1007/s11423-017-9535-0
Westheimer, J., and Kahne, J. (2004). What kind of citizen? The politics of educating for democracy. Am. Educ. Res. J. 41, 237–269. doi: 10.3102/00028312041002237
Willingham, D. T. (2012). When can you trust the experts? How to tell good science from bad in education. 1st Edn. San Francisco, CA: Jossey-Bass, a Wiley imprint.
Zeichner, K. (2010). Rethinking the connections between campus courses and field experiences in college- and university-based teacher education. J. Teach. Educ. 61, 89–99. doi: 10.1177/0022487109347671
Keywords: educational epistemology, falsifiability, inferential constraint, Situated Learning Theory, theory testing
Citation: Affognon DA (2026) Situated learning and the problem of scientific inference. Front. Educ. 11:1725228. doi: 10.3389/feduc.2026.1725228
Edited by:
Víctor Hugo Fernández-Bedoya, Universidad Nacional Mayor de San Marcos, PeruReviewed by:
Paulo Brazão, University of Madeira, PortugalYasir Riady, Indonesia Open University, Indonesia
Copyright © 2026 Affognon. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Don A. Affognon, YWZmb2dub25AdW5sdi5uZXZhZGEuZWR1