Abstract
Psychologically distant events—those removed in time, space, people, or likelihood—pose epistemic challenges because specific details are often unknown or subject to change. Construal-level theory (CLT; Trope & Liberman, 2010) suggests that people respond to this epistemic challenge by engaging in high-level construal, mentally representing distant events by their abstract, essential features. Focusing on what is common to different potential manifestations of events allows people to leverage available information while anticipating likely variability. As events become psychologically closer, people can adaptively shift to low-level construal, representing events idiosyncratically based on the increased availability and reliability of detailed information. These ideas and related implications have been examined in a range of studies over the past three decades.
Calderon and colleagues (2026) provided an important data point in this literature, conducting a multilab test of the effect of between-subjects manipulations of distance on responses to the Behavioral Identification Form (BIF; Vallacher & Wegner, 1989). Although originally created as an individual-difference measure of action identification, the BIF was used in earlier CLT work as a measure of participants’ mental representation. The self-report scale presents a series of target actions (e.g., “paying the rent”) and asks respondents to select, for each one, one of two restatements that best captures how they might think about the action: a choice option that is more abstract (i.e., ends-related; e.g., “maintaining a place to live”) versus one that is more concrete (i.e., means-related; e.g., “writing a check”). The underlying logic is that the more that respondents are engaged in abstract representation, the more they should prefer the abstract, end-related choice options. Using this measure, Calderon and colleagues found minimal support for the prediction that distance prompts abstract, high-level construal (vs. concrete, low-level construal).
We take these findings at face value, assuming that they are substantively correct: that at the point in time and in the contexts in which these data were collected, there is little to no effect of distance manipulated across subjects on responses to this measure of action identification. In the present commentary, we consider the meaning of these findings in the context of emerging work that sheds insight into when distance-construal effects are more or less likely to be observed. We believe that this serves two simultaneous goals. First and of most direct current relevance, it offers critical context for understanding Calderon and colleagues’ (2026) findings and their implications for CLT. Second, at a broader level, it illustrates the potential advances that can come from efforts to refine and optimize the methods of a mature literature. Whether through explicit paradigm optimization or a more bottom-up process of methodological refinement over time, such efforts not only support greater methodological sensitivity but can also spur theoretical development by bringing implicit theoretical assumptions into sharper relief and generating new theoretical questions.
The Role of Methodological Refinement and Optimization in Supporting Theory
The goal of initial research on a psychological phenomenon or theoretical proposition is primarily discovery and demonstration: Researchers test an idea by developing paradigms that allow them to operationalize their variables and investigate whether the findings are consistent with their theoretical claims. We argue, however, that as a literature matures, it is critical not only to replicate key effects but also to engage in methodological refinement and optimization: an iterative, recursive process between theorizing and method testing that allows one to explore what factors affect the reliability and magnitude of a given effect with the goal of identifying features and paradigms that produce the largest and most robust effects. Such refinement is not only useful methodologically in that it spurs innovation of more reliable and sensitive paradigms, but it is also useful conceptually because observing how methodological variation influences an effect can lead researchers to refine their theorizing and curb potential theoretical overreach. This process also serves to support replication science because it can increase confidence in an effect and make replications more informative. Failed replications, for example, are much more compelling when they use methods vetted to produce the most reliable effects under current conditions.
Although initial demonstrations of an effect are highly incentivized, in our experience, efforts to explore and develop optimized methods can sometimes be dismissed in the field as “just a replication,” leading to disincentives to do such work. By contrast, in other sciences, such efforts are celebrated: Consider, for example, tests of the equivalence principle—the notion that objects of different weights fall at the same rate. Rather than drop objects of different weights off the Leaning Tower of Pisa as Galileo was reputed to have done, contemporary physicists have instead capitalized on emerging technologies to replicate these findings with substantially greater precision, such as observing the behavior of falling atoms (Asenbaum et al., 2020). 1 We argue that this sort of innovation and testing of novel paradigms to replicate established phenomena, supported by efforts to understand methods as a source of variability in the size and replicability of an effect, needs to become a more prominent feature of psychological research.
Critically, efforts to refine and optimize methods either explicitly or implicitly acknowledge the changing conditions under which research on a phenomenon may be conducted in terms of both grappling with the implications of shifting contextual conditions for existing paradigms and leveraging new technologies to enhance methodological possibilities. Indeed, changing contextual conditions may reveal important vulnerabilities in initially promising paradigms, which may help researchers to better refine both their methods and their theorizing. For example, the broad migration of research to online survey platforms may have traded off efficiency and engagement: To the degree that methods that rely on attentive participants are less efficacious in a digitized environment, researchers should update their methods to make them more engaging and/or their theorizing to conceptually consider the role of attention.
In what follows, we briefly spotlight recent work that has begun to speak to the issue of methodological refinement and optimization in the context of CLT. Note that only some of this research was conducted with methodological refinement explicitly in mind; much of it was not but is nonetheless informative. We integrate this emerging work to identify several overarching principles that together offer insight into when CLT effects are more or less likely to emerge, highlighting theoretical implications of these findings and new questions that they raise.
Principles Supporting Optimized Tests of the Effect of Distance on Construal Level
CLT proposes that distance influences individuals’ mental representations of objects or events. Because gauging mental representation directly is difficult or arguably impossible, early research used a variety of indirect approaches to explore whether there might be evidence consistent with this assertion. More recent articles have introduced advances that recognize that (a) initial studies were often guided by demonstration rather than refinement/optimization goals and (b) technological advances can be leveraged to improve methodology. Cumulatively, these newer articles offer basic principles that can be used to optimize tests of CLT, highlighting the importance of (a) capturing variation in individuals’ spontaneous construal, the key dependent variable, while engaging participants both cognitively and motivationally; (b) using within- rather than between-subjects designs; and (c) effectively anchoring distance. We briefly review each of these points below, highlighting their methodological implications and several theory-relevant questions that they raise.
Capturing spontaneous mental construal
CLT’s central prediction is that distance influences spontaneous situational construal. The BIF, as used by Calderon and colleagues (2026), is appealing as a construal-level assessment for its ease of use and directness yet has critical methodological drawbacks (Nguyen et al., 2023). First, the BIF—originally developed as an individual-difference measure—is not optimized for situational usage. 2 More critically, it fails to capture spontaneous mental representation: Participants must select one of two fixed responses (abstract vs. concrete) for each item. Unfortunately, preference for an option may be independent of abstraction. For example, if a response is irrelevant to one’s circumstances (e.g., “writing a check” for rent when one uses Venmo), participants may disfavor it regardless of their construal level. This consideration is especially important given how dated the BIF is and its potential cultural specificity.
To address these concerns, Nguyen et al. (2023) developed the “dynamic BIF.” Participants first generate their own abstract and concrete restatements, ensuring response options reflect how they might actually construe the activities. They then consider those activities in specific contexts (e.g., near vs. far) and choose between their own restatements. This approach leverages modern survey platforms, which can easily tailor surveys based on participants’ prior responses, something that was not practical in the era of paper-and-pen surveys. Using this approach, Nguyen et al. found consistent and reliable effects of temporal distance.
An additional benefit of this approach is that participants are encouraged to cognitively engage with the task; by later choosing between restatements they themselves provided, the task also has motivational relevance. Calderon and colleagues (2026) reported a high rate of attention-check failure (incorrect responding about condition), suggesting a lack of engagement with the traditional BIF. This is especially striking for the temporal studies, in which the manipulation repeated on every item: That more than 20% failed to correctly identify their condition raises concerns about how engaged participants more generally were with the task.
Other approaches might be even more effective in capturing spontaneous construal because they allow individuals to describe events as they themselves conceptualize them in the moment. Natural language processing (NLP), for example, codes participants’ descriptions of events at scale. 3 Early NLP approaches indexed concreteness at the word level (using dictionaries of norms from Brysbaert et al., 2014), replicating distance-construal effects in people’s spontaneous communications in real-world settings (e.g., Twitter messages; Bhatia & Walasek, 2016; Snefjella & Kuperman, 2015). However, word-level metrics may not capture concreteness at the sentence level (Johnson et al., 2020; Yeomans, 2021); newer articles have leveraged more advanced approaches. Le, Hildebrand, et al. (2026) showed that although dictionary-based analyses using the Brysbaert et al. (2014) approach revealed predicted differences of temporal distance (especially using within-subjects designs; see following point), judgments made by human raters revealed effects almost 10 times larger. A machine-learning algorithm was able to mirror these judgments, suggesting scalable implementation. Advancement in neural networks and large language models can also enhance NLP approaches (Gamoran et al., 2024; Le, Hildebrand, et al., 2026). For instance, Gamoran et al. (2024) trained a neural network to code narratives that varied across three distance dimensions and found that increases in a single dimension of distance increased abstraction and that abstraction increased monotonically when events were distanced on more than one dimension.
NLP, however, relies on text data, and the act of communication itself may modify mental representation. Levit et al. (2025) instructed participants to visualize events and evaluate the detail of their own mental representations. Across psychological distances, people rated their mental images of near versus distant events as more detailed. Adapted from reality-monitoring theory (Johnson et al., 1988), this approach complements text coding because it requires no mediation by language.
Cumulatively, the above findings highlight important conceptual and methodological insights. First, CLT makes predictions about spontaneous mental construal. Although other approaches may at times show effects, measures that capture variation in how people describe events in their own terms offer greater fidelity to the theory. As measures get more removed from this—such as when they ask participants to react to experimenter-provided possibilities or ask people about aspects about the context they may or may not have considered—the theoretical grounding of the prediction is more tenuous. Second, engagement with the task at hand is important for both pragmatic and theoretically relevant reasons. Conceptually, CLT has largely been approached as a cognitive theory. However, the underlying epistemic logic is motivational: People turn to abstract construal to represent distant events because they want to connect to such events. In the absence of desired connection, construal effects should be attenuated, an idea that most of the construal literature has not considered. Although Nguyen et al. (2023) did not set out to study this issue directly, in considering the contrast between their methods and the traditional BIF, we note that the dynamic BIF inherently engages individuals more in the task and may be experienced as more relevant to motivation. Future research should consider more definitively how motivational relevance—whether task-related, personality-dependent, or as affected by the larger environment—might moderate effects of distance on construal.
Within-persons versus between-persons effects and the importance of contrasts
CLT argues that people respond to the epistemic challenges posed by changes in distance by modulating their level of mental representation, representing events in more high-level terms as they increase in distance and in more low-level terms as they increase in proximity. Tests of CLT have often adopted between-subjects designs in which participants consider distant or near events. The favoring of between-subjects designs is somewhat surprising given considerable individual-level variation in abstract thinking (Vallacher & Wegner, 1989; Yin et al, 2025). Theoretically, what is most relevant is not differences across people but differences within people: The same event is construed differently by the same person as a function of distance. Accordingly, within-subjects investigations of CLT not only provide greater statistical power but also enhance construct validity by better modeling the phenomenon. Indeed, within-subjects tests of CLT are considerably more robust and reliable than between-subjects tests (e.g., Grinfeld et al., 2024; Le, Hildebrand, et al., 2026; Nguyen et al., 2023).
Note that what is critical may not be the within-subjects design itself but providing people with freedom to move around a general tendency. When asked to report their construal, people arguably have two impulses: to share their general way of thinking (individual differences) and their specific manner of thinking in the given context (situational variation). Within-subjects designs address this tension by allowing participants to express an overall tendency even as they vary their responses by context. Between-subjects designs can also prompt this freedom by providing contextual contrast. Grinfeld et al. (2024) demonstrated this in the context of hypotheticality. Whereas a within-subjects design showed an effect on the BIF, a traditional between-subjects design did not. However, the predicted CLT effect emerged between subjects after incorporating a contextual contrast (e.g., “Consider a hypothetical situation rather than a real one”), mimicking a within-subjects structure. Including a contrast thus appears to be critical, encouraging participants to respond with their contextualized rather than general tendency.
The above findings are not simply a methodological curiosity but raise important conceptual issues. The logic of CLT suggests that people adjust their construal in response to greater versus lesser distance, consistent with a within-subjects effect. Within-subjects designs and between-subjects designs that provide a contextual contrast each lead to more robust construal findings. This latter point is intriguing but not yet precisely understood. What does a contextual contrast provide that affords construal effects? One possibility, alluded to above, is that it signals to individuals to respond with their contextualized rather than generalized tendencies. Building on this intuition, future research might explore factors beyond basic trait-level differences that lead people to report more general abstract versus concrete construal. Perhaps, for example, the desire to project an image of being a “big-picture” person versus a “detail-oriented” individual affects how people respond. Alternatively, as we discuss in more detail below, it is plausible that a contextual contrast “fixes” the focal distance point in a more reliable fashion than when such a contrast is not included. Finally, given recent work suggesting that many people hold accurate lay beliefs about the benefits of high-level and low-level construal (Le, Nguyen, et al., 2026; Nguyen et al., 2019), it is possible that by making distance more salient, a contextual contrast increases the likelihood that people will apply this knowledge in their construal of events. This would be broadly consistent with CLT while representing a new avenue for further exploration.
Reliable invocation of near and far
As alluded to earlier, an additional key methodological issue highlighted by recent research is the challenge of reliably establishing what constitutes “near” versus “far.” Distance points (e.g., “tomorrow”) are often picked arbitrarily; yet these can dramatically affect resultant effect sizes. For example, Le, Nguyen, et al. (2026) examined people’s judgments of the usefulness of high-level versus low-level construal for thinking about events that spanned an array of time points from today to 10 years from now. The relative advantage of using high-level versus low-level construal was negligible when distance from instantiated as 1 week from now but steadily increased before reaching an asymptote at 5 years from now. In other words, the size of the effect changed dynamically depending on the time point used, spotlighting it as a critical decision in the optimization of temporal-distance effects. Calibrating across distance dimensions is even more problematic (e.g., Is “1 year” equivalent to “1 mile”?), making comparing effect sizes across dimensions conceptually difficult.
Situational factors can also affect the interpretation of distance. Grinfeld et al.’s (2024) work demonstrates the utility of adding a distance contrast. In addition to highlighting the contextualized nature of the task, introducing a contrast provides an essential anchor for the distance manipulation. For example, contrasting “tomorrow” with “a year from now” renders it close, whereas contrasting it with “today” might render it far. Explicitly adopting contrasts ensures participants conceptualize near and far points similarly, controlling for noise. This is especially important because technology and culture continuously shift what is understood to be near and far: Studies must reliably operationalize distance within a given cultural and historical context.
The issue of distance calibration not only represents a methodological concern but also highlights important theoretical questions. CLT introduced the term “psychological distance” to provide a unifying way of considering effects of different dimensions that make something more or less part of one’s current experience; this represented a critical advance in the landscape of a literature that had earlier approached these dimensions as largely distinct. However, empirical research on CLT primarily manipulated distance by varying objective distance, and only a handful of articles have attempted to quantify either how objective distance translates to subjective distance or the functional form of the relationship between distance and construal (e.g., Snefjella & Kuperman, 2015; Soderberg et al., 2015). This work has generally adopted a between-subjects approach to this question and looked for regularities across participants. However, methodological appreciation of the importance of variability in distance perception raises the question of whether the objective/subjective translation of distance is better considered at an individual-difference level. In addition, to the degree that distance activates construal because it poses an epistemic challenge, it is critical to consider not only subjective distance per se but also the expectation of variability that comes with it because this is likely to be the important trigger of shifts in construal.
Conclusion
A developing body of work on when CLT effects are more likely to emerge provides crucial context for understanding and drawing conclusions from Calderon and colleagues’ (2026) findings. Distance-construal effects are more reliably obtained under conceptually and methodologically sensitive conditions; they are fickle when these conditions are unmet. It may be tempting for readers unfamiliar with these recent efforts to conclude that Calderon and colleagues’ findings suggest that studying the association between distance and construal is a dead end, a line of thinking that is not worthy of further investigation. Instead, we hope that these findings coupled with those we review here spur renewed interest in CLT’s core ideas and advance a new set of related theoretical questions.
More generally, we stress the importance of research that updates and improves methods used to test theory, especially as a literature matures; such efforts push researchers to think deeply not only about methods but also the underlying theory. From this broader perspective, although the principles for optimizing tests of CLT may or may not generalize to other theories, we believe they serve as a case study of the value that such efforts can offer. Furthermore, the studies we review here were often bottom-up in their approach. Although we integrate these to extract several basic principles that we believe emerge from this work, we encourage scholars to consider how to best conduct future work of this nature in a systematic and programmatic manner.
