Illuminating the relative dominance of awareness and pervasiveness over visibility in organizational ICT affordances

Abstract

Contemporary discourse on the affordances of organizational information and communication technologies (ICTs) has largely been captivated by the allure of visibility. This article challenges that glare by elucidating the overlooked yet pivotal roles of a set of other organizational ICT affordances. Through a dominance analysis, our findings illuminate that awareness—the capacity of ICTs to link information and actors in an ongoing digital tapestry—and pervasiveness—the widespread nature, across time and space, of digital content and interactions—hold greater explanatory power compared to visibility in understanding some types of interactions fostered by ICTs (communication frequency, information-sharing quality at work [within and across departments], and identity processes [departmental and organizational]). By spotlighting the explanatory strength of affordances such as awareness and pervasiveness and somewhat dimming the role of visibility, this study urges scholars and practitioners alike to broaden their focus on the affordances of media in the digital workplace.

Contemporary work environments are rife with information and communication technologies (ICTs) that have been developed, diffused, and incorporated into organizational processes and members’ lives. In this increasingly digital world of work, scholars have attempted to make sense of how organizational ICTs influence organizational dynamics and individual outcomes (e.g., Ellison et al., 2015; Leidner et al., 2018; Leonardi & Vaast, 2017). One recent approach to understanding the uses and effects of ICTs is the affordances framework (e.g., Carr, 2020; Evans et al., 2017). Research has identified (Treem & Leonardi, 2013), conceptualized (Evans et al., 2017), and operationalized various affordances of ICTs (Fox & McEwan, 2017; Rice et al., 2017), and examined the consequences of affordances beyond the uses of specific ICTs (Chen et al., 2020; van Zoonen et al., 2023). The affordance perspective argues that to robustly theorize about the mediated organizational communication environment, we need to look beyond considerations of particular channels and set our sights on the ways in which ICTs may afford different people different possibilities for communicative action (Carr, 2020; Flanagin, 2020).

Despite the identification of numerous affordances, including association, awareness, editability, evaluability, persistence, personalization, pervasiveness, replicability, scalability, searchability, sharing, signaling, and visibility (Evans et al., 2017; Fox & McEwan, 2017; Manata & Spottswood, 2022; Rice et al., 2017; Scarlett & Zeilinger, 2019; Treem & Leonardi, 2013), among others, research has spotlighted the affordance of visibility. First, initial work on communication visibility theory (Engelbrecht et al., 2019; Leonardi, 2014, 2015; Treem et al., 2020; 2024) identified how technologies present possibilities to obtain metaknowledge through both message transparency and network translucence. In addition, Baym and Ellison (2023) argued that visibility emerges as one of the four central themes for understanding technology-mediated communication dynamics and the future of work.

Studies using communication visibility labeled this affordance “a hallmark of computer-mediated communication” (Treem et al., 2020, p. 44), the “root affordance” (Flyverbom et al., 2016, p. 101; Lane et al., 2018; van Zoonen et al., 2023), and “central to contemporary digital developments” (Flyverbom, 2022, p. 4). Furthermore, research has scrutinized the ways in which increased digitization gives rise to new forms of behavioral visibility (Lane et al, 2018; Leonardi & Treem, 2020), even leading to concerns that our lives have become “overlit” (Flyverbom, 2022, p. 1). Thus, it is no surprise that studies have firmly put the study of visibility in their crosshairs (e.g., Chen et al., 2020a; 2020b; Engelbrecht et al., 2019; Stohl et al., 2016; Treem et al., 2020; 2024; van Osch & Steinfield, 2018; van Zoonen & Sivunen, 2023). Indeed, Table 1 shows that “visibility” has received the bulk of coverage of affordances in social science (49.5%) and business/management (68.4%) journals. So, the current application of an affordance framework in organizational communication research tends to disproportionally gaze upon visibility and thus dim the glow of other affordances.

Instead, our inquiry focuses on the relative importance of these affordances, making several contributions. First, we argue that the prevailing highlighting of visibility as the premier affordance is incomplete; that approach is more like a single-frequency laser than the full-spectrum sun. We summarize a central set of affordances of organizational ICTs based on Rice et al. (2017). Then, applying dominance analysis, we rank them according to their contribution to explaining mediated communication frequency, information-sharing quality, and departmental and organizational identification. Our analysis offers robust evidence that challenges and thus expands upon conventional wisdom, which holds that visibility is the brightest affordance in the rainbow of organizational communication.

Second, by identifying which affordances are more or less associated with social- and task-related outcomes in an organizational setting (ICT use, information sharing, and identification), the results can inform organizations about how different affordances may have both overlapping and distinct implications. A better understanding of which affordances matter for what outcomes may help organizations overcome the challenge of balancing a complex and interdependent set of ICT affordances in pursuit of organizational goals (Safadi, 2024).

Finally, from a methodological perspective, while “dominance analysis is one of the most utilized relative importance metrics in organizational sciences” (Braun et al., 2019, p. 593), its application in communication research is infrequent. By describing and applying this methodological approach, we hope to familiarize communication scholars with the currently recommended method for assessing relative importance in regression models.

Theoretical framework

Affordances

Organizational ICTs, including enterprise social media, increasingly afford a rich set of communication, sharing, and socialization possibilities (Rice et al., 2017; Volkoff & Strong, 2013). To better understand how technologies are perceived and used in organizational settings, scholars have increasingly adopted a socio-material perspective to escape the confines of cognitivist and deterministic models (Faraj & Azad, 2012). Central to this idea is Gibson’s (1979) theory of affordances, which, when applied to organizational ICT use, emphasizes the interdependencies between the potentialities of technologies and users’ perceptions and goals in understanding how people interact with, use, and experience technology (Leonardi, 2013; for a recent analysis and review, see Ronzhyn et al., 2023). These affordances can broadly be understood as possibilities for action and have been used to understand individuals’ interactions with technologies across a variety of disciplines, including communication studies (Treem & Leonardi, 2013), management studies (Leonardi & Vaast, 2017), information systems (Safadi 2024), and organizational behavior (van Zoonen et al., 2021).

For this study, we apply the concept of organizational media affordances, defined as “relationships among action possibilities to which agents perceive they could apply one or more media within its/their potential features/capabilities/constraints, relative to the agents’ needs or purposes, aggregated within or across media contexts, and within or across organizational contexts” (Rice et al., 2017, p. 110). Rather than treating ICTs as separate technological objects, the affordance approach offers greater opportunities to compare results across different ICTs and contexts and highlights the notion that any individual may experience a varying combination of different affordances associated with organizational ICTs (Faraj & Azad, 2012).

Relative importance of affordances

Ronzhyn et al. (2023) concluded that “The four-item taxonomy of social media affordances (visibility, editability, persistence, association) introduced by Treem and Leonardi (2013) dominates the scientific discourse” (p. 3173). However, as Table 1 shows, scholars have firmly directed their view to one affordance in particular—visibility (Engelbrecht et al., 2019; Leonardi, 2014; 2015; van Zoonen & Sivunen, 2023). While we do not dispute the importance of visibility, we suggest that adjusting our conceptual and empirical lenses to a more complete understanding of affordances may advance scholarship on how individuals in organizations use technologies with what implications. Studies taking a quantitative approach (many affordance studies are qualitative) have relied on regression models to identify the implications of various sets of affordances (with a specific focus on visibility). The question of interest becomes understanding the extent to which each affordance is relatively more important (i.e., explains variance) to a criterion outcome. Broadly, relative importance “refers to the contribution a variable makes to the prediction of a criterion variable by itself and in combination with other predictor variables” (Tonidandel & LeBreton, 2011, p. 2). Assessing the relative strength of explanatory variables or predictors on outcome criteria can have important theoretical and practical implications (Braun et al., 2019). For example, a manager may want to know whether social media use (i.e., x₁) is more or less important than in-person meetings (i.e., x₂) in explaining or predicting team performance (i.e., y).

Outcome criteria validity

Some affordances may matter relatively more than others, depending on the criteria of interest. If organizations are interested in improving a given criterion, it matters how the organization and users engage with ICT affordances that are relatively more appropriate and useful for achieving that goal. Further, to assess the relative contributions of affordances, it is helpful to analyze their associations with a diverse set of relevant outcomes to determine whether those rankings are generalizable. In particular, organizational ICTs can “afford a rich set of communication, sharing, and socialization capabilities” (Safadi, 2024, p. 347). Thus, our goal is not to test a model of the effects of affordances on outcomes; rather, it is to assess the relative influence of central affordances and whether the ordering is consistent across a small set of outcomes (i.e., criterion validity).

First, central to organizational functioning is communication, and thus organizational ICT use. Much of the media affordance literature focuses on how media can facilitate or constrain such communication (Fox & McEwan, 2017; Hampton, 2016). To the extent that organizational ICT affordances make it easier, more effective, or more efficient to communicate with one or more organizational members or groups, the more those members may use those ICTs. Thus, we are interested in how different affordances are associated with the frequency of organizational ICT use.

Second, sharing is examined by evaluating information-sharing quality. Much research emphasizes the crucial contribution to organizational success of ICTs by facilitating timely, accurate, and relevant knowledge and information awareness, exchange, sharing, transfer, brokering, or re-use (Ellison et al., 2015; Engelbrecht et al., 2019; Leonardi, 2014, 2015; Majchrzak et al., 2013; van Zoonen & Sivunen, 2023). Such processes may occur at multiple levels (dyad, group/team, department, organization) and may be more or less targeted. Notably, communicating with co-workers in the contexts of within and across departmental boundaries is crucial as organizations become more complex, virtual, networked, distributed, and project-oriented (Boswell & Olson-Buchanan, 2007; Chen et al., 2020; Ellison et al., 2015; Leonardi, 2013; Leonardi & Vaast, 2017; Rice, 2017; Stohl et al., 2016; Van Osch & Steinfield, 2016; van Zoonen et al., 2020; van Zoonen & Sivunen, 2023).

Third, socialization is evaluated by scrutinizing the role of affordances in relation to organizational identification, defined as “the perception of oneness with or belongingness to an organization” (Mael & Ashforth, 1992, p. 104), and generally considered one implication of socialization. Positive or strong identification has a wide variety of employee implications, such as job satisfaction, citizenship behaviors, and work performance (Weisman et al., 2023). In particular, Weisman and colleagues review research on both positive and negative relationships of organizational communication channels and digital transformation with organizational identification. In the context of fluid, fragmented, and mediated organizational forms and working relationships noted above (e.g., Leonardi, 2013; Volkoff & Strong, 2013), it may become more difficult for workers to become socialized and thus identify with their department or organization, and achieve departmental or organizational goals. Thus, organizational ICT affordances may be associated with both departmental and organizational identification (Chen et al., 2020; Hampton, 2016; Leidner et al., 2018; Rice, 2017; van Zoonen et al., 2020).

Research question

Establishing the relative strength of predictors in a regression model is an important empirical question for informing subsequent theory and selecting the subset of predictors that optimize criterion-related validity. For example, if we are interested in improving job satisfaction, a researcher may select a set of predictors, including various job characteristics such as autonomy, feedback, task significance, or skill variety. If autonomy affects job satisfaction more strongly within the context of any subset of the remaining predictors, then efforts to strengthen autonomy would be an effective use of resources. Organizational ICT affordances are expected to influence important organizational processes related to how people connect and collaborate. Hence, we first identify a parsimonious and central set of affordances and analyze their relative importance to organizational outcomes related to how people connect and collaborate through ICTs—frequency of mediated communication, information-sharing quality, and identification. As such, we examine whether the current emphasis on visibility may require recalibration.

RQ1: What is the relative importance of different organizational ICT affordances in explaining (a) mediated communication frequency, (b) information-sharing quality within departments, (c) information sharing across departments, (d) departmental identification, and (e) organizational identification?

Method

Procedure

This study relies on two cross-sectional samples of employees of a Finnish public-sector broadcasting company. A survey link was sent to all employees with an invitation letter, organizational authorization and contact, and a link to the company’s intranet for a study description. After two weeks and two reminders, we obtained, for Sample 1, N = 460 (response rate 13.6%) in October 2015. In early 2018, we repeated the process, obtaining Sample 2, N = 279 (8.2%). Utilizing the two samples allows for an assessment of the replicability and validity of the findings. As several respondents (n = 87) responded to surveys 1 and 2, we examined whether the inclusion or exclusion of these respondents affected the outcomes; it did not, so we used the full sample.

Samples

Sample 1 Respondents were, on average, 49 years old (SD = 9.17) and reported working 38.52 hours per week (SD = 4.79). The average organizational tenure was 19.73 years (SD = 10.16), and 54% were female. In total, 49.6% reported having obtained a university degree, 27.1% an applied science degree, and 23.1% obtained some vocational education degree. Sample 2 Respondents reported being 48 years old (SD = 8.66) and working 39 hours per week on average (SD = 11.41). They indicated organizational tenure of 18.09 years (SD = 10.15). Slightly over half the respondents were female (51%) and most obtained either a university degree (45.5%) or an applied science degree (26.6%).

Measures

Table 2 provides the measures used in both samples in this study, with corresponding factor loadings. All statements were answered using a 1 = strongly disagree to 7 = strongly agree scale unless indicated otherwise.

Affordances

We examined six affordances of organizational ICTs previously iteratively developed, operationalized, and validated by Rice and colleagues (2017; Table 1, p. 113), with their subdimensions validated and conceptually defined by Manata and Spottswood (2022; see Table 1, p. 1325; Table 2, p. 1327; and Table 4, p. 1130). These were indicated by 21 items representing the affordances of visibility, editability, self-presentation, awareness, pervasiveness, and searchability. For each item, respondents were instructed: “Think about the extent to which you agree that these activities are currently possible (whether you actually do them or not), using the various media (email, phones, instant messaging, intranet, social media, etc) available at [organization]. Throughout, “others” and “people” refer to current employees of [organization name].”

Mediated communication frequency

Mediated communication frequency was measured by asking respondents to indicate the frequency with which they used six organizational ICTs available to them within the organization: email, telephone, chat messages, audio-conference calls, video conference calls, and enterprise social media (1 = never to 9 = many times a day; Boswell & Olson-Buchanan, 2007; Park et al., 2011).

Information-sharing quality

Information-sharing quality was assessed by adopting four items from Fonner and Roloff (2010). We inquired about information-sharing quality first within the respondent’s department and then across the organization’s departments.

Identification

Two levels of identification were examined: departmental identification and organizational identification. Four items were adopted from Mael and Ashforth (1992) to examine each identification, respectively.

Analysis

We first examine the reliability and validity of our measurement instrument through confirmatory factor analysis. The goodness of fit was examined by inspecting the χ² and degrees of freedom, the Tucker-Lewis Index (TLI), and the Root Mean Square Error of Approximation (RMSEA). TLI values of .90 and .95 indicate acceptable and excellent model fit, whereas RMSEA values less than .05 and .08 reflect a close and reasonable fit to the data (Hu & Bentler, 1999; Marsh et al., 2004).

Second, we employed dominance analysis (DA: Azen & Budescu, 2003) to investigate the relative importance of the six organizational ICT affordances in explaining communication frequency, information-sharing quality, and identification. We used the ‘Domin’ Stata implementation developed by Luchman (2021). DA is ideally suited to establish relative importance among predictors in a multiple regression model (Braun et al., 2019). Notably, while the relative importance of predictors in a regression equation is frequently inferred from standardized regression weights, bivariate correlations, or even p-values (Budescu & Azen, 2004), such approaches are limited, wrong, and may lead to mistakes (Budescu, 1993). To overcome these limitations, DA is the current recommended approach to establish relative importance (Braun et al., 2019).

The key benefit of DA is that it provides a more comprehensive measure of predictor importance by considering all possible model combinations and accounting for shared variance and multicollinearity (Azen & Budescu, 2003; Budescu, 1993). This is important as the typical assumption that regressors are uncorrelated is generally untenable (Braun et al., 2019). We apply DA to examine relative importance, defined as “the proportionate contribution each [independent variable] makes to R², considering both its direct effect (i.e., its correlation with the [dependent variable]) and its effect when combined with the other [independent] variables in the regression equation’’ (Johnson & LeBreton, 2004, p. 240). This is a meaningful distinction from alternative approaches because it involves comparing each p predictor’s incremental value across all 2^p − 1 sub-models that involve that predictor and using the set of incremental values that result (Braun et al., 2019). In other words, in conditions of intercorrelated predictors, DA demonstrates that “a predictor X_i, can be unequivocally judged to be more important than its competitor, X_j, only if it outperforms it in each of the 2^(p-2) subset models” (Budescu & Azen, 2004, p. 343).

We used the standardized general dominance statistic (as opposed to conditional or complete dominance statistics; Luo & Azen, 2013). Standardized dominance is the percent of the overall explained variance allocated to each compared item. This standardized general dominance statistic is less stringent than complete and conditional dominance but is useful for additive decomposition of the fit statistic into components attributable to each IV. The general dominance statistics are derived as the weighted arithmetic average of the marginal contribution to overall fit statistics for each explanatory variable. This approach, akin to Shapley value decomposition, allows for establishing the relative importance of each affordance (Luchman, 2021).

Results

Confirmatory factor analysis

We estimated a measurement model that included 11 latent factors—i.e., six affordances, information sharing within and across departments, identification with the department and the organization, and communication technology use. The CFA demonstrated adequate model fit for Sample 1 (χ² = 1717.22 df = 791, p < .001; TLI = .917; and RMSEA = .051) and Sample 2 (χ² = 1225.38 df = 677, p < .001; TLI = .903; RMSEA = .054). We tested several alternative measurement models (see Table 3). We tested two second-order models treating the affordances as indicators of the same second-order construct, ‘ICT affordances’ (see Manata & Spottswood, 2022). These second-order CFA models indicate slightly worse model fit compared to the retained models for Sample 1 (N = 460; Δχ² = 131.70, Δdf = 27, p < .001; ΔTLI = .005; ΔRMSEA = .001) and Sample 2 (N = 279; Δχ² = 84.58, Δdf = 25, p < .001; ΔTLI = .007; ΔRMSEA = .002). In addition, two single-order models were tested, treating ICT affordances as one unidimensional construct. Again, these models indicated significantly worse model fit for Sample 1 (N = 460; Δχ² = 2223.45, Δdf = 40, p < .001; ΔTLI = .182; ΔRMSEA = .049) and Sample 2 (N = 460; Δχ² = 869.07, Δdf = 30, p < .001; ΔTLI = .138; ΔRMSEA = .030). Hence, our initial CFA model is robust against these alternative specifications.

Table 4 provides the reliability and validity statistics for both samples. First, the reliability coefficients, the maximum reliability (H), and composite reliabilities all exceed .70. For Sample 1, the reliabilities range between .81 and .96; for Sample 2, between .77 and .93. Moreover, convergent validity is examined through the average variance extracted (AVE). The AVE values for mediated technology use ranged between 0.40 and 0.50. While these values fall slightly below the conventional threshold of 0.50 for convergent validity, the high composite reliability (CR ≥ 0.70) provides sufficient justification for assuming convergent validity (Fornell & Larcker, 1981). We also note that communication technology use is sometimes conceptualized as a formative rather than reflective construct (e.g., Park et al., 2011). As such, metrics based on internal consistency may not be ideally suited for evaluating their validity (Bollen & Lennox, 1991). For all other constructs, the AVE values exceed the threshold of .50, ranging between .53 and .77 in Sample 1 and .53 and .66 in Sample 2.

Furthermore, no discriminant validity concerns emerged for Sample 1, as the maximum shared variance among constructs was smaller than the average variance extracted for each construct. Similarly, the square root of the AVE was greater than the inter-construct correlations. However, in Sample 2, the square root for the AVE for self-presentation (.72) was smaller than its correlation with pervasiveness (.88). Similarly, the maximum shared variance was greater (.77) than the AVE (.52). Arguably, the extent one engages in self-presentation overlaps with the pervasiveness of one’s presence online. Since this presents discriminant validity concerns, self-presentation was dropped from the Sample 2 model. The revised model indicated no remaining reliability or validity concerns (see Table 4). In addition, the model is more parsimonious and demonstrated a slightly better model fit compared to the initial model based on a chi-square difference test (Δχ² = 149.26, Δdf = 110, p = .008: see Table 3).

Finally, because both samples represent cross-sectional self-reported data, common method variance was examined through the common latent factor technique. The results indicate only small (Cohen, 1988) changes in the standardized factor loadings (< .200) for either sample, indicating the absence of common method variance. Hence, we proceed to the DA.

Dominance analyses

Table 5 reports the general dominance results for both samples, responding to RQ1.

Sample 1

For Sample 1, the DA estimates 2⁶-1 = 63 sub-models to establish the relative importance of the six affordances in explaining each criterion variable. The results (Table 5) indicate that the affordances jointly predict a very small portion of the respective variance in mediated communication frequency (5.1%), information-sharing quality within the respondent’s department (6.4%), information-sharing quality across departments (4.9%), departmental identification (7.8%), and organizational identification (10%). The DA demonstrates that awareness ranks first for explaining mediated communication frequency (% R² = 25.4%), information-sharing quality within the department (% R² = 50.5%), and departmental identification (% R² = 32.9%), and second for information-sharing quality across departments (% R² = 30.8%), outranked only by pervasiveness (% R² = 31.5%). For organizational identification, awareness ranks third (% R² = 23.9%), outranked by self-presentation (% R² = 30.9%), and editability (% R² = 26.7%). Notably, pervasiveness seems important for mediated communication frequency and information-sharing quality within and across departments (ranking 1^st or 2^nd), but not for departmental (ranking 6^th) or organizational identification (ranking 4^th).

Substantially smaller portions of the explained variances are attributed to visibility and searchability for all five outcome measures, both ranking either 4^th, 5^th, or 6^th, except visibility, ranked 3^rd for information-sharing quality across departments. Overall, these results indicate that the affordance of awareness, and slightly less so pervasiveness, are consistently among the strongest explanations of variance in the outcome criteria.

Sample 2

For Sample 2, for each criterion, the DA estimates 2⁵-1 = 31 sub-models to establish the relative importance of the five affordances (as noted above, self-presentation was not included) in explaining the five outcome variables. As with Sample 1, organizational ICT affordances only explain small respective portions of the variance in mediated communication frequency (8.5%), information-sharing quality within the department (8.5%), information-sharing quality across departments (13.8%), and departmental (6.2%) and organizational identification (13.7%) (see Table 4).

Inspecting the general dominance statistics, the awareness affordance is the largest contributor to mediated communication frequency (% R² = 31.3%) and information-sharing quality within the department (% R² = 39.1%), 2^nd largest for departmental (% R² = 14.4%) and organizational identification (% R² = 29.1%), and the 3^rd largest contributor to information-sharing quality across departments (% R² = 23.8%). In addition, the results again confirm that pervasiveness is one of the biggest contributors, ranking 1^st for information-sharing quality across departments (% R² = 38.9%), as well as for departmental (% R² = 68.6%) and organizational identification (% R² = 48.6%). Furthermore, pervasiveness ranks 2^nd for information-sharing quality within departments (% R² = 26.3%) and for mediated communication frequency (% R² = 21.2%). Finally, visibility is consistently among the least dominant affordances, with the 4^th smallest contribution to explaining mediated communication frequency (% R² = 16.9%), information-sharing quality across departments (% R² = 7.8%), and organizational identification (% R² = 6.3%), and 5^th for information-sharing quality within the department (% R² = 7.6%) and departmental identification (% R² = 5.2%). Searchability is also ranked 4^th or 5^th for all the outcomes except being 3^rd for organizational identification (% R² = 10.7%).

Finally, for both samples, we conducted additional analyses to examine the robustness of our findings. First, we examined the relative importance of affordances across different mediated communication indicators separately (e.g., email, telephone, social media). In 14 separate models across the two samples, awareness, pervasiveness, or both, consistently outperformed visibility in explaining variance in the frequency of technology use. Second, the correlation matrix (Table 4) indicated some significant correlations between gender, work hours, and the criterion of interest. Hence, we conducted additional analyses based on demographic splits (gender and work hours), but those also supported the robustness of our results across these subgroups.

Discussion

The results demonstrate that, overall, organizational ICT affordances are relatively weak explanations of mediated communication frequency, information-sharing quality within and across departments, and departmental and organizational identification. This is to be expected, as no research proposes or finds that such central and enduring organizational phenomena are associated primarily with affordances; many other factors play a role (Weisman et al., 2023). However, the DA clearly demonstrates it is not the visibility afforded by organizational ICTs that drives communication frequency, information-sharing quality, and identification, but rather awareness and pervasiveness do. Our findings compellingly advocate for a consideration of a more comprehensive set of action potentialities afforded by organizational ICTs. Hence, we propose a reassessment of the prevailing orthodoxy that dully privileges visibility, advocating instead for an expanded analytical lens that magnifies a broader set of affordances.

Theoretical implications

Based upon a review of prior affordance literature (Rice et al., 2017, Table 1), our conceptualization of awareness includes being aware of information in one’s own or other departments, cognizant of activities, opinions, or locations of others, keeping up-to-date with projects and with policies and norms. Awareness “is foundational to communication, mediated or otherwise” (Hampton, 2016, p. 112). Our operationalization of pervasiveness includes quick responses, communication across time and place, and across strengths of work relationships. By demonstrating the importance of awareness and pervasiveness relative to other affordances, our findings build on recent studies that highlighted how various affordances (e.g., visibility and persistence) might yield diverse individual and organizational outcomes (van Zoonen et al., 2023). Thus, emphasizing other affordances in addition to visibility could enrich our understanding of these (and presumably similar) outcomes.

To illustrate our point, we consider the theoretical implications of our findings in light of one previous study that focused on visibility. van Zoonen and Sivunen (2023) reported on the relationship between aspects of communication visibility and knowledge brokering in organizations. The study concluded that the afforded communication visibility may facilitate knowledge brokering by providing organizational members equal “vision advantages”—i.e., the ability to see what knowledge exists in the organization—regardless of their position in the social network. The current study results highlight the importance of awareness and pervasiveness in predicting information-sharing quality and identification. This suggests that the context-relevant and continuous interaction reflected in awareness and pervasiveness can lead to even more effective knowledge brokering. Awareness and pervasiveness may ensure that information is not only visible, but also noticed and contextualized (awareness), and integrated across space, time, and networks (pervasiveness). A ‘traditional’ focus on visibility might argue that knowledge brokering is primarily about overcoming barriers to information flow (van Zoonen & Sivunen, 2023). However, awareness and pervasiveness may point to the importance of continuous and embedded knowledge processes, whereas visibility may focus on more episodic knowledge sharing. This is in line with recent research demonstrating that the persistence of information reduced the need to engage in supplemental work, while the visibility of information may generate pressure to immediately act on the information that is made visible (van Zoonen et al., 2023).

Practical implications

Our results have important implications for organizational strategy supporting communication and collaboration within and across departments. First, organizations may benefit by designing and implementing technologies that afford awareness and pervasiveness, and emphasizing these affordances in training and workshops. Instead of only prioritizing tools and platforms that maximize visibility, organizations should also consider how and to what extent technologies enhance continuity and context of information flows. Applying a broader spectrum of affordances may aid in improving knowledge management and organizational socialization.

Second, organizational ICT affordances are more helpful in explaining variance in information-sharing quality across departments and organizational identification, when compared to mediated communication frequency, information-sharing quality within departments, or departmental identification. This suggests that affordances associated with ICTs may be particularly helpful to enhance information flows to more distant, as opposed to proximate, colleagues.

Finally, for individual users of organizational ICTs, the findings provide directions for more effective information management. Employees should engage affordances beyond visibility, by seeking out ways to provide timely and contextually relevant information through actualizing perceived awareness and pervasiveness affordances. However, as research has demonstrated the potentially negative implication of visibility on stress (van Zoonen et al., 2023) and communication overload (van Zoonen et al., 2022), awareness and pervasiveness may also impose demands and distractions on workers. As such, effective technology management and digital skills are crucial in effectively filtering and prioritizing information.

Limitations and future research

The present study has several limitations, though the findings offer a range of future research directions. First, the data underlying this study, though obtained from two samples two years apart, are from one organization, are cross-sectional, and are self-reported (though we found no evidence of common method variance). These characteristics limit opportunities for generalizability and prevent any conclusions regarding the directionality or causality of tested relationships, which we underscore by using terms such as “explanatory” rather than “predictive” or “predictor.” Future research should explore reciprocal affordance causality as it is likely that employees who perceive a wider range of affordances end up using some technologies more frequently (e.g., Rice et al., 2017), while in turn, more frequent uses of specific technologies may expand the scope and level of perceived affordances for users (e.g., van Zoonen & Sivunen, 2023). In addition, the data were collected in 2015 and 2018. While the results demonstrate consistency over these years, technological advances and social context may change, possibly giving rise to new affordances and implications. Hence, ongoing research on the predictive validity of ICT affordances is needed. Collaborative future research might obtain multiple prior datasets measuring organizational affordances and relevant outcomes, and conduct similar dominance analyses to see how general our results are. Further, conducting a systematic or meta-analytical review might contribute important theoretical insights into the conceptual prominence of visibility relative to a host or different organizational ICT affordances (see Table 1).

Second, this study focused on establishing the relative importance of a specific and parsimonious set of organizational ICT affordances based on Rice et al. (2017). However, these are not the only affordances of organizational ICTs. Rice et al. (2017) conducted an extensive literature review, thematic analysis, and construct review, identifying 11 primary affordances (association, awareness, editability, evaluability, persistence, personalization, pervasiveness, searchability, sharing, signaling, and visibility). Measurement validation showed that these indicated the six latent affordances used in this current study. Manata and Spottswood (2022) replicated the identification of the 11 affordances and used those separately instead of the resulting six affordances. Much literature focuses on just a few of the general affordances. For example, Treem and Leonardi (2013), emphasize association, editability, persistence, and visibility. Thus, we feel that the six affordances analyzed here indicate 11 justified and validated underlying possible affordances, represent the major affordances, and cover more than most studies do, but of course, do not represent all affordances mentioned in the literature. Future research may build on previous attempts to further validate the factor structure of organizational ICT affordances. However, while higher-order models may offer a more parsimonious structure, they risk conflating distinctions and nuances between affordances, thereby masking their relative contributions. Given our study’s focus on establishing the relative importance of different affordances through dominance analysis, retaining first-order factors ensures that these nuances remain analytically and theoretically interpretable.

Third, the conceptualization of organizational ICT affordances represents an overall perception of affordances available through the range of existing ICTs to an organizational member. That is, the results do not speak to the relative dominance and explanatory power of affordances of specific ICTs (nor does most of the current ICT affordance research). However, we do note that Manata and Spottswood’s (2022) analysis of the 11 distinct affordances found similar factor structures in one study asking about the extent to which the social media affordances were possible, and a second study asking about how much each affordance influenced the respondent’s stress (with subsamples assessing their preferred platform of Facebook, Instagram, or Snapchat for attaining emotional support). The two Manata and Spottswood contexts are fairly different from each other and from the organizational context in our study. Yet they found the same underlying 11 affordances as in Rice et al. (2017), the basis for our study. Thus, our results about the relative influence of visibility vs. awareness and pervasiveness may be applicable across a range of contexts. Nonetheless, future research could conduct dominance analyses of specific affordances of specific organizational ICTs.

Fourth, we did not examine the interaction or hierarchy among affordances and how this may strengthen or weaken their relationships with possible outcomes. The concept of a possible hierarchy or nesting among affordances was raised early on (Gaver, 1991), does appear in the physical tools/task literature (Wagman et al., 2016), and has begun receiving attention in the ICT affordance area (Ronzhyn et al., 2023, who referred to higher- (general) and lower- (specific) level affordances). Hierarchies or interactions can be important, as the implications of visibility may be more profound when coupled with increased awareness (Engelbrecht et al., 2019; Treem & Leonardi, 2013). Alternatively, searchability may mitigate the impact of pervasiveness because the ability to search for information in ICTs may make the importance of potentially continuous exposure to communication with others (i.e., awareness, pervasiveness) less important for information-sharing quality (for example). Hence, future research may move beyond identifying the importance of individual affordances (whether for specific media or not) and examine how their hierarchy and interdependencies, especially over time, may improve explanatory power and practical implications. In addition, future research is needed to establish a deeper understanding of which affordances matter more, when, under which conditions, and for which outcomes. For instance, visibility may matter more than pervasiveness when the criterion of interest is organizational learning (Leonardi, 2014) or technology-related stress (van Zoonen et al., 2023).

Conclusion

This study shines a beacon on the relative importance of a diverse set of affordances in explaining behavioral (technology use), cognitive (knowledge-sharing quality), and affective (identification) responses. Overall, our findings imply that incorporating a multitude of affordances, and specifically awareness and pervasiveness, could lead to a more comprehensive understanding of how ICTs, information, and networks are integrated and utilized in organizations. We demonstrated that awareness and pervasiveness are relatively stronger explanations for communication frequency, information-sharing quality, and identification than is visibility. We further speculated that a reorientation on affordances beyond visibility may allow researchers to rethink how organizational ICTs may shape organizational communication and collaboration, such as in knowledge brokering—e.g., a shift from episodic knowledge visibility to more contextual and continuous processes.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Funding

The research is supported by a grant from the Academy of Finland (grant number 318416) awarded to Anu Sivunen. The funder was not involved in the development, execution, or reporting of this study in any way.

Conflicts of interest: The authors of this manuscript declare that they have no known conflicts of interest to report. The authors affirm that there are no financial, personal, or professional interests that could be construed to have influenced the work reported in this article.

References