Different ways new information technologies influence conventional organizational practices and employment relationships: The case of cybervetting for personnel selection

The functions and goals of contemporary personnel selection

Personnel selection is often framed as a multi-stage process where employers acquire information needed to screen out undesirable applicants and to identify desirable candidates efficiently (Dipboye, 2014); however, selection also accomplishes other functions. In particular, selection helps organizations achieve relational goals. For example, in addition to screening applicants to create a manageable pool, first interviews also offer occasions for rapport building, applicant information seeking about organizations, and early organizational identification (Miller and Buzzanell, 1996). Even less interactive selection practices (e.g. background checks) can serve relational and information goals since applicant perceptions at each stage affect the development of organizational trust and identification (Anderson et al., 2010). Selection, thus, involves more than information acquisition and use. Selection is also crucial for effective socialization, helping to introduce an organization’s culture and management style (Anderson, 2001) and laying the foundation for the psychological contract between employees and employers (Herriot, 1989). Yet, although relational functions affect individual (e.g. employee retention and satisfaction; Gilliland, 1994) and organizational outcomes (e.g. organizational reputation, productivity, competitive advantage; Ployhart and Weekley, 2010), relational functions are often backgrounded in practice; overshadowed by research and practice focused on optimizing information-seeking tactics and outcomes (Derous and De Witte, 2001).

Researchers who attend to personnel selection’s relational functions often focus on how applicants evaluate and respond to the perceived fairness of selection procedures. When applicants consider selection procedures just (Gilliland, 1994), they assess the organization more favorably and express greater intent to accept offers and establish an employment relationship (Anderson et al., 2010). Applicants deem selection procedures fair if procedures meet expectations for participation and control; offer transparency; provide feedback; respect privacy; and gather job-relevant information (Derous et al., 2003). Yet, as popularly practiced, cybervetting violates procedural justice expectations: applicants consider the practice less fair, less job-relevant, and more invasive than conventional techniques like résumé screening (Berkelaar, 2014). Experts also express concerns. Not only does cybervetting lack structure (the gold standard for accurate and just selection; see Roth et al., 2016), cybervetting’s potential to create disparate impacts suggests ethical and legal implications as well (Clark and Roberts, 2010). Procedural injustice lowers organizational evaluations and intent to accept offers, even as it increases litigation (Stoughton et al., 2015).

Procedural justice perspectives thus suggest cybervetting and résumé screening are not analogous, at least from applicant perspectives. Applicants rarely receive feedback on their résumés. Yet, most people consider résumé information job-relevant, in part because they have greater control over, awareness of, and involvement in deciding what information is included and how on which résumé. In contrast, online information used in cybervetting is made visible by sources beyond applicants themselves. Along with applicants’ (unintentional) self-disclosures, other people and/or technologies also make applicant information visible online. Plus, online information artifacts are selected, re-ordered and re-constructed by employers and the technologies employers use in ways increasingly outside of applicants’ control (Berkelaar et al., 2016). Background checks and cybervetting as practiced are also not analogous as background checks in the US have permission and due process requirements that increase transparency, respect privacy, and offer feedback and recourse for applicants.

Even as cybervetting increases concerns of procedural injustice, technology-enabled practices promise to improve efficiency and provide more information. Employers have long sought ways to optimize selection by increasing the efficiency and amount of information available. In recent decades, employers have privileged information and communication technology (ICT) solutions that automate and replace analog practices (e.g. electronic applications; résumé scanning; Chapman and Webster, 2003). Research on ICT use in selection emphasized the degree to which ICTs alter information accuracy or outcomes (e.g. comparing online versus face-to-face interviews, tests or applications; Chapman and Webster, 2003; Strohmeier, 2009; or social media versus traditional personality assessments; Brown and Vaughn, 2011). Yet, such research rarely considers how ICT use affects personnel selection’s relational functions despite links between effective employment relationships and competitive advantage (Dipboye, 2014); and ICTs effects on organizational processes and employment relationships. Longitudinal research shows that the information ICTs make visible shapes organizational processes, relationships and outcomes in subtle and dramatic ways (Zuboff, 1988). When acknowledged, relational functions are rarely the focus of this research: Chapman and Webster (2003: 118) briefly noted that employers missed personal contact after moving to computerized application screening as employers rationalized that ICTs would allow more ‘face time’ with ‘the best’. Research on whether later interactions increased in length or quality was not found.

Until recently, research on ICT use in selection focused on organizational technologies. Yet, the ICTs enabling cybervetting differ in form and affordances. Organizational or enterprise technologies include ICTs purchased, developed or leased by employing organizations. They differ from the ICTs enabling cybervetting in their goals and the types and range of information made visible (Treem and Leonardi, 2012). Namely, the uses and affordances of ICTs implicated in cybervetting alter opportunities for participation and control; expand what employers count as job-relevant information; and thus, extend the times, contexts and sites in which workers must manage present and future careers and employment (Berkelaar et al., 2016).

Yet, despite differences in applicant perceptions and the underlying ICTs used, research continues to frame cybervetting as an extension or automation of existing selection practices. On the surface, framing cybervetting as an extension of, or a complement to, existing practices such as background checks, pre-screening or résumé checks for ‘red flags’ seems intuitive. Framing cybervetting as a social media background check provides a sense of familiarity that informs and validates behavior (Berkelaar et al., 2015). Yet, such analogies oversimplify the varied ways in which a technology’s materiality (i.e. physicality, form, affordances) entangle with sociality (e.g. norms, values, discourses, goals) to create the possibility of divergent, and sometimes multiple, trajectories. Considering the entanglement of a technology’s materiality and its sociality could help explain mixed results from cybervetting research (e.g. although evaluations of Facebook™ profiles offer accurate predictors of Big-5 personality traits linked to performance (Brown and Vaughn, 2011), cybervetting does not always appear to predict job performance (Van Iddekinge et al., 2013)). The common assumption that cybervetting offers a digital analog to an existing offline practice fails to acknowledge how social factors shape and inform organizational processes (Leonardi, 2012): assuming cybervetting only complements, extends or automates a conventional practice does not consider how employers’ sensemaking about cybervetting’s functions or about the possibilities afforded by particular ICTs becomes entangled with ICT uses to shape organizational processes and outcomes.

People’s talk about ICTs and their associated practices offer clues as to what ICTs will likely do or become in practice, especially during early phases of technological appropriation. By examining competing or coalescing motivations for technologies and their associated practices, researchers can gain insight into the developing spirit of a technology: how people think technology should be used and is being used; that is, its goals, values and functions (DeSanctis and Poole, 1994). Over time, practices like cybervetting and their underlying ICTs intersect with norms, perceptions, attitudes and so forth to create, sustain or disrupt organizational processes (Zuboff, 1988). Thus, it is not that ICTs intrinsically alter organizational processes; rather, it is the new organizational processes enabled by ICTs and people’s sensemaking about those ICTs that alter processes and outcomes (DeSanctis and Poole, 1994). That is, people ‘decide how they will let the technology influence their work’ based on social understandings and material constraints (Leonardi, 2012: 32).

Technology’s materiality – its physicality and form (Leonardi, 2012) – is important to this process. Without the distinct characteristics of social media and Internet technologies, cybervetting, as practiced, would be unlikely. Yet, materiality is not deterministic. Search engines, personal websites and blogs existed prior to cybervetting’s emergence. People’s increased information sharing – enabled, but not determined, by social media and search engine affordances, along with popular discourse that framed information found on social media or via search engines as ‘public’ – helped construct cybervetting as appropriate, even necessary (Berkelaar, 2014). That is, ICTs material properties melded with social norms, practices and discourses to inform organizational practice, affecting people’s perceptions of what a technology can do (i.e. affordances), and consequently people’s goals, and their behavior (Leonardi, 2012).

Because talk often reveals why and for what ends people engage in specific practices (Alvesson and Kärreman, 2000), one way to understand to what extent and in what ways a practice might replicate, shift, displace or otherwise shape existing practices involves studying how employers frame the functions of a practice, here, how they make sense of what cybervetting is for. Thus, to understand whether and how cybervetting might affect personnel selection, I asked:

Existing research has started to address these questions; however, the answers are incomplete because surveys either presuppose a relatively specific fixed set of functions for cybervetting, assume cybervetting occurs at one specific stage, or isolate cybervetting from the overall process. Yet, research hints at a more diverse range of possible or perceived functions.

By examining the language used to describe cybervetting, this study helps illuminate how employers constrain or expand what is considered possible and desirable given new ICTs. Focusing on employers’ talk about cybervetting’s reported functions and goals answers numerous calls to examine the communicative processes underlying selection (Dipboye, 2014), and the affordances and uses of non-organizational ICTs on selection specifically (Berkelaar et al., 2015; Roth et al., 2016), and organizational processes generally (DeSanctis and Poole, 1994; Leonardi, 2012). This study also offers an exemplar of an emerging sociotechnical practice that can help illustrate how the functions ascribed to a specific technology-enabled practice affect conventional practices, as well as its anticipated effects on larger social processes and outcomes.

Method

Findings

Participants described cybervetting practices in ways consistent with extractive information seeking. They also ascribed three primary functions to cybervetting: screening, efficiency and relational. Initially, these findings seem to reinforce extant efficiency and information screening goals of personnel selection. Yet, considered together, these findings also complicate taken-for-granted assumptions that cybervetting is analogous to, or an extension of, conventional background checks or résumé screening. Rather, these data suggest some employers frame extractive practices as interactive and relational – a framing consequential for selection processes and outcomes including employment relationships.

As context for participant responses, I offer some details on the reported prevalence of cybervetting. Without prompting, more than three-quarters of these participants directly acknowledged looking online for applicant information, although the reported timing and tools varied somewhat. Some participants also reported cybervetting current employees. Participants who did not directly acknowledge cybervetting often referenced ‘legal’ or ‘compliance’ departments, describing how ‘from a legal perspective, I’m basing everything on the résumé,’ and ‘only using employment-related’ information, on occasion noting how cybervetting violated ‘personal ethics.’ Yet, even participants who explicitly denied cybervetting when directly asked, still reported that they and/or other organizational members known to them engaged in behaviors consistent with cybervetting – for example, using Google to ‘clarify résumé details’ or noting ‘they might come across information’ online if they were ‘somehow connected’ to an applicant. Such language situates cybervetting within known, expected, common and legal selection procedures and provides a context for the three functions ascribed to cybervetting.

Discussion

In describing employers’ sensemaking about cybervetting, this study provides insight into the likely, if unintended, results of technology-enabled practices on personnel selection, employment relationships and broader organizational processes (see Berkelaar, 2014; Dipboye, 2014; Leonardi, 2012). Understanding how cybervetting ‘works’ in contemporary selection matters given the individual, organizational and societal consequences of personnel selection. Cybervetting also offers a useful exemplar to help illustrate how talk can provide clues into whether and how emerging sociotechnical practices may affect conventional practices. These findings suggest as last two key theoretical contributions.

First, this study highlights the crucial role of sensemaking in shaping the trajectory of emerging sociotechnical practices within broader social processes. Most extant research presumes cybervetting replaces a conventional information-seeking practice that most people consider relatively extractive, albeit more interactive than cybervetting (i.e. résumé screening; interviews). Yet, these data suggest that employers may also consider cybervetting a replacement for the relational functions of more interactive practices. This is not to say that because some employers frame cybervetting as relational or interactive, that cybervetting is necessarily relational or interactive. Rather, when people ascribe different functions to cybervetting, they change what behavior is considered desirable, possible and likely, especially when frames are subtle (Shah et al., 2010). Such frames are real in their consequences. Specifically, when participants framed cybervetting as relational or interactive, employers considered and sometimes reported acting on the possibility that cybervetting could replace the first interview or social aspects of interviews, even as they reported cybervetting behaviors consistent with extractive information seeking. In contrast, employers who did not frame cybervetting as relational or interactive, but framed it as serving efficient and/or screening functions and fulfilling informational or efficiency values, reported using cybervetting to supplement conventional practices. The few employers who felt cybervetting was not efficient, accurate and/or fair tended to reject cybervetting themselves ‘at least for now’ (although they often expected or wanted others to cybervet). Thus, the functions people ascribe to an emerging practice can constrain or enable possibilities for whether and how an emerging practice becomes situated within a broader organizational process.

These data also suggest that salient social values shape whether and how cybervetting fit into the overall selection process. Efficiency and the perceived availability of more information offered dominant values by which employers evaluated the relative effectiveness of cybervetting at serving desired functions. In reporting whether and how they incorporated cybervetting into their overall selection process, participants also drew on perceived affordances of ICTs (e.g. information visibility; ‘social nature’). Employers combined these affordances into their sensemaking about the functions ascribed to cybervetting (screening, efficiency, relational) and salient individual or collective values (e.g. ‘efficiency,’ ‘sufficiency,’ ‘humaneness,’ ethicality, ‘legality’) and then acted based on entangled (often implicit) assessments of the social and technical aspects of emergent practices. Such behavior aligns with a long tradition in the sociology of technology whereby technological development involves contesting, negotiating and (re)defining what it means to say that a particular technology ‘works’ given particular material and social conditions (see Leonardi, 2012 for summary).

Beyond providing insight into personnel selection, this study also offers a useful exemplar for understanding how emerging sociotechnical practices evolve generally. Specifically, I propose that whether people reported supplementing or replacing a conventional practice with an emerging practice, or rejecting the emerging practice, seems to depend on at least two key factors: (1) the salient values linked to the practices being considered; and (2) the degree to which employers felt the functions ascribed to the emerging practice aligned with those of the conventional practice (see Figure 1). Although functions and values are entangled in these data, this framework proposes a means of disentangling and testing values and functions as potential causal factors.

OPEN IN VIEWER

Figure 1.

How will an emerging practice affect a conventional practice?

That is, this framework proposes that when people frame an emerging practice in a way that highly aligns with salient values and the ascribed functions of a conventional practice, they are likely to displace the conventional practice with the emerging practice, especially when the conventional practice has a lower alignment with salient values. Moreover, I suspect such processes and outcomes are not exclusive to personnel selection. Consider the emergence of mobile-app-enabled ridesharing organizations (e.g. Uber, Lyft). Imagine a woman who compares ridesharing (emerging practice) to taxis (conventional practice), frames the function of transportation as providing on-demand transport, with a salient value of affordability. She would likely replace taxi service with ridesharing. In contrast, when an emerging practice highly aligns with the ascribed function of a conventional practice, yet does not align with the salient value as highly, the emerging practice is more likely to supplement, rather than replace the conventional practice. In the ridesharing example, if the passenger’s salient value was safety, she might supplement taxi service with ridesharing only if ridesharing drivers were verified. Conversely, if an emerging practice is comparatively lower than the selected conventional practice in terms of salient value(s) and ascribed function(s), people will tend to reject the emerging practice. Imagine a man comparing ridesharing (emerging) to his current form of transportation (conventional, e.g. personal car) and who views the function of transportation as moving between locations without waiting. His salient value is control. This man would likely to reject ridesharing. One could also imagine situations where low value alignment might constitute a personal or professional ethical violation, in which case people might reject the emerging practice, change their value, or reframe the practice to align with the value.

What remains unclear is the anticipated trajectory when an emerging practice highly aligns with salient values, but is relatively low on ascribed functions. Although the lack of data for these conditions might be an artifact of data collection; should a practice emerge under this fourth set of conditions, I tentatively suggest it might supplement an existing practice until it was considered to have a comparative functional advantage within a ‘reasonable’ amount of time, or otherwise it would be rejected. It also remains unclear how and when people decide on the conventional practice with which to compare the emerging practice; whether ascribed functions reflect people’s beliefs or operate exclusively persuasive strategies to accomplish organizational goals; or how contextual or temporal changes shift behavioral choices. This framework also does not consider whether a practice actually offers a functional advantage. More research is needed.

This proposed framework aligns with notions that institutional and organizational norms, values and discourse shape how people use technology and how a technology (or set of technologies) unfolds as part of a sociotechnical process (DeSanctis and Poole, 1994; Leonardi, 2012; Orlikowski and Scott, 2008). By focusing on how people compare the ascribed functions and alignment of salient values of conventional and emerging practices, this study extends conversations on sociomaterial practices. In so doing, it provides a tractable means for testing the evolution of other sociotechnical practices central to contemporary life. Specifically, this study reinforces the need to ‘challenge the separation of technology, work and organization,’ but rather to consider how technology (e.g. mobile apps), work (e.g. gig economy) and organizations (e.g. Uber, Lyft, city councils) overlap and entangle as equivoque not entailments. Whereas equivoque allow for ambiguity and multiple possible interpretations, entailments suggest logical and necessary consequences (Orlikowski and Scott, 2008: 433). As a technology-enabled practice, cybervetting is equivoque – cybervetting did not (and does not) have to evolve in one particular way. Rather, alternate interpretations about the functions of cybervetting and its underlying technologies, along with differences in salient values afford opportunities for different sociotechnical evolutions that might better address the instrumental and ethical challenges of personnel selection (see e.g. Clark and Roberts, 2010; Roth et al., 2016) and ongoing concerns about the problematic trajectory of contemporary employment relationships (e.g. Berkelaar, 2014). This framework also provides a way of understanding the space of action for other contemporary and controversy technology-work-organization entanglements (e.g. the gig economy; electronic health records) potentially offering a means of predicting current or likely trajectories and offering options for resisting deterministic trajectories. In sum, this study encourages people to question the functions and values ascribed to practices considering empirical data on actual functions and values enabled by particular practices.

Second, these findings also highlight the need to understand the different modes of relating in organizations, including how people make sense of employment relationships. Within this study, a subset of participants considered cybervetting a tool for relational development or interaction even as they reported using extractive means for cybervetting. Although there were at least two people involved in the movement of information – insofar as one was the seeker and one was the target – cybervetting, as reportedly practiced, rarely offered a reciprocal exchange of information, despite being framed by some participants as relational. Conventional modes of relating assume reciprocal, usually overt, information exchange (whether synchronous, asynchronous, asymmetrical or symmetrical). However, these findings suggest employers may relate to applicants, or rather applicant information, via covert information extractions. By leveraging the information and clandestine affordances and features of search and social ICTs, extractive processes separate the interactive functions of relating to another person from the informational functions of relating to another person. ‘Relating’ with a target from this perspective does not necessarily involve the target relating with the seeker via back and forth interactions, but rather ‘relating’ unidirectionally via information about the target.

Unidirectional targeted relating of one individual to another via information cues aligns with media research on parasocial interactions (Giles, 2002). Parasocial interaction or parasocial relationships describe ‘relationships’ in which people respond to a media figure as if in a typical social relationship. Historically, parasocial relationships target celebrities or fictional characters. Because of the growing amount of information visible online, people can now relate parasocially to non-celebrities. Although parasocial interactions exhibit many of the same qualities of social relationships for the person directing the relationship, one person remains a ‘stranger.’ Such estrangement could prove problematic in employment relationships. Even as extractive communication seems to offer a functional alternative to some key challenges of contemporary selection; it does not come without costs. Although cybervetting may offer more efficient information seeking, the distribution of relational control in extractive information seeking, and by extension parasocial employment relationships, undermine applicant needs for voice, awareness and space for action. Such imbalances have implications for procedural justice (Gilliland, 1994) and the establishment of a healthy psychological contracts and employment relationships (Derous and de Witte, 2001; Herriot, 1989).

These findings challenge scholars to reexamine the fundamental notions of what constitutes an employment relationship – an imminently practical consideration given ongoing debates about what constitutes a worker in a gig economy. This study highlights and extends conversations about the ways in which new ICTs make it possible for parasocial relationships to be established more widely, by more people, and in more relational contexts. Perhaps employment relationships also exist on a continuum from parasocial to social determined by the varying combinations and orders of social versus information encounters. Understanding how people make sense of and practice one-way (online) information-based ‘relationships’ can expand research on parasocial relationships, while providing insight into the development and structure of contemporary employment relationships.

These findings also offer corresponding practical insights. First, the findings reinforce the need to pay attention to sensemaking about the sociomaterial aspects of organizational processes. By framing cybervetting as a chance to build a sense of rapport and connection, employers could justify focusing on more immediate information and instrumental goals like assessing information during the time-intensive practice of interviewing, or excluding one or more interviews altogether. Such choices could (unintentionally) decrease the proportion of relatively more interactive information seeking, potentially delaying or otherwise altering the development of the employer-employee psychological contract (see Derous and de Witte, 2001; Herriot, 1989). Practitioners would benefit from considering communicative approaches when evaluating emerging practices and their desired and likely effects on organizational processes such as personnel selection. Such an approach would include at least two foci. Practitioners should consider their goals for personnel selection overall, moving beyond job screening and fit to consider how competent interactions can improve selection, retention and engagement through effective socialization and identification. Practitioners should also engage in research and reflection on which strategies best accomplish relational and informational goals. Such an approach could allow employers to be more strategic with limited time and fiscal resources, while providing opportunities to encourage just, fair and effective selection processes. As suggested in recent European Union considerations of the ‘right to be forgotten,’ policymakers also should consider how to legislate information in ways that offer people greater voice, interaction and/or control over digital artifacts given visibility, persistence and replicability affordances of online information; while also recognizing employers’ needs for accurate, job-relevant information. Plus, given the growing using of big data and algorithms, policymakers and practitioners should also consider how computational cybervetting might affect practices, processes and outcomes related to voice, fairness and personal information control.

By attending to talk of what a tool or practice does or should do, practitioners can better understand the assumptions and implications of sensemaking surrounding a specific practice (DeSanctis and Poole, 1994; Orlikowski and Gash, 1994). When pooled with attentiveness to technology appropriations and organizational processes, such an approach helps practitioners anticipate and strategically intervene to encourage desirable trajectories and outcomes and redirect undesirable ones. Relatedly, employers should attend to the language used when describing practices as well as their imagined relationships with the people online information references: What do I (others) think I am (we are) up to when engaging in a particular practice or overall process (Alvesson, 2003)? Who is that I (they) think I am (they are) relating to? Attending to language helps reveal assumptions about what constitutes possible and desirable actions and outcomes given material opportunities and constraints (Phillips and Oswick, 2012).

These findings provide avenues for future work. Since this study employs a purposive sample, it should not be used to make generalizable claims. Rather, the preliminary framework proposed here should be tested to help disentangle how values and ascribed functions influence the evolution of (sociotechnical) organizational practices and processes. Specifically, future research should draw on experimental, field and survey approaches to help examine under what conditions and to what extent employers’ talk affects conventional practices, the overall selection process, and short- and long-term outcomes of cybervetting specifically, and technology-enabled practices generally. Research into the frequency, timing and intensity of cybervetting could offer missing insights or help untangle mixed results of existing research. Broadly speaking, scholars should consider to what extent the proposed framework holds across different practices, processes, technologies, contexts and over time; and whether and how different degrees of relationality are enacted across time and contexts. That is, how and in what ways do sociotechnical practices and employment relationships evolve?

In conclusion, this study addresses Orlikowski and Scott’s (2008) call to consider the role and influence of technology on organizational practices and processes. It encourages scholars and practitioners to consider how ascribed functions and salient values influence the trajectories and outcomes of sociotechnical practices, rather than assuming technology-enabled practices are altogether different or altogether the same as conventional analogs. Specifically, findings encourage people to consider how talk affects the likely, potentially unanticipated, consequences of cybervetting on personnel selection practice and employment relationships. Findings also offer a preliminary, testable framework to help explain how sociotechnical practices emerge, with what likely effects; offering avenues for future research, practice and intervention.