Emphasizing worker identification with skills to increase helping and productivity in production: A field experiment

Franke, Henrik; Kwasnitschka, Daniel; Schmutz, Jan B.; Netland, Torbjørn H. Article — Published Version Emphasizing worker identification with skills to increase helping and productivity in production: A field experiment Journal of Operations Management Provided in Cooperation with: John Wiley & Sons Suggested Citation: Franke, Henrik; Kwasnitschka, Daniel; Schmutz, Jan B.; Netland, Torbjørn H. (2024) : Emphasizing worker identification with skills to increase helping and productivity in production: A field experiment, Journal of Operations Management, ISSN 1873-1317, Wiley Periodicals, Inc., Boston, MA, Vol. 70, Iss. 5, pp. 712-732, https://doi.org/10.1002/joom.1300 This Version is available at: https://hdl.handle.net/10419/306096 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by-nc/4.0/ RESEARCH ARTICLE Emphasizing worker identification with skills to increase helping and productivity in production: A field experiment Henrik Franke 1,2 | Daniel Kwasnitschka 2 | Jan B. Schmutz 3 | Torbjørn H. Netland 2 1 Faculty of Management, Economics and Social Sciences, University of Cologne, Cologne, Germany 2 Department of Management, Technology, and Economics, ETH Zürich, Zurich, Switzerland 3 Department of Psychology, University of Zurich, Zurich, Switzerland Correspondence Henrik Franke, Faculty of Management, Economics and Social Sciences, University of Cologne, Sibille-Hartmann-Str. 2-8, 50969 Cologne, Germany. Email: [email protected] Funding information ETH Foundation SEED Grant, Grant/Award Number: SEED-06 21-1 Handling Editors: Aravind Chandrasekaran, Rogelio Oliva, and Bradley Staats Abstract Can productivity improve if workers identify more with the skills they use in their work environment? This paper reports the results of an experimental design that was peer-reviewed prior to collecting data. The research setting is a global manufacturer using a novel smartwatch-based system for distributing work tasks among factory floor workers. Drawing on the concepts of identification and helping in organizations, we hypothesized that fostering workers' identification with their own skills could serve as a mechanism to enhance helping behavior on the factory floor, which should improve productivity. We designed a compound skill-fostering treatment consisting of communication, meetings, and exercises regarding individual skills. We treat one large factory area for 2 weeks and keep a similar area in a sister factory as a control group for comparison in a difference-in-difference model. The results show that nudging skill identification increases workers' identification with skills, but we do not find evidence for increased helping behavior or increased productivity. Our results help develop theory around multiple sub-identities and provide guidance for future studies seeking to enhance identification in organizations. KEYWORDS digitalization, helping behavior, identification at work Highlights •A 4-week-long field intervention was conducted in a digitalized factory aiming to increase workers' identification with their skills. •The experimental design was reviewed and accepted by the JOM Editors before data collection. •Identification with skills increased and emotional exhaustion decreased. •Expected positive effects of the treatment on helping behavior and productivity were not supported. •Managers and future studies should try to disentangle identification with skills and machines. Received: 30 April 2022 Revised: 15 February 2024 Accepted: 24 February 2024 DOI: 10.1002/joom.1300 This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. © 2024 The Authors. Journal of Operations Management published by Wiley Periodicals LLC on behalf of Association for Supply Chain Management, Inc. 712 J Oper Manag. 2024;70:712–732. wileyonlinelibrary.com/journal/joom 1|INTRODUCTION Workers are more motivated to exert effort when they identify with their immediate work environment (Albert et al., 2000;Duttonetal.,1994). In fabrication operations, workers' identification with a set of machines often comes naturally since managers allocate workers to machines to solve two problems: detecting interruptions as soon as possible in hard-to-oversee facilities and matching technical problems with the appropriate skill. This way, a worker can closely monitor the assigned machines' status and develop appropriate skills over time. In this context, skill can be seen as the ability of a worker to operate or solve an interruption of a specific machine type. Due to the workermachine allocation, workers naturally attach meaning to the allocated machines and incorporate them into their self-concept: they identify with the machines (Miscenko & Day, 2016). Even when allocating workers to specific machines, factory floors are usually organized in teams that rely on citizenship behavior among workers, such as helping behavior (Cantor & Jin, 2019). We define helping as the voluntary exchange of work among workers that is not directly or explicitly recognized by the formal reward system (Organ, 1988). Thus, the tight allocation of the workforce to machines can be useful for monitoring large-scale production and driving identification with a few machines. However, it may crowd out the potential to identify with other features of the factory floor that could increase productivity via helping behavior. This problem becomes particularly salient after a digital transformation of the factory because digitalization can eliminate the problems that the strict allocation of workers to specific machines intended to solve; it can make all machine status digitally available and dynamically match factory floor tasks with workers' codified skills. The inherent challenge of motivating employees to show helping behavior is relevant in various contexts, including manufacturing. Interestingly, a digital transformation can have positive and negative effects on job resources (Parker & Grote, 2022). On the one hand, new digital technologies can potentially reduce the need for human interaction and coordination. For example, algorithms can automatically allocate human workers to tasks (Bai et al., 2022). On the other hand, new digital technologies can increase the transparency on the shop floor, highlighting where problems are and automatically ping someone to help. This information is necessary but not sufficient for helping behavior and introduces the challenge to encourage helping among workers. For the past 4 years, we have studied a company that has experienced this double-edged problem first-hand: a globally operating manufacturer headquartered in Italy. The company has implemented a digital worker-task matching system using smartwatches, machine sensors, and a digital back end (see Figure 1). The company in our research setting produces millions of complex metal parts daily, and its workers receive information about, for instance, machine interruptions from machines directly and in real-time via smartwatches in case they are available and have the appropriate codified skill in their profile. Codified skills are the binary codifications of human skills in a digital system. The new technology has changed the past organization based on machine groups into a new organization based on skills. The main goal of the system is to pool skills across machines to achieve higher productivity. The company aims to be at the forefront of digital transformation in its industry, has undergone far-reaching changes in its production organization, and faces unique behavioral challenges grounded in their progressive digital transformation. As known, profound technological changes can positively and negatively affect humans at work (Parker & Grote, 2022). Therefore, our study chooses a behavioral angle and considers workers' wellbeing important to an organization's social sustainability. In theory, workers equipped with the smartwatch would no longer restrict their work to a specific area or set of machines but focus on any machine that requires their skill, as all solvable tasks appear on the smartwatch as interruptions occur (see Figure 1). In practice, however, workers have established their work identification based on machines for years, and the change of identification lags the technology change. The new technical production logic focuses on matching activities and codified skills, whereas the socio-psychological concept of identification still focuses on machines. Workers prefer to continue working on a limited set of machines since workers strongly identify with these machines. Employees are drawn to “their” machines because they are familiar with and part of their accustomed surroundings. Sociologists have described this processasprocessualinteractionsthathelphumansunderstand and construct their reality and their concept of self within that reality (Gecas, 1982). While this division of work by machines indeed also has benefits in reducing walking distances and skill-specificity, the narrow identification with and feeling of responsibility for machines may make employees reluctant to help in other areas, which is a success factor in production (Cantor & Jin, 2019). Forcing workerstoworkonmachinesoutsidethescopeoftheir work identification bears the risk of upsetting workers, resulting in low satisfaction and eventually negatively impacting productivity. Instead, voluntary willingness to help is preferred. Notably, in our research setting, the lack of helping behavior is not a temporary adoption difficulty of the FRANKE ET AL.713 system but has persisted since the system was introduced in 2020. Based on the theoretical and practical problem outlined above, we derive the following research question: Can we improve helping behavior and productivity by increasing workers' identification with skills? The potential contribution of addressing this question to operations management theory lies in proposing a shift in worker identification as a concealed requirement for realizing the advantages of digitalization in production settings. Specifically, this study can contribute to the behavioral operations management discussion on worker identification and helping behavior. This behavioral operations literature has been silent on identification in manufacturing and instead centered on similar problems in software use and delivery contexts (Bagozzi & Dholakia, 2006; Ta et al., 2018). At the same time, the community has initiated a discussion on how helping behavior can improve operations (Cantor & Jin, 2019). Our work can broaden the discussion of identification from software and delivery to manufacturing and continue the helping behavior research by examining whether a treatment can incrementally affect workers' identification patterns to favor helping behavior and productivity. Our research builds on the concept of identification in organizations (Albert et al., 2000; Ashforth et al., 2008) and draws motivation from recent management theory around the coexistence of multiple sources of identification (Bataille & Vough, 2022)—such as machines and skills. Using this lens, we aim to draw novel operations-specific implications for helping behavior and productivity on the factory floor of a highly digitalized manufacturer. Thereby, we offer identification as a potential complement to the behavioral operations literature on how task design, interdependence, incentive systems, or motivation can contribute to worker collaboration (De Vries et al., 2016; Franke et al., 2022; Schoenherr et al., 2017; Siemsen et al., 2007). 2|THEORETICAL BACKGROUND 2.1 |Work identification and identity The literature on identification and identity has a long tradition but is also heterogeneous and lacks universal definitions (Albert et al., 2000; Miscenko & Day, 2016). Our study follows the idea of personal identification at the individual level of inclusiveness (Brewer, 1991; Brewer & Gardner, 1996), which argues that any collection of meanings can define a worker's self-concept at work (Gecas, 1982). Examples can be roles at work, membership in social groups, or—as in our study—one's professional skills or the machinery one uses. Gecas (1982) summarizes two alternative views of how identification can emerge: via an individual's interactions with the environment or via the roles of an individual. Both are valid sources of identification, yet this study focuses more on how humans connect to features of their work through interactions. In line with this approach, Miscenko and Day (2016) have proposed that “identity refers to the meaning of a particular entity (i.e., role, organization) that is internalized as part of the self-concept”and that “identification is a cognitive/psychological/emotional attachment that an individual FIGURE 1 Impressions from the factory floor (left), the smartwatch (middle), and the digital front end (right). 714 FRANKE ET AL. makes to a role, team, organization, or other entity” (p. 217). In other words, the former is a state, and the latter can be interpreted as a behavioral process. These definitions do not conflict with the conceptualizations of identity and identification as socially constructed (i.e., Turner & Tajfel, 1986) but are complements. We focus on personal identity since this concept commonly distinguishes individuals, whereas social identity focuses on differences between groups that can define in-groups and out-groups with their identity (Ashforth et al., 2008). We use the term identification in our study instead of identity since workers literally seem to be attached to their machines and are reluctant to go elsewhere on the factory floor to help. Identification is a useful term in our study as the notion of identification as a process matches the reality on the factory floor: workers become attached through their work. Furthermore, this view of identification fits the idea that attachments are somewhat fluid and can be changed via management practices. However, we acknowledge that both identification and identity concepts are inextricably connected on the conceptual level (Ashforth et al., 2008), point to studies that discuss their relation (Dukerich et al., 2002; Dutton et al., 1994), and note that scholars often treated them as synonyms (Miscenko & Day, 2016). In summary, while we review the consolidated literature on identification and identity and our arguments would allow using both terms, we opt for “identification”for its stronger resonance with our research context. We review the consolidated literature on both concepts. Identification matters in operations contexts. Scholars have examined the common identification of production workers with their production area and compared it to sports fans who identify with their team (Urda & Loch, 2013). The study showed that when a worker is unexpectedly rewarded, it may trigger guilt among other workers in that area as they start to ask why they were not good enough. However, no guilt was measured among supporters of the same sports team in a comparable situation. Thus, mechanisms and outcomes around identification are not identical in production and other settings, which motivates behavioral operations examinations. Several studies define identity and the target of identification as the social group in the operations management literature. They focus on hiring and in-group membership (Casoria et al., 2022; Del Carpio & Guadalupe, 2022), trust in transport services (Ta et al., 2018), groups making donations (Charness & Holder, 2019), or identification with buyers in supply chains (Corsten et al., 2011). Only a few operations management studies chose our focus on individual-level personal identification and on how features other than social group membership affect identification and outcomes (e.g., Reagans, 2005). To the best of our knowledge, the behavioral operations management literature has not examined the effects of identification on the factory floor, and all prominent review articles omit the concept (Bendoly et al., 2006; Bendoly, Croson, et al., 2010; Croson et al., 2013; Donohue et al., 2020; Fahimnia et al., 2019). A recent review of the literature concludes: “it is clear that decision making in practice continues to be heavily influenced by human judgment, even with regard to highly automated and supposedly objective systems.”(Fahimnia et al., 2019, p. 29). Identification is one element that may explain human judgment in decisions about whether to help or not. 2.2 |Helping behavior To improve productivity, manufacturing relies on workers to help each other to improve overall performance (Cantor & Jin, 2019). The operations management literature has established that correctly designing incentives can encourage collaboration (e.g., De Vries et al., 2016;Siemsen et al., 2007), yet a trade-off exists between the use of explicit incentives and possible concerns of crowding out intrinsic contributionsacrossmanycontexts(Decietal.,1999). This is especially true when incentives are closely tied to performanceandwhenqualityisessential,asiscommonlythe case in factory floor operations (Cerasoli et al., 2014). Therefore, the literature has begun to focus on fostering voluntary worker behavior in addition to the research on incentives. In broad terms, helping behavior is part of “organizational citizenship behavior,”which is defined as “individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system, and that in the aggregate promotes the effective functioning of the organization”and is a synonym for altruism (Organ, 1988, p. 4). It is useful to specifically observe helping as one factor of organizational citizenship behavior in the context of operations since collaboration is immediately relevant for performance. Despite the scarce coverage of helping behavior in the operations literature, it is a relevant concept in practice, as examples from the company in our research setting can illustrate: When I know that a machine [of a co-worker] is not running “clean,”say the [robot] arm keeps getting the alignment of the part wrong while inserting parts into the press, I will look out for tasks on his machine. I can help out when he is on break or away for some reason. However, such helpful behavior is not given in a production context. Consistently, the team leader described FRANKE ET AL.715 a situation where helping could have avoided performance losses: I came in this morning, and my dashboard showed me that this machine had been interrupted for 24 minutes already. That's ridiculous! One guy is in the lab, and the other is chatting. They go “they have people over there. Not my job”although it really is. This grinds my gears. The operations management literature has hitherto not addressed the link between personal identification and helping behavior or other types of cooperation. Previous contributions have addressed related concepts, such as the effects of within-group interactions on performance in project contexts (Bendoly, Thomas, & Capra, 2010)orfunctional dominance, which can reduce cooperation and affect the performance of cross-functional teams (Franke et al., 2022;Malhotraetal.,2017). Moreover, conflicts can reduce team cooperation and the performance of crossfunctional teams (Franke et al., 2021; Oliva & Watson, 2011). Cantor and Jin (2019)werethefirstto examine questions about voluntary help in production explicitly. Their study finds that workers who are more aware of others' efforts will more likely detect performance differences and attribute them to a lack of motivation, which reduces helping behavior. The paper suggests that creating interdependence between the workers' performance can encourage helping. Interdependence is the extent to which employees depend on other group members to carry out work effectively (Bachrach et al., 2006; Van Der Vegt et al., 2003). Interdependence is a critical factor regarding helping behavior, both in the operations and general management literature. Studies agree that interdependence can drive team collaboration by making it a necessity (Cantor & Jin, 2019; Schoenherr et al., 2017). Interdependence makes one's own success dependent on others' work. Thereby, it benefits those who help others via a self-serving element that transcends the altruistic help concept. The management literature commonly examines helping behavior and related concepts in such interdependent teams. For instance, scholars have accumulated evidence supporting that collaboration or felt obligations to help, as well as concepts that derive from them (cohesion, information exchange, etc.), drive performance in work teams (Kilcullen et al., 2022; Lorinkova & Perry, 2019; Mathieu et al., 2008;Mathieu et al., 2019; Mesmer-Magnus & DeChurch, 2009). However, not all production tasks are interdependent or can be changed to become interdependent. Unlike in assembly flow lines or cellular manufacturing, largescale automatized and digitalized mass unit-production settings require workers to monitor production and autonomously intervene when interruptions occur rather than actively collaborating as a team. Consistently, research has shown that interdependence is an important boundary condition for conclusions around helping behavior (Bachrach et al., 2006). This study examines an under-researched, non-interdependent operational setting to help explain the scarcely understood relation between helping and productivity from an identification standpoint. 2.3 |Helping behavior and identification The most natural conclusion from intersecting research on helping and identification is that to encourage helping within a group, it is useful to emphasize employees' identification with that group, be it their immediate work group or the entire organization (Dukerich et al., 2002; Dutton et al., 1994; Janssen & Huang, 2008; Van Der Vegt et al., 2003; Wu et al., 2016). However, although production workers may be organized in teams, the workgroup may not be a salient enough feature of their work, making it difficult for workers to identify with it. Factory floor management can face a dilemma between the lack of salient teamwork and their strong reliance on voluntary help since an enforcing mechanism like interdependence is often absent. This is commonly the case in production environments such as the one we address in this study: highly automatized and digitalized mass unit-production settings in which workers monitor and intervene but seldom actively collaborate as a team. Thus, the above-cited findings from the management literature are sound but not necessarily transferable. This illustrates a research gap at the intersection of helping behavior and identification regarding the unforeseen challenges that cutting-edge digital technology imposes in manufacturing. Instead of identifying with the team, workers tend to develop strong identification with machines via their traditional assignment to the equipment, as motivated at the start of the paper. We acknowledge that, as any human, production workers would likely respond positively to interventions or training that build team cohesion (Chiniara & Bentein, 2018;Hu& Liden, 2015). However, the effects will unlikely persist as daily routines on the factory floor are still determined by the production technology that does not reflect active teamwork. Therefore, we propose an alternative avenue to increase helping behavior and productivity on the factory floor via identification with workers' skills. 716 FRANKE ET AL. 3|HYPOTHESIS DEVELOPMENT 3.1 |Effect of identification with skills on productivity Employees show higher levels of motivation at work when they identify with the features of their job (Albert et al., 2000; Dutton et al., 1994). Problems in highly automatized and digitalized manufacturing are often complex and involve sophisticated machinery. Thus, these problems require motivation that makes workers focus and pay careful attention when machines are interrupted. Workers who identify more strongly with their skills focus more on the immediate task since using their skills reinforces their self-concept. Focusing on work is not a tedious exercise for them but can be a source of job satisfaction when their skills and work align, as should be the case when a digital task-allocation system matches tasks with skills (Vignoles et al., 2006). Specifically, when workers are actively solving a machine interruption, higher identification with their skills likely enables them to recall aspects of their expertise, transfer knowledge from one problem to the next, or apply the skills they have more effectively. In other words, skills that are sources of identification for a worker are likely of higher quality and more thoroughly applied, which can positively affect several facets of productivity. It can reduce the downtime of machines and processing time of interruptions and increase the availability of machines to produce parts. It can also reduce future interruptions of the interrupted machine by contributing to more sustainable problem-solving on the factory floor, further improving productivity. Finally, higher identification can also reduce the latency of pending work tasks on the factory floor when workers respond faster to tasks that provide self-reinforcing value to them. This reduces the duration of interruptions waiting unaddressed and increases operational productivity in production. Thus, we hypothesize: Hypothesis 1. Higher levels of identification with skills will be associated with higher levels of productivity in digitalized production environments. 3.2 |Effect of identification with skills on helping behavior We expect stronger identification with skills on the factory floor to increase productivity (Hypothesis 1). We propose that one important and hitherto overlooked mechanism on how identification with skills contributes to productivity relies on mutual helping on the factory floor. Helping behavior requires awareness of other workers, machines, tasks, or general entities around one's traditional scope of responsibilities in production (Cantor & Jin, 2019). Simply put, without being aware, one cannot make the decision to help. Identification is more than awareness; it means that individuals form a psychological relationship with an entity. For a digitalized production setting using smartwatches for task allocation, workers are made aware of opportunities to exchange work with co-workers via the watch. We argue that their choice to help depends partly on their relation to their own skills. This relation is characterized in the literature as an emotional investment that individuals make in an evaluation process (Ashforth et al., 2008;Tajfel&Turner,1982). This leads to what Miscenko and Day (2016) call attachment to an entity of the work environment, such as one's own skills. When workers in a production facility do not identify strongly with their skills, they naturally search for other sources to define their self-concept. These can be any entities but are unlikely the team, as teamwork is less salient in highly automated and digitalized productions. The entities that workers identify with instead may or may not encourage helping due to their inherent nature. Identifying with a part of the product spectrum, for instance, would likely focus the scope of workers' awareness—a precondition to identification—on those products and, therefore, reduce helping in times when other products are scheduled or in areas where these other products are assembled simultaneously. When workers identify with the machines in their scope of responsibilities, they focus their attention on those machines and likely feel reluctant to help when machines that they do not identify with as strongly face problems. Other examples may drive helping, too yet on average, a lack of identification with their own skills can be associated with lower levels of helping behavior compared to the inverse and clearer case. When workers identify strongly with their skills, they also indirectly identify with all tasks on the factory floor that require these particular skills. Applying the skills in the production is a way to enact workers' identification, and any task that fits their skill profile is a potential source of self-verification. Thus, stronger identification with skills will motivate workers to apply them as often as possible to reinforce their self-concept. Research has shown that individuals draw job satisfaction from identification elements that drive self-esteem and efficacy (Vignoles et al., 2006), such as solving a production task drawing on one's own abilities. This motivation to apply skills does not distinguish between tasks that one was originally assigned and tasks that lie outside one's scope of responsibility. Instead, any completed task can provide FRANKE ET AL.717 self-verification. Thus, helping others who may not have the required skills or are busy with other tasks is desirable from an identification and motivation standpoint when workers strongly identify with their production skills. Thus, we hypothesize: Hypothesis 2. Workers who strongly identify with their skills in production will show more helping behavior compared to workers who identify less with their skills. 3.3 |Effect of helping behavior on productivity The link between helping behavior and productivity remains untested in the literature so far (Cantor & Jin, 2019). In non-interdependent operations, it is only necessary for workers to help each other when all workers with the appropriate skills for solving an open task are busy. The required information, namely the codified skill, its live availability, and the need for its application, can be made fully transparent in digitalized factories. Thus, from a utilization perspective, helping behaviors balance workload across the factory floor. Our study focuses on interruptions of semi-automated machines. Helping behavior can reduce the time a machine is interrupted and waiting for an operator. Thus, it will directly contribute to higher productivity by reducing machine downtimes and increasing unit output. Thus, we hypothesize: Hypothesis 3. Higher levels of helping behavior in automatized and digitalized production environments will be associated with higher levels of productivity. TheconceptualresearchmodelcanbeseeninFigure2. 4|METHODOLOGY This study was pre-registered and conditionally accepted by the Journal of Operations Management prior to conducting the field experiment. Subsequently, we conducted the field experiment as planned and described next. We ran a field experiment introducing an intervention that drives the identification of workers with their individual skills in a manufacturing company to stimulate helping behavior. The nature of the experimental design is a pre-post quasi field experiment. In addition to the treatment group that went through a pretest-posttest design, we simultaneously assessed a control group with no treatment using the difference-in-difference (DiD) technique. Control groups that potentially met the parallel trends assumption (i.e., have a similar pretreatment productivity trajectory) were available within the same plant and in three other plants of the company. The company led the selection of the treatment group and the treatment design. In this process, the research team ensured a stratified random sample selection and that an appropriate control group was selected by testing the parallel trends assumption of the DiD design. 4.1 |Experimental setting The collaborating company is a large supplier and developer of metallurgy parts for automobiles, aerospace equipment, and consumer applications. Simply put, metallurgy uses pressing and thermal treatment to bring metal into shape, which allows the creation of more complex geometries than any chipping processes like lathing or milling would allow. The company produces millions of parts daily, has several thousands of customers worldwide, and employs thousands of employees in an extensive network of plants worldwide. The company provided full access to its facilities and the digital systems to allow the implementation of the field experiment. 4.2 |Experimental treatment The treatment took advantage of the worker-task matching system using smartwatches. The system offers tasks to workers via a list that workers can choose from (see Figure 1). Importantly, the list is individualized such that only those workers who have the required codified skills FIGURE 2 Conceptual research model. We aggregate all variable to the team and shift for analysis. 718 FRANKE ET AL. see a specific task. To enable this, all skills are codified in a matrix that features all available tasks on one dimension and all anonymous worker IDs on the other dimension. The tasks are differentiated by various machine types to make sure that only workers who know a particular machine type will work on its interruptions. The “skill matrix”is a part of the back-end system and is not visible or salient to workers. The main authority to change the skill matrix is with the team leaders. For that reason and due to their intangible nature, skills and their digital codifications are an arguably less salient feature of the factory floor today. The treatment addressed this lack of presence on the factory floor by introducing “skill weeks.”The company occasionally promotes special themes that feature communication to raise awareness or workshops and training. During a recent safety week, for example, workers were motivated to report safety hazards, could order new safety shoes, and participated in first aid training. “Skill weeks”encouraged the reflection on individuals' skills in several ways: first, banners and posters made the initiative visible and transported the goal to raise awareness of how important workers' skills are to the factory. Second, workers participated in voluntary meetings that encouraged exploring, reviewing, and reflecting upon skills and their digital codifications in the skill matrix. These sessions lasted between 10 and 15 min and encouraged workers to reflect on what skills they possess, which ones are important, which ones constitute bottlenecks, or possible training needs they may have. This was accomplished by letting workers map their individual skills and development trajectories. The script of the meetings is shown in Appendix A. Third, workers received information about their skill use in real-time via their smartwatches, similar to micro-interventions in the medical sciences (e.g., Baumel et al., 2020; Fuller-Tyszkiewicz et al., 2019). Messages displayed via the watch included information on the most frequently used skills of the current and previous shifts. Figure B1 shows an example of how the message was displayed on workers' smartwatches. This combination of activities is typical for a themed week at the company and all directly address workers' skills. The company agreed to hold the “skill weeks”for 2 weeks to increase the chance of change in the identification of workers. The treatment focused on emphasizing skills, and skills are defined by the type of machine that a worker can operate and troubleshoot. This natural connection between skill and machine type is inherent to the production system, if not to any production that operates several machines of the same type. Our study acknowledges this inherent connection, but the treatment did not emphasize it since we want to increase identification with skills, not machines. Thus, the treatment avoided making references to machines (see Appendix Aand Figure B1) but focused on skills instead. In other words, our treatment focused on identifying with the skill that fits a whole set of machines instead of only a few. It thus included the potential to broaden a worker's action radius in the factory to enhance broader collaboration among workers. We expected that the treatment would strengthen the cognitive and emotional attachment that workers make to their own skills, thus, their identification with skills (Ashforth et al., 2008; Miscenko & Day, 2016). The manipulation was unobtrusive as it neither directly concerns helping behavior nor productivity. We adopted the definition of altruistic helping behavior grounded in the idea of voluntary, not mandated, helping behavior (Cantor & Jin, 2019; Organ, 1988). Therefore, the treatment did not introduce a new policy or incentive that explicitly or implicitly enforces helping. 4.3 |Manipulation check We use several sources of information to verify that workers have experienced the treatment. First, we assessed the attendance of workers in group sessions targeted at discussing skills. Only one worker chose not to attend one of the 12 sessions. Second, we questioned workers on whether they had noticed and read banners or posters announcing the treatment and initiatives that were part of it. Most workers noticed the banners. Finally, we conducted a manipulation check (i.e., if the intervention induces a change in skill identification) using a survey question measuring how strongly workers identify with their skills at work (see Appendix C). We report a significant increase in the scale instrument as part of the results. 4.4 |Sample and pre-test survey The sample comprises workers in a production area of the company that performs the pressing process of small metal components. The entire production process includes forming, calibrating, and thermal treatment. We focus on the two steps of forming and calibrating as they include many parallel automatic processes that rely on workers to help each other when interruptions occur. Our study focuses on the presses. Each press has an integrated palletizer for automatically storing pressed parts. We do not focus on the thermal treatment as it is a separate step in the production process since it is not batched, as the pressing processed, but involves more flow. The treatment group was a factory in Germany that FRANKE ET AL.719 is pivotal for establishing one's professional identity within a specific field. While emotional exhaustion is only a control variable in our work, the interplay between identity at work and burnout has been a subject of interest in various studies. A significant correlation between professional identity and burnout has been documented across diverse industries (Chen et al., 2020; Sabanciogullari & Dogan, 2015; Zhang et al., 2021). Our intervention specifically targeted an element of professional identity—the enhancement of workers' identification with their skills. By focusing on this aspect, we aimed to bolster the professional self-concept, thereby strengthening the individual's perception of themselves as integral members of their profession. This enhanced self-concept could potentially make work tasks feel less burdensome, as they are seen as intrinsic to one's identity rather than external obligations. Consequently, this shift in perception may lead to a decreased risk of burnout (Zhang et al., 2021). These findings imply that our intervention likely augmented the professional identity, serving as a buffer against burnout. This has substantial implications for the well-being of workers and their intentions to remain with the organization (Campbell et al., 2013). Future studies must further investigate these effects in more detail. 6.5 |Summary and limitations We have discussed that (i) our manipulation may not have been strong enough to show an effect in a real factory; (ii) that two mechanisms based on two different sub-identities may be canceling each other out; (iii) that our treatment design may have been overly conservative and therefore was limited in its strength; (iv) and that our treatment has shown an unforeseen but welcome reduction in emotional exhaustion. These insights deliver new contributions to the discussion on identification at work, the research on burnout, and for future treatment design in the context of identification with skills. We also contribute to the behavioral operations literature on worker collaboration. We complement the prior work that has focused on the configuration of incentives (De Vries et al., 2016; Siemsen et al., 2007) and extend the discussion around what factors of intrinsic motivation can drive performance in operational teams (Franke et al., 2022). Specifically, we suggest that when teams are not implicitly forced to cooperate via task interdependence (Bachrach et al., 2006; Schoenherr et al., 2017), it is important to disentangle the competing effects of identification with skills and machines. Beyond these values of our study, it is worth noting that it faces several inherent limitations that inform future research that may support our theory in other contexts. First, we are examining identification in a specific digitalized production system that involves operators that solve randomly occurring machine interruptions. Whenever work tasks do not appear randomly but according to a predetermined sequence, we cannot transfer our null results to those settings. This holds particularly true for other technology-enabled production organizations, especially those that follow a different process type compared to the batched mass production in our case. Digitalized assembly lines or job shops may deliver different conclusions. Assembly lines, for example, are characterized by high interdependency between process steps, which is not the case in our study. In our experiment, machine interruptions occur independently from each other. Moreover, job shops may include a much broader range of activities and different machines that can drive cognitive switching costs when workers move from one task to another. Finally, the cooperating company is a traditional manufacturer, and the workforce has an average tenure of about 10 years. Our results may not compare to organizations without a long tradition or ad-hoc production where workers have built fewer long-standing relations with each other. 6.6 |Managerial contribution This study highlights a little-known mechanism on the factory floor: identification. We argued initially that identification with skills could be a non-tangible and low-cost leverage point to improve helping behavior—but we could not prove its effect in a field experiment in the manufacturing industry. Our study contributes to illuminating workers' identification change as an important managerial challenge. We were able to increase workers' identification with their own skills by about 7% over the course of 2 weeks, applying an intensive program of meetings, notifications via wearables, and visible posters and roll-ups. Reaping the rewards of higher identification with skills in the work context has proven to be a truly challenging endeavor since we could not show that the higher identification also translated into more cooperation among the workforce or overall productivity. However, we could observe a reduction in workers' emotional exhaustion. Our discussion, considering our theory and based on managerial insights, suggests that emphasizing skills on factory floors may also automatically emphasize the relevance of the machines and equipment. Skills and machines are inherently connected in our study. This leads to a possible cancelation effect: identification with skills may enhance helping, but this effect remains concealed since workers' increasing identification with 726 FRANKE ET AL. specific machines makes them simultaneously avoid helping at machines that are not “theirs.”This dual process requires further exploration in studies on identification in manufacturing. FUNDING INFORMATION This project receives funding from a research grant at ETH Zurich (CareerSeed Grant SEED-06 21-1). We received no funding from the industrial partner. Open Access funding enabled and organized by Projekt DEAL. [Correction added on 13 April 2024, after first online publication: Projekt DEAL funding statement has been added.] ETHICS STATEMENT This research design has been accepted after review by the Ethics Commission for research with human participants at ETH Zurich. ORCID Henrik Franke https://orcid.org/0000-0002-2336-4312 Daniel Kwasnitschka https://orcid.org/0009-0004-68899662 Jan B. Schmutz https://orcid.org/0000-0002-0181-807X Torbjørn H. Netland https://orcid.org/0000-0001-73821051 REFERENCES Adair, J. G. (1984). The Hawthorne effect: A reconsideration of the methodological artifact. Journal of Applied Psychology,69(2), 334–345. Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckmann (Eds.), Action control (pp. 11–39). Springer. Albert, S., Ashforth, B. E., & Dutton, J. E. (2000). Organizational identity and identification: Charting new waters and building new bridges. Academy of Management Review,25(1), 13–17. Ashforth, B. E., Harrison, S. H., & Corley, K. G. (2008). Identification in organizations: An examination of four fundamental questions. Journal of Management,34(3), 325–374. Bachrach, D. G., & Bendoly, E. (2011). Rigor in behavioral experiments: A basic primer for supply chain management researchers. Journal of Supply Chain Management,47(3), 5–8. Bachrach, D. G., Powell, B. C., Collins, B. J., & Richey, R. G. (2006). Effects of task interdependence on the relationship between helping behavior and group performance. Journal of Applied Psychology,91(6), 1396–1405. Bagozzi, R. P., & Dholakia, U. M. (2006). Open source software user communities: A study of participation in Linux user groups. Management Science,52(7), 1099–1115. Bai, B., Dai, H., Zhang, D. J., Zhang, F., & Hu, H. (2022). The impacts of algorithmic work assignment on fairness perceptions and productivity: Evidence from field experiments. Manufacturing & Service Operations Management,24(6), 3060–3078. Bataille, C. D., & Vough, H. C. (2022). More than the sum of my parts: An intrapersonal network approach to identity work in response to identity opportunities and threats. Academy of Management Review,47(1), 93–115. Baumel, A., Fleming, T., & Schueller, S. M. (2020). Digital micro interventions for behavioral and mental health gains: Core components and conceptualization of digital micro intervention care. Journal of Medical Internet Research,22(10), 1–14. Bendoly, E., Croson, R., Goncalves, P., & Schultz, K. (2010). Bodies of knowledge for research in behavioral operations. Production and Operations Management,19(4), 434–452. Bendoly, E., Donohue, K., & Schultz, K. L. (2006). Behavior in operations management: Assessing recent findings and revisiting old assumptions. Journal of Operations Management,24(6), 737–752. Bendoly, E., Thomas, D., & Capra, M. (2010). Multilevel social dynamics considerations for project management decision makers: Antecedents and implications of group member tie development. Decision Sciences,41(3), 459–490. Brewer, M. B. (1991). The social self: On being the same and different at the same time. Personality and Social Psychology Bulletin, 17(5), 475–482. Brewer, M. B., & Gardner, W. (1996). Who is this “We”? Levels of collective identity and self representations. Journal of Personality and Social Psychology,71(1), 83–93. Burke, P. J. (1991). Identity processes and social stress. American Sociological Review,6, 836–849. Campbell, N. S., Perry, S. J., Maertz, C. P., Jr., Allen, D. G., & Griffeth, R. W. (2013). All you need is…resources: The effects of justice and support on burnout and turnover. Human Relations,66(6), 759–782. Cantor, D. E., & Jin, Y. (2019). Theoretical and empirical evidence of behavioral and production line factors that influence helping behavior. Journal of Operations Management,65, 312–332. Casoria, F., Reuben, E., & Rott, C. (2022). The effect of group identity on hiring decisions with incomplete information. Management Science,86(8), 5557–6354. Cerasoli, C. P., Nicklin, J. M., & Ford, M. T. (2014). Intrinsic motivation and extrinsic incentives jointly predict performance: A 40-year meta-analysis. Psychological Bulletin,140(4), 980–1008. Charness, G., & Holder, P. (2019). Charity in the laboratory: Matching, competition, and group identity. Management Science, 65(3), 1398–1407. Chen, H., Liu, F., Pang, L., Liu, F., Fang, T., Wen, Y., Chen, S., Xie, Z., Zhang, X., & Zhao, Y. (2020). Are you tired of working amid the pandemic? The role of professional identity and job satisfaction against job burnout. International Journal of Environmental Research and Public Health,17(24), 9188. Chiniara, M., & Bentein, K. (2018). The servant leadership advantage: When perceiving low differentiation in leader-member relationship quality influences team cohesion, team task performance and service OCB. The Leadership Quarterly,29(2), 333–345. Cooper, W. H., & Richardson, A. J. (1986). Unfair comparisons. Journal of Applied Psychology,71(2), 179–184. Corsten, D., Gruen, T., & Peyinghaus, M. (2011). The effects of supplier-to-buyer identification on operational performance— An empirical investigation of inter-organizational identification in automotive relationships. Journal of Operations Management,29(6), 549–560. Croson, R., Schultz, K., Siemsen, E., & Yeo, M. (2013). Behavioral operations: The state of the field. Journal of Operations Management,31(1), 1–5. FRANKE ET AL.727 De Quidt, J., Haushofer, J., & Roth, C. (2018). Measuring and bounding experimenter demand. American Economic Review, 108(11), 3266–3302. De Vries, J., De Koster, R., & Stam, D. (2016). Aligning order picking methods, incentive systems, and regulatory focus to increase performance. Production and Operations Management, 25(8), 1363–1376. Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin,125(6), 627–668. Del Carpio, L., & Guadalupe, M. (2022). More women in tech? Evidence from a field experiment addressing social identity. Management Science,68(5), 3196–3218. Donohue, K., Özer, Ö., & Zheng, Y. (2020). Behavioral operations: Past, present, and future. Manufacturing & Service Operations Management,22(1), 191–202. Dukerich, J. M., Golden, B. R., & Shortell, S. M. (2002). Beauty is in the eye of the beholder: The impact of organizational identification, identity, and image on the cooperative behaviors of physicians. Administrative Science Quarterly,47(3), 507–533. Dutton, J. E., Dukerich, J. M., & Harquail, C. V. (1994). Organizational images and member identification. Administrative Science Quarterly,39(2), 239–263. Eckerd, S., DuHadway, S., Bendoly, E., Carter, C. R., & Kaufmann, L. (2020). On making experimental design choices: Discussions on the use and challenges of demand effects, incentives, deception, samples, and vignettes. Journal of Operations Management,67(2), 261–275. Fahimnia, B., Pournader, M., Siemsen, E., Bendoly, E., & Wang, C. (2019). Behavioral operations and supply chain management— A review and literature mapping. Decision Sciences,50(6), 1127–1183. Franke, H., Eckerd, S., & Foerstl, K. (2022). Rising to the top: Motivational forces influencing status conflict in sourcing teams. Production and Operations Management,31(3), 963–983. Franke, H., Foerstl, K., & Heese, H. S. (2021). The interaction effect of goal misalignment and metaknowledge distribution on team decision making in operations and supply chain management. Decision Sciences,52(2), 331–361. Fuller-Tyszkiewicz, M., Richardson, B., Lewis, V., Linardon, J., Mills, J., Juknaitis, K., Lewis, C., Coulson, K., O'Donnell, R., & Arulkadacham, L. J. C. I. H. B. (2019). A randomized trial exploring mindfulness and gratitude exercises as eHealth-based micro-interventions for improving body satisfaction. Computers in Human Behavior,95,58–65. Gao, Y., Li, M., & Sun, S. (2023). Field experiments in operations management. Journal of Operations Management,69, 676–701. Gecas, V. (1982). The self-concept. Annual Review of Sociology,8, 1–33. Grant, A. M., & Wall, T. D. (2009). The neglected science and art of quasi-experimentation: Why-to, when-to, and how-to advice for organizational researchers. Organizational Research Methods, 12(4), 653–686. Hu, J., & Liden, R. C. (2015). Making a difference in the teamwork: Linking team prosocial motivation to team processes and effectiveness. Academy of Management Journal,58(4), 1102–1127. Janssen, O., & Huang, X. (2008). Us and me: Team identification and individual differentiation as complementary drivers of team members' citizenship and creative behaviors. Journal of Management,34(1), 69–88. Johnson, M. D., Morgeson, F. P., & Hekman, D. R. (2012). Cognitive and affective identification: Exploring the links between different forms of social identification and personality with work attitudes and behavior. Journal of Organizational Behavior,33(8), 1142–1167. Kilcullen, M., Bisbey, T. M., Rosen, M., & Salas, E. (2022). Does team orientation matter? A state-of-the-science review, metaanalysis, and multilevel framework. Journal of Organizational Behavior,44, 355–375. Kristensen, T. S., Borritz, M., Villadsen, E., & Christensen, K. B. (2005). The Copenhagen burnout inventory: A new tool for the assessment of burnout. Work & Stress,19(3), 192–207. Landis, R. S., James, L. R., Lance, C. E., Pierce, C. A., & Rogelberg, S. G. (2014). When is nothing something? Editorial for the null results special issue of Journal of Business and Psychology.Journal of Business and Psychology,29(2), 163–167. Lorinkova, N. M., & Perry, S. J. (2019). The importance of groupfocused transformational leadership and felt obligation for helping and group performance. Journal of Organizational Behavior,40(3), 231–247. Malhotra, M. K., Ahire, S., & Shang, G. (2017). Mitigating the impact of functional dominance in cross-functional process improvement teams. Decision Sciences,48(1), 39–70. Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of Psychology,52(1), 397–422. Mathieu, J., Maynard, M. T., Rapp, T., & Gilson, L. (2008). Team effectiveness 1997-2007: A review of recent advancements and a glimpse into the future. Journal of Management,34(3), 410–476. Mathieu, J. E., Gallagher, P. T., Domingo, M. A., & Klock, E. A. (2019). Embracing complexity: Reviewing the past decade of team effectiveness research. Annual Review of Organizational Psychology and Organizational Behavior,6,17–46. Mervis, J. (2014). Why null results rarely see the light of day. Science,345(6200), 992. Mesmer-Magnus, J. R., & DeChurch, L. A. (2009). Information sharing and team performance: A meta-analysis. Journal of Applied Psychology,94(2), 535–546. Miller, C. C., & Bamberger, P. (2016). Exploring emergent and poorly understood phenomena in the strangest of places: The footprint of discovery in replications, meta-analyses, and null findings. Academy of Management Discoveries,2(4), 313–319. Miscenko, D., & Day, D. V. (2016). Identity and identification at work. Organizational Psychology Review,6(3), 215–247. Oliva, R., & Watson, N. (2011). Cross-functional alignment in supply chain planning: A case study of sales and operations planning. Journal of Operations Management,29(5), 434–448. Organ, D. W. (1988). Organizational citizenship behavior: The good soldier syndrome. Lexington Books. Parker, S. K., & Grote, G. (2022). Automation, algorithms, and beyond: Why work design matters more than ever in a digital world. Applied Psychology,71(4), 1171–1204. Pratt, M. G., Rockmann, K. W., & Kaufmann, J. B. (2006). Constructing professional identity: The role of work and identity learning cycles in the customization of identity among medical residents. Academy of Management Journal,49(2), 235–262. 728 FRANKE ET AL. Rea, D., Froehle, C., Masterson, S., Stettler, B., Fermann, G., & Pancioli, A. (2021). Unequal but fair: Incorporating distributive justice in operational allocation models. Production and Operations Management,30(7), 2304–2320. Reagans, R. (2005). Preferences, identity, and competition: Predicting tie strength from demographic data. Management Science, 51(9), 1374–1383. Sabanciogullari, S., & Dogan, S. (2015). Relationship between job satisfaction, professional identity and intention to leave the profession among nurses in Turkey. Journal of Nursing Management,23(8), 1076–1085. Schoenherr, T., Bendoly, E., Bachrach, D. G., & Hood, A. C. (2017). Task interdependence impacts on reciprocity in IT implementation teams: Bringing out the worst in us, or driving responsibility? Production and Operations Management,26(4), 667–685. Siemsen, E., Balasubramanian, S., & Roth, A. V. (2007). Incentives that induce task-related effort, helping, and knowledge sharing in workgroups. Management Science,53(10), 1533–1550. Ta, H., Esper, T. L., & Hofer, A. R. (2018). Designing crowdsourced delivery systems: The effect of driver disclosure and ethnic similarity. Journal of Operations Management,60,19–33. Tajfel, H., & Turner, J. C. (1982). Social psychology of intergroup relations. Annual Review of Psychology,33(1), 1–39. Turner, J. C., & Tajfel, H. (1986). The social identity theory of intergroup behavior. In A. Worchel (Ed.), Psychology of intergroup relations (pp. 7–24). Nelson. Urda, J., & Loch, C. H. (2013). Social preferences and emotions as regulators of behavior in processes. Journal of Operations Management,31(1), 6–23. Van Der Vegt, G. S., Van De Vliert, E., & Oosterhof, A. (2003). Informational dissimilarity and organizational citizenship behavior: The role of intrateam interdependence and team identification. Academy of Management Journal,46(6), 715–727. Van Dyne, L., & LePine, J. A. (1998). Helping and voice extra-role behaviors: Evidence of construct and predictive validity. Academy of Management Journal,41(1), 108–119. Vignoles, V. L., Regalia, C., Manzi, C., Golledge, J., & Scabini, E. (2006). Beyond self-esteem: Influence of multiple motives on identity construction. Journal of Personality and Social Psychology,90(2), 308–333. Wanous, J. P., Reichers, A. E., & Hudy, M. J. (1997). Overall job satisfaction: How good are single-item measures? Journal of Applied Psychology,82(2), 247–252. Wu, C.-H., Liu, J., Kwan, H. K., & Lee, C. (2016). Why and when workplace ostracism inhibits organizational citizenship behaviors: An organizational identification perspective. Journal of Applied Psychology,101(3), 362–378. Zhang, T., Feng, J., Jiang, H., Shen, X., Pu, B., & Gan, Y. (2021). Association of professional identity, job satisfaction and burnout with turnover intention among general practitioners in China: Evidence from a national survey. BMC Health Services Research,21,1–11. How to cite this article: Franke, H., Kwasnitschka, D., Schmutz, J. B., & Netland, T. H. (2024). Emphasizing worker identification with skills to increase helping and productivity in production: A field experiment. Journal of Operations Management,70(5), 712–732. https:// doi.org/10.1002/joom.1300 FRANKE ET AL.729 APPENDIX A Script for the Meetings of the Skills Weeks (to be distributed among researchers and team leaders beforehand) Script for 4 Mini-Meetings on Skills, each lasting 10–15 min (Responsible: researchers and managers; blinded) Guideline for all of us: The meeting should solely focus on individual skills, not on other aspects such as skill distribution within the team, team collaboration, compensation, the smartwatch system, technical issues, and so on. Meeting 1: •Employees receive an empty skill matrix and fill in their skills for solving interruptions. •Then, each person receives their matrix as it is stored in the system. •Objective: Comparison of self-assessment and system assessment. •If desired, the employee can discuss the results with the team leader and request changes (control variable). Meeting 2: •HR provides information about current or upcoming skill development trainings. •HR informs about current or future needs of (COMPANY) pertaining to skills. •Development opportunities for employees' skills at (COMPANY). •Illustrate ways and requirements to acquire set-up skills, for example. Meeting 3: •Employees receive their individual skill matrix as it is stored in the system. •Which skills do I want to learn in addition or expand upon? •Which skills can I learn particularly quickly based on my existing skills? Which ones are more challenging? •How does this align with (COMPANY)'s offerings and plans (referring to Meeting 2)? Meeting 4: •Employees receive their individual skill matrix as it is stored in the system. •Which skills outside of the skills matrix do you want to learn or expand? •Which skills are important for production that are not represented in the system? After the Skills Weeks: •Mutual feedback. What went well? What needs improvement? Objective Participants Approach •Promote identification of employees with their skills •Employees of one shift •Interactive •One team leader •Open communication •One or two researchers •No performance measurement or evaluation as goal 730 FRANKE ET AL. APPENDIX B APPENDIX C C.1 |SCALE INSTRUMENTS This appendix shows the single-item measures that were gathered via the smartwatch during every shift in the treatment group and the full measure structure. Short measures such as single-item measures are appropriate when constructs are unambiguous and have been used in operations management (Rea et al., 2021; Wanous et al., 1997). We show the dropped items (white background) along with the single items we used in the study (gray background). Identification with Skills (subscale of Johnson et al., 2012) (Not at all 1—Very much 5) •My self-identity is based on the skills I use at work. (dropped) •My skills are very important to my sense of who I am at work. •My sense of self overlaps with the skills I required for my work. (dropped) •If someone criticized my skills that would influence how I thought about myself. (dropped) •I identify strongly with my skills at work. (dropped) Helping Behavior (short scale, Van Dyne & LePine, 1998) (Not at all 1—Very much 5) •I help others in this group with their work responsibilities. •I get involved to benefit this work team. (dropped) •I volunteer to do things for this team. (dropped) •I assist others in this group with their work for the benefit of this team. (dropped) Intention to Help (based on Ajzen, 1985) (Very unlikely 1—very likely 7) •How likely is it that you will help a colleague in the next shift? Identification with Machines (subscale of Johnson et al., 2012) (Not at all 1—Very much 5) •My self-identity is based in part on the machines I commonly use at work. (dropped) •The machines I commonly use are very important to my sense of who I am at work. •My sense of self overlaps with the machines I commonly use for my work. (dropped) FIGURE B1 Skill message delivered via the smartwatch. The message shows codes for those skills that are the top 5 most frequently used skills of a worker. These codes are commonly known among management and the workforce. Messages like these were delivered at the beginning of each shift, reporting on the previous shift of that particular worker. A similar message was delivered at halftime of a shift, reporting on the current day for a particular worker. FRANKE ET AL.731 •If someone criticized the machines I commonly use at work that would influence how I thought about myself. (dropped) •I identify strongly with the machines that I commonly use at work. (dropped) Emotional Exhaustion (subscale of Kristensen et al., 2005) (Never 1—Always 5) How often are the following statements true? (dropped) •Do you feel burnt out because of your work? •Does your work frustrate you? (dropped) •Do you have enough time for family and friends during leisure time? (dropped). APPENDIX D TABLE D1 Main analysis for Hypothesis 1: DiD regression predicting productivity. Estimate Standard error tValue p-Value CI lower bound CI upper bound Intercept 0.433 0.128 3.393 .001 0.180 0.686 Treated group (group fixed effect) 0.058 0.055 1.050 .296 0.168 0.052 Post (time fixed effect) 0.049 0.028 1.749 .083 0.104 0.006 Late shift 0.025 0.020 1.253 .213 0.015 0.065 Night shift 0.103 0.021 4.84 .000 0.145 0.061 Workload 0.230 0.580 0.397 .692 0.918 1.378 Treated group Post 0.009 0.036 0.261 .795 0.081 0.062 Note: Regression uses robust standard errors; Including the survey-based control variables was not possible in the DiD regressions since the control group could not be surveyed in the field; R 2 -adjusted: 0.349. TABLE D2 Main analysis for Hypothesis 2: DiD regression predicting helping behavior. Estimate Standard error tValue p-Value CI lower bound CI upper bound (Intercept) 0.424 0.168 2.522 .013 0.091 0.757 Treated group (group fixed effect) 0.098 0.068 1.454 .149 0.232 0.036 Post (time fixed effect) 0.046 0.036 1.301 .196 0.024 0.117 Late shift 0.015 0.025 0.600 .549 0.034 0.064 Night shift 0.023 0.024 0.963 .338 0.070 0.024 Workload 0.245 0.726 0.338 .736 1.192 1.683 Treated group Post 0.017 0.04 0.417 .677 0.063 0.097 Note: Regression uses robust standard errors; Including the survey-based control variables was not possible in the DiD regressions since they are only available for the treatment group; R 2 -adjusted: 0.228. TABLE D3 Main analysis for Hypothesis 3: Regression predicting productivity. Estimate Standard error tValue p-Value CI lower bound CI upper bound Intercept 0.490 0.202 2.424 .019 0.084 0.896 Helping behavior 0.090 0.088 1.022 .311 0.265 0.086 Post (time fixed effect) 0.028 0.031 0.924 .360 0.090 0.033 Late shift 0.040 0.023 1.736 .088 0.006 0.087 Night shift 0.010 0.023 0.422 .674 0.057 0.037 Workload 0.844 0.722 1.169 .248 2.292 0.605 Emotional exhaustion 0.018 0.020 0.900 .372 0.059 0.022 Identification with machines 0.023 0.029 0.786 .435 0.035 0.081 Note: Regression uses robust standard errors; Group fixed effects cannot be modeled in this regression because the bottom two control variables are only available for the treatment group; R 2 -adjusted: 0.162. 732 FRANKE ET AL.