1 Introduction

Trust is an essential component of workplace relationships. Having a trustworthy supervisor enhances subordinates’ performance at work, including task performance and contextual performance, and increases their job satisfaction (Goris et al. 2003; Colquitt et al. 2007; Ugwu et al. 2016; Legood et al. 2020). Moreover, feeling trusted (i.e., felt trust) by the supervisor enhances one’s task performance, job engagement, and organizational citizenship behavior (OCB) (Salamon and Robinson 2008; Lau et al. 2014, 2018; Skiba and Wildman 2019; Zheng et al. 2019). The current study explores the relationship between feeling trusted by supervisors and OCB intention, and investigates specifically the moderators of this relationship.

Prior research on trust from the followers’ perspective had extensively concentrated on their trust towards their leader (Zheng et al. 2019), and only a few have focused on the influence of followers’ feeling of being trusted by the leader. Studies on felt trust have proliferated in recent years with some addressing the relationship between felt trust and job performance and the mediators of this relationship (e.g., Skiba and Wildman 2019; Zheng et al. 2019). However, the moderators of this relationship have rarely been explored. Identifying and examining boundary conditions within the felt trust literature represents a needed theoretical progression. Moderation studies have the potential to contribute to organizational practices. This current study intends to fill the knowledge gap in the literature of felt trust by examining three potential moderators based on the trust model proposed by Mayer et al. (1995) and five perspectives of the mechanism of felt trust (Lau et al. 2014; Baer et al. 2015; Skiba and Wildman 2019; Zheng et al. 2019).

This study addressed the research question of how trustor characteristics and the cost of being trusted place limitations on the relationship between felt trust and job performance. This exploration not only responds to the argument of Lau et al. (2014) but can justify underlying dynamics of felt trust effect on job performance proposed by prior studies (e.g., Lau et al. 2014; Baer et al. 2015; Skiba and Wildman 2019; Zheng et al. 2019). According to the model of trust (Mayer et al. 1995), trust and its outcomes involve characteristics of the trustee, characteristics of the trustor, and perceived risks in the relationship. In the same vein, this study selected supervisor behavioral integrity and non-work relationship between the subordinate and the supervisor as potential moderators of the relationship between felt trust and OCB intention. In terms of perceived risk or cost involved, this study selected perceived stress from OCB as a potential moderator.

The justification of the three variables as potential moderators in the relationship between felt trust and OCB intention relies on five mechanisms of felt trust proposed by prior studies. The five perspectives, each with their respective mediator, explain how felt trust influences job performance. These include the following: (1) social exchange perspective, which specified felt obligation as the mediator (e.g., Salamon and Robinson 2008); (2) self-evaluative perspective, which specified organization-based self-esteem (OBSE) as the mediator (e.g., Lau et al. 2014); (3) social cognitive perspective, which specified occupational self-efficacy (OSE) as the mediator (Zheng et al. 2019); (4) self-determination perspective, which specified perceived autonomy as the mediator (Skiba and Wildman 2019); and (5) conservation of resources perspective, which specified perceived workload and emotional exhaustion as mediators (Baer et al. 2015). Based on the aforementioned perspectives, supervisor behavioral integrity may moderate the effects of felt trust on OCB intention through its enhancement of OBSE, OSE, and feelings of pride. Further, close relationship outside work with the supervisor may moderate the effect of felt trust on OCB intention through its enhancement of felt obligation to reciprocate. Lastly, perceived OCB stress may moderate the effect of felt trust on OCB intention through its effect on perceived workload.

What constitutes a theoretical contribution may involve the building blocks of a theory which include what (which factors are relevant), how (how are the factors related), and why (what are the underlying dynamics that justify the selection of factors and the proposed causal relationships), or involve who, where, and when (the boundary conditions of a theoretical model) (Whetten 1989). The theoretical contributions of this study mainly rely on the boundary conditions of the model of felt trust effect on job performance (see Fig. 1). In addition, the examination of the three boundary factors (moderators) contributes to the why questions of the felt trust model.

Fig. 1
figure 1

Theoretical model of felt trust. Note The gray ones are studied variables

Full size image

In terms of theoretical contributions to boundary conditions, Prior research has addressed the relationship between felt trust and job performance and the mediators of this relationship (e.g., Skiba and Wildman 2019; Zheng et al. 2019). What has rarely been examined is the moderators of this relationship. Incorporating the model of trust (Mayer et al. 1995) and mechanisms of felt trust proposed by prior studies (e.g., Salamon and Robinson 2008; Lau et al. 2014; Baer et al. 2015; Zheng et al. 2019; Skiba and Wildman 2019), this study explored how trustor characteristics and the cost of being trusted intervene in the relationship between felt trust and job performance.

Behavioral integrity is vital for building positive interpersonal relationships (Simons 2002; Tsai and Tsai 2022). A lack of supervisor behavioral integrity in the work environment may serve as a barrier, diminishing the potential benefits of supervisors’ trust on employees’ OBSE, OSE, and feelings of pride, and vice versa. Workplace relationships notably shape individuals’ work behaviors and outcomes (Trefalt 2013). Close relationship outside work with the supervisor may amplifying the positive effects of employees’ felt trust based on the social exchange perspective (Cropanzano and Mitchell 2005). OCB is not without cost (e.g., Bergeron 2007; Bolino et al. 2015). However, how the cost influences the relationship between felt trust and OCB has not yet been fully clarified. Baer et al. (2015) suggested that the nature of employees’ tasks might be an important factor in understanding the dynamics of feeling trusted. Perceived OCB stress may be another contextual limit of the benefits of felt trust.

In terms of theoretical contributions to underlying dynamics of felt trust effect, there are inconsistent results on the mediating role of OBSE in the relationship between felt trust and job performance. Zheng et al. (2019) found a significant mediating effect of OBSE on the relationship between felt trust and OCBO (organizational citizenship behavior-organization), contrary to the finding of Lau et al. (2014). Our results provide new evidence to solve the controversy. Moreover, there exists inconsistency on the mediating role of felt obligation in the relationship between felt trust and job performance. Lau et al. (2018) found that felt obligation to reciprocate is a significant mediator in the relationship between felt trust and OCB, contrary to the finding of Skiba and Wildman’s (2019) study. Our results offer new evidence justifying whether felt obligation is a mediator in the relationship between felt trust and job performance.

2 Theoretical perspectives and hypothesis development

2.1 Felt trust and OCB

The core concept of trust is the willingness to make oneself vulnerable to another in which it is expected to yield positive outcomes without consciously monitoring its effects (Lewicki et al. 2006; Mayer et al. 1995). These expectations are based on the perceptions of the trustee’s trustworthiness, e.g., benevolence, integrity, and ability (Mayer et al. 1995). In contrast, felt trust is the trustee’s perception that the trustor is willing to rely on the trustee and expects positive outcomes (Lau et al. 2014; Williams 2016). Lester and Brower (2003) defined felt trust in the workplace as the extent to which subordinates perceive that their supervisor considers them as trustworthy.

Empirical evidence revealed that felt trust enhances OCB 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” (Organ, 1988, p. 4). Several studies demonstrated that felt trust is a significant predictor of OCB, organizational commitment, and job performance (Lester and Brower 2003; Lau and Lam 2008; Brower et al. 2009; Kim et al. 2022). Lau et al. (2014) found that felt trust enhances task performance and OCB. Skiba and Wildman (2019) found that felt trust increases job engagement which is an antecedent of OCB (Ko et al., 2018). Bush et al. (2021) found that prevention-focused ethical leadership has a negative indirect effect on OCBO through felt trust; promotion-focused ethical leadership has a positive indirect effect on OCBO through felt trust. Cho et al. (2021) found that a leader’s humble leadership behaviors are positively related to employees’ OCB via feeling trusted by their supervisor. Chen et al. (2021) found that trusted leaders are more likely to have strong feelings of obligation toward their subordinates and to further engage in benevolent leadership behavior to repay the kindness. All these studies pointed out that felt trust plays a significant role in facilitating OCB.

2.2 Mechanisms of the effect of felt trust

A few studies have focused on the mechanism of felt trust. Lau et al. (2018) called for rigorous theoretical perspectives to explain the relationship between felt trust and job performance. Previous literature proposed five perspectives to explain the effect of felt trust, which are: (1) social exchange, (2) self-evaluative, (3) social cognitive, (4) self-determination, and (5) resource conservation.

In terms of the social exchange perspective, Skiba and Wildman (2019) argued that supervisors are the direct providers of many resources that their subordinates value. Supervisors can initiate an exchange relationship with a subordinate by demonstrating trust through delegating high-risk assignments. When subordinates observe that their supervisors rely on them, they will anticipate greater opportunities to develop new marketable skills and demonstrate competence in exchange for valued resources in the future. Therefore, trusted subordinates will feel more obligated to exert extra efforts to reciprocate and expand the social exchange relationship. Similarly, Brower et al. (2009) asserted that when supervisors trust their subordinates, the latter are more likely to exhibit favorable attitudes toward the exchange relationship, which makes them feel obligated to reciprocate. In addition, Salamon and Robinson (2008) argued that felt trust can be regarded as a rewarding psychological commodity which may evoke feelings of obligation to reciprocate, which would lead to increased work performance in employees.

Empirical evidence showed mixed results in support of felt obligation as a mediator of the relationship between felt trust and job performance. Salamon and Robinson (2008) found that the relationship between felt trust and sales performance is fully mediated by responsibility norms. Moreover, Skiba and Wildman (2019) found that felt trust increases job engagement through increased felt obligation. However, Lau et al. (2018) found a lack of support for the social exchange perspective.

The self-evaluative perspective suggests that individuals are motivated to enhance or maintain self-esteem (Gecas 1982). Felt trust triggers pride, a positive emotion strongly related to self-esteem (Baer et al. 2015). In the workplace, supervisors represent sources of salient social information. Supervisors’ trust, representing a positive evaluation of employees, is likely to enhance employee OBSE (Lau et al. 2014), which is the degree to which an individual believes that he/she is capable, significant, and worthy of being a member of the organization (Pierce and Gardner 2004). Consequently, based on the self-evaluative perspective, positive information from the supervisor is likely to enhance employee OBSE, which may lead to employees making extra efforts to improve work performance to remain valuable to the trustor (Lau et al. 2014; Zheng et al. 2019).

Empirical evidence showed mixed results in support of OBSE as a mediator of the relationship between felt trust and job performance. Lau et al. (2014) found that the mediating effect of OBSE on the relationship between felt trust and job performance is significant for task performance only, and is not significant for OCBI (organizational citizenship behavior-individual) and OCBO. In contrast, Zheng et al. (2019) found that the mediating effect of OBSE on the relationship between felt trust and job performance is significant only for OCBO, and is not for OCBI and task performance.

Self-efficacy refers to the individuals’ beliefs about their ability to achieve goals through their own actions, which strongly influences their thought, motivation, action, and performance (Bandura 1986). According to the social cognitive theory, social persuasion is an important source of individual self-efficacy. Felt trust may function as a type of social persuasion to enhance self-efficacy (Zheng et al. 2019). In other words, felt trust may increase task effort, persistence, and job performance through its enhancement of self-efficacy. Empirical evidence supported self-efficacy as a mediator predicted by the social cognitive perspective. Zheng et al. (2019) found that occupational self-efficacy (OSE) significantly mediates the relationship of felt trust with both in-role and extra-role performance even when the OBSE is controlled.

The self-determination theory asserts that individuals’ perceptions of autonomy—the perceived freedom to act according to one’s own volitions—facilitates their intrinsic motivation, engagement, and well-being (Gagné and Deci 2005). Since managers are more likely to delegate an important task to a trusted subordinate, the empowerment granted would motivate the subordinate to exhibit more productive and prosocial behaviors intrinsically (Brower et al. 2009). Subordinates who feel trusted are likely to feel that they are able to make decisions according to their own volitions, leading to increased perceived autonomy and added effort to work beyond their prescribed roles (Skiba and Wildman 2019). Empirical evidence showed support for the mediator predicted by the self-determination perspective. Skiba and Wildman (2019) found that felt trust strongly predicts perceptions of autonomy which mediates the effects of felt trust on increased work engagement and decreased turnover intention.

Finally, conservation of resources theory (COR, Hobfoll 1989) asserts that stress increases whenever resources are lost or threatened, or whenever investment of time and effort does not yield expected resource gains. In accordance with this theory, Baer et al. (2015) found that felt trust triggers a perceived increase in workload because it is associated with additional assignments and reputation maintenance concerns, which could result in resource loss. Consequently, felt trust makes employees feel emotionally exhausted, which impacts employee job performance negatively.

In summary, the above five perspectives proposed divergent mediators for the relationship between felt trust and job performance. As shown in Fig. 1, felt trust may induce psychological responses of self-efficacy, self-esteem, perceived autonomy, felt obligation, perceived workload, and emotional exhaustion. The psychological responses determine the types of motivated behaviors and consequent outcomes.

2.3 Potential moderators in the relationship between felt trust and OCB intention

Despite having a growing number of studies on the effects of felt trust, research on its moderators remains few. One study done by Skiba and Wildman (2019) found that felt trust and trust in supervisor interacts to reduce turnover intention. Understanding context sensitivity and discovering the inherent limiting conditions are important for theory development because observations are embedded and must be understood within the context (Whetten 1989). In practice, identifying variables that act as moderators is important to develop effective intervention strategies.

Mayer et al. (1995) proposed a model of trust which regards characteristics of the trustor, the trustee, and the role of risk as the three main factors to influence trust relationship and outcomes. Accordingly, we added three potential moderators to the theoretical model summarized from prior studies on mediators of felt trust mechanism (see Fig. 1). Psychological responses of being trusted may depend on who the trustor is and how the trustee interprets the information. The extent of a trustee’s motivation depends on the perceived cost associated with the trust relationship. Therefore, the characteristics of the trustor, the characteristics of the trustee, and the perceived cost of being trusted should play the role of moderators.

Based on this model, we selected two potential moderators associated with the characteristic of the trustor (supervisor behavioral integrity and non-work relationship with supervisor) and a potential moderator related to the perceived cost in trust relationship (stress from OCB). Trust from a supervisor with behavioral integrity may enhance employee self-efficacy and self-esteem. Non-work relationship with supervisor may enhance felt obligation to reciprocate. Perceived stress of OCB may increase perceived workload and emotional exhaustion.

2.3.1 Behavioural integrity

Behavioral integrity is essential for building positive interpersonal and organizational relationships (Simons 2002; Tsai and Tsai 2022). Simons (2002) defined behavioral integrity as the perceived alignment between an actor’s words and deeds. Moreover, integrity is regarded as a virtue (Audi and Murphy 2006) or as a discrete component of good character (Palanski and Yammarino 2007). A leader characterized by behavioral integrity, a universal moral virtue, would attain followers’ respect.

Supervisor behavioral integrity may moderate the relationship between felt trust and OCB intention of subordinates through its influence on OSE and OBSE. In terms of the OBSE mechanism, when employees perceive that a significant other trusts them, e.g. a respectable supervisor, their self-esteem is enhanced, making them exert greater efforts at work beyond their prescribed roles (Lau et al. 2014). In addition, Lau et al. (2008) found that the higher the organizational rank of the trustor, the stronger the effect of felt trust. Therefore, supervisor behavioral integrity may strengthen the relationship between felt trust and OCB intention through its influence on OBSE.

In terms of the OSE mechanism, the trustee’s perceptions of the persuader’s credibility become vital when efficacy information is received through social persuasion from a trustor (Schunk 1985; Bandura 1986). Furthermore, trustees may experience a heightened sense of self-efficacy if they are persuaded by someone knowledgeable and trustworthy (Mellor et al. 2006). Supervisors are usually regarded as knowledgeable, and those with behavioral integrity are considered credible and trustworthy (Simons 2002). Therefore, supervisor behavioral integrity may augment the relationship between felt trust and OCB intention of subordinates. Based on the aforementioned points, the following hypothesis is proposed:

H1

Supervisor behavioral integrity moderates the relationship between felt trust and OCB intention of subordinates, such that the relationship is stronger when behavioral integrity is greater.

2.3.2 Non-work relationship

There are two domains of workplace relationship: work and non-work. This study focused purely on the non-work domain to obtain unequivocal results. Workplace relationships, the ongoing connections between people in a workplace, notably shape individuals’ work behaviors and outcomes (Trefalt 2013) and determine who gets asked for help and who gets help (Casciaro and Lobo 2008).

Non-work relationship with the trustor may moderate the relationship between felt trust and OCB intention. From the social exchange perspective (Cropanzano and Mitchell 2005), close relationships are the result of long-term interactions in which people exchange social resources. As individuals gather, exchange resources, and meet each other’s expectations, their exchange relationships become deepened and strengthened (Ferris et al. 2009; Ballinger and Rockmann 2017). Therefore, the felt obligation to reciprocate derived from felt trust would be stronger when trustees are in a close relationship with their trustors. Therefore, this study proposes the following:

H2

Non-work relationship with the supervisor moderates the relationship between felt trust and OCB intention of subordinates, such that the relationship is stronger when non-work relationship is closer.

2.3.3 Stress from OCB

OCB refers to employee behavior that is more discretionary, is less likely to be formally rewarded, and contributes to organizational functioning (Organ 1997). The COR theory states that individuals strive to obtain, retain, and protect the quality and quantity of their resources, and try to limit any factors that threaten the security of these resources that allow them to deal with and overcome challenges in their life (Hobfoll 1989). These valued resources may be personal, such as self-esteem, material, such as money, or conditional, such as social status. When people’s resources are threatened or lost, or when people fail to make gains after investing their resources, psychological or physical stress will occur (Hobfoll 2002). Performing OCB may engender stress because it increases resource loss with uncertain resource gain.

OCB involves the decision to invest cognitive, emotional, and physical resources in activities beyond role requirements. Internal tensions will arise and result in feelings of fatigue because these resources are limited (Bolino et al. 2015). Performing OCB involves personal costs; thus, it can make one feel drained and depleted (Bolino and Turnley 2005; Bergeron 2007), which can contribute to employee overload and stress (Bolino and Turnley 2005).

Different types of OCB consume varied resources and may exert divergent effects on the relationship between felt trust and OCB intention. Bolino et al. (2015) found that citizenship fatigue—a state of feeling worn out, tired, or on edge attributed to engaging in OCB—is negatively related to subsequent acts of OCB. The authors suggested that future studies explore the relationship between specific dimensions of OCB and OCB fatigue. Baer et al. (2015) asserted that the nature of employees’ tasks may be a critical factor in understanding the dynamics of felt trust. Much of the cost associated with trust involves additional tasks assigned to trusted employees, because when supervisors trust subordinates, they usually delegate them additional responsibilities and important decision-making authorities (Skiba and Wildman 2019). Similar to the effect of OCB fatigue, extra-role tasks may result in increased perceived workload and stress, which may reduce OCB intention. Individuals would engage in an activity if they perceive that a balance between resource loss and gain can be achieved (Hobfoll 1989). Given a fixed level of resource gains from felt trust, it can be predicted that the greater the perceived OCB stress, the lesser the OCB intention. Therefore, the following hypothesis is proposed:

H3

Perceived OCB stress moderates the relationship between felt trust and OCB intention, such that the relationship is weaker when perceived OCB stress is higher.

3 Method

The purpose of this study was to examine the moderating effects of behavioral integrity, non-work relationship, and perceived OCB stress on the relationship between felt trust and OCB intention. This involves how individuals integrate these attributes when deciding to engage in OCB. One of the most common approaches for evaluating how people integrate multiple attributes with one another is through the Information Integration Theory (IIT, Anderson 1982, 2013). IIT has been widely used as a method for multi-attribute analysis and as a theoretical basis for studies associated with attitude formation and heuristics (e.g., Adaval 2003; Kahneman 2003; Heimann et al. 2015; Yang 2019); thus, this study employed the IIT methodology for research design.

3.1 Materials

In this experiment, two OCB judgment tasks were developed, with each task consisting of 8 scenarios. The scenarios were constructed based on a three-factor design: felt trust (low vs. high) × supervisor behavioral integrity (low vs. high) × non-work relationship with supervisor (low vs. high), 2 × 2 × 2. The manipulations are described in detail in the next sections.

Judgment task of OCB The judgment task in this experiment was based on the vignette technique (Finch 1987; Poulou 2001). The vignette technique involves presenting participants with a set of hypothetical scenarios and characters. The technique can be employed to determine attitudes regardless of whether participants have had any direct experience of a scenario. Participants can be asked about what they think the character in the vignette should do and what they would do. “Should” questions focus participants’ attention on the normative dimension of situations and “would” questions focus on the pragmatic dimension (Poulou 2001). “Would” questions may involve personal experiences in the workplace in addition to the stated factors in the scenario. The purpose of this study was to probe the general attitudes about OCB intention under the consideration of the three studied factors. Therefore, participants were asked what they think the character in the vignette “should” do.

To meet the study purpose, the judgment task was designed following vignette methodology: (1) X (the character) is one of three subordinates of the supervisor; (2) the supervisor hopes someone would volunteer to take a task; (3) X is now thinking about whether he/she should take the task, considering to what extent to which the supervisor trusts X, the supervisor demonstrates behavioural integrity, and the supervisor is close with X; then, (4) participants were asked what they think the character in the vignette “should” do (In your opinion, would X feel he/she should take the task?), using a 20-point scale (1 for certainly not and 20 for certainly yes), as suggested by Anderson (1982). Participants’ rating can reflect the general attitudes about OCB intention under the consideration of the three studied factors.

This study designed two types of OCB judgment tasks to represent high and low perceived OCB stress. Stress refers to the perceived or actual threat on physical and/or psychological homeostasis of the human body (Andreou et al. 2011). Rodell and Judge (2009) assessed job stress using the individual’s perceived level of workload, time urgency, job responsibility, and job complexity. Accordingly, this study designed the two OCB tasks based on five facets, which are: (1) time and effort to input, (2) interference with work and life, (3) task uncertainty, (4) task difficulty, and (5) time pressure.

Individual initiative which involves task-related behaviours at a level beyond what is required is a type of OCB (Podsakoff et al. 2000). Below is the task representing high perceived OCB stress (Task H), which was designed to reflect individual initiative.

Y Company has just started to fulfil orders for a new product. However, the yield of the new product has not met the standard. The supervisor of the yield improvement section is under considerable pressure. X is one of the three members of the section. The supervisor considers all three members to be busy with their own work. It is not easy to assign a person for the new challenge. Therefore, the supervisor hopes someone will volunteer to solve the new problem. The person who undertakes this task needs to study difficult technical materials after work, and he/she needs to work harder in order to free up time to solve the problem as soon as possible. X is now thinking about whether he/she should take the task.

Instructing and helping a new hire is also a type of OCB (Podsakoff et al. 2000). Low perceived OCB stress task (Task L) stated below was designed based on this type of OCB.

X is one of the three staff in the company’s training and development section. A new employee comes to report today. The section supervisor hopes to find a staff willing to spare some time to help the new employee become familiar with the environment and the job, and to occasionally assist the new employee to complete unfamiliar works. The supervisor considers all three staff to be fully occupied with their respective tasks; thus, it is difficult to decide who should be assigned to help the new employee. Therefore, the supervisor hopes for someone to volunteer to help. X is now thinking about whether he/she should take the task.

To check whether the task manipulation was successful, the five facets previously mentioned were used to evaluate the degree of perceived OCB stress. One sample item is, “Time and effort needed for Task H is… (rate the item from extremely low to extremely high)”. A total of 30 participants consisting of 12 females and 18 males, aged between 23 and 56 years (M = 33.0, SD = 8.4), and with a job tenure between 1 and 30 years (M = 10.3, SD = 7.3), were recruited. They were asked to evaluate the two tasks in relation to the five facets using a 7-point scale (1 for extremely low and 7 standing for extremely high). The Cronbach’s alpha for the five items were 0.76 for Task L and 0.81 for Task H. The results of the analysis confirmed that Task H had higher perceived stress (M = 5.7, SD = 1.0) than Task L (M = 3.7, SD = 1.1; t (29) = 7.99, p < 0.001).

Felt trust There are two dimensions included for felt trust: reliance and disclosure (Gillespie 2003). Previous studies found that reliance, and not disclosure, affects employee attitudes and behaviors (Lau et al. 2014; Zheng et al. 2019); therefore, this study focused on reliance.

Lester and Brower (2003) defined felt trust in the workplace as the extent to which subordinates perceive that their supervisor considers them as trustworthy. Past studies that measured trust in organizational settings suggested that competence and responsibility are the central elements of perceived trustworthiness (Cook and Wall 1980; Butler 1991). McAllister (2017) created a six-item scale to measure cognitive-based trust, the reliability estimate (Cronbach’s alpha) of which for this measure is 0.91. Five of the six items (with Lambda of confirmatory factor analysis) are “This person approaches his/her job with professionalism and dedication (0.90)”; “Given this person’s track record, I see no reason to doubt his/her competence and preparation for the job (0.86)”; “I can rely on this person not to make my job more difficult by careless work (0.81)”; “Most people, even those who aren’t close friends of this individual, trust and respect him/her as a coworker (0.77)”; “Other work associates of mine who must interact with this individual consider him/her to be trustworthy (0.90)”. Schaubroeck et al. (2011) chose three items from McAllister’s (2017) scale to measure trust and found them to be highly reliable (0.90). Accordingly, we used competence and responsibility to manipulate felt trust. For the high (low) level of felt trust, the statement used was, “My supervisor (does not) trusts my professional ability and thinks that I am (not) responsible.”

In terms of measurement reliability, although it is not possible to test the reliability of the manipulation in this type of research design, competence and responsibility are core elements of trust and had been examined by researchers as mentioned. Regarding manipulation check, we did not check this manipulation because it is apparent that the degree of felt trust in the high-level statement is greater than that in the low-level one. Moreover, the analysis results of participants’ judgments indicated that the marginal mean of high felt trust (12.00 with standard error = 0.15) was significantly greater than that of low felt trust (8.56 with standard error = 0.18); p < 0.001, applying Bonferroni correction for multiple comparisons). Therefore, the manipulation of felt trust should be valid.

Supervisor behavioral integrity Behavioral integrity was manipulated based on the scale proposed by Simons et al. (2007), which is a well-accepted scale among other studies (e.g., Andrews et al. 2015; Way et al. 2018). The 8 items of this scale are “There is a match between my manager’s words and actions”; “My manager delivers on his/her promises”; “My manager practices what he/she preaches”; “My manager does what he/she says he/she will do”; “When my manager promises something, I can be certain that it will happen”; “My manager conducts himself/herself by the same values he/she talks about”; “My manager shows the same priorities that he/she describes”; and “If my manager says he/she is going to do something, he/she will”. Simons et al. (2007) found the eight-item scale reliable (Cronbach’s alpha = 0.87). For this study, the statement of high (low) level of behavioral integrity used was “My supervisor (does not) delivers on promises and (does not) practices what he/she preaches”.

In terms of measurement reliability, although it is not possible to test the reliability of the manipulation in this type of research design, “delivers on promises” and “practices what one preaches” represent the essential behavioral patterns of behavioral integrity defined by Simons et al. (2002; 2007); , scale which had been found to be highly reliable by researchers (e.g., Andrews et al. 2015; Way et al. 2018). Regarding manipulation check, we did not check this manipulation because it is apparent that the degree of behavioral integrity in the high-level statement is greater than that in the low-level one. Moreover, the analysis results of participants’ judgments indicated that the marginal mean of high behavioral integrity (12.94 with standard error = 0.17) was significantly greater than that of low behavioral integrity (8.62 with standard error = 0.17; p < 0.001, applying Bonferroni correction for multiple comparisons). Therefore, the manipulation of behavioral integrity should be valid.

Non-work relationship with supervisor Law et al. (2000) developed a six-item scale for measuring close personal tie (Cronbach’s alpha = 0.84), which had been used in previous studies and found to be reliable (e.g. Chen and Tjosvold 2006; Zhang et al. 2015). The six items are: “During holidays or after office hours, I would call my supervisor or visit him/her”; “My supervisor invites me to his/her home for lunch or dinner”; “On special occasions such as my supervisor’s birthday, I would definitely visit my supervisor and send him/her gifts”; “I always actively share with my supervisor about my thoughts, problems, needs and feelings”; “I care about and have a good understanding of my supervisor’s family and work conditions”; and “When there are conflicting opinions, I will definitely stand on my supervisor’s side”. The core concepts in this scale are “spend much time together after office hour” and “do something for those only in close personal relationship”. Drawing upon this scale, this study designed the high level of non-work relationship as: “My supervisor and I often get together after work. We have a close personal relationship”; and for the low level of non-work relationship as: “My supervisor and I only have a business relationship, and our personal relationship is distant.”

In terms of measurement reliability, although it is not possible to test the reliability of the manipulation in this type of research design, “spend much time together after office hour” and “in close personal relationship” are core concepts of non-work relationship and had been found to be internally consistent by researchers. Regarding manipulation check, we did not check this manipulation because it is apparent that the degree of non-work relationship with supervisor in the high-level statement is greater than that in the low-level one. Moreover, the analysis results of participants’ judgments indicated that the marginal mean of close non-work relationship with supervisor (12.23 with standard error = 0.15) was significantly greater than that of distant non-work relationship with supervisor (9.33 with standard error = 0.17; p < 0.001, applying Bonferroni correction for multiple comparisons). Therefore, the manipulation of non-work relationship with supervisor should be valid.

Consequently, an example scenario encompassing “high felt trust” × “high behavioral integrity” × “close non-work relationship” in the situation of high perceived OCB stress for this study is as follows:

Y Company has just started to fulfil orders for a new product. However, the yield of the new product has not met the standard. The supervisor of the yield improvement section is under considerable pressure. X is one of the three members of the section. The supervisor considers all three members to be busy with their own work. It is not easy to assign a person to take on the new challenge; therefore, the supervisor hopes someone will volunteer to solve the new problem. The person who undertakes this task needs to study difficult technical materials after work, and he/she needs to work harder in order to free up time to solve the problem as soon as possible. X is now thinking about whether he/she should take the task. X thinks: ‘My supervisor trusts my professional ability and thinks that I am responsible. My supervisor delivers on his/her promises and practices what he/she preaches. Also, my supervisor and I often get together after work, and we have a close personal relationship.’ In your opinion, would X feel he/she should take the task?

OCB intention As mentioned, OCB intention was measured by asking participants to rate the extent to which would X feel he/she should take the task? using a 20-point scale, suggested by Anderson (1982).

3.2 Procedure

Participants made 8 judgments for one of the two tasks. To enhance the reliability of their judgments, they were asked to make the 8 judgments twice, as suggested by Anderson (1982), resulting in 16 judgments. Responses to the two replications were averaged for further analysis.

The experiment was conducted inside a computer lab in small groups. Initially, the process, the judgment task, the three factors to be considered for judgments, two example scenarios, and the rating scale were described. Next, the participants were asked to read the instructions, take a practice session with four judgments, and then make formal judgments. The computer program presented the judgment scenarios randomly for each participant and each replication. The practice session required the participants to consider the four judgments for 60 s, allowing them to better calibrate their rating scale. In order to promote the participants’ thoughtfulness about their judgments, a 2 s time interval was provided between sentences in a scenario. The time intervals between two contiguous sentences were 3 and 1 s in the first and second replication respectively. There was also a 2 s delay between the display of the whole scenario and making the judgment. The time interval between the display of a scenario and the next was 2 s and between two replications was 1 min. The whole process took approximately 20 min.

3.3 Participants

This experiment was conducted in Taiwan. The data was collected from February 2018 to June 2019. The participants, who are all Taiwanese, were recruited by advertising the study on social media. A total of 316 participants joined the study, which is comprised of 166 women (52.5%) and 150 men (47.5%), aged between 20 and 68 years (M = 37.3, SD = 9.9) who had full-time or part-time work for at least one year, and had a job tenure of between 1 and 34 years (M = 9.5, SD = 7.7). Moreover, the participants had worked with their current direct supervisor between 1 and 30 years (M = 5.9, SD = 5.6). The sectors where they were working in at the time of this research included: service companies (43.7%), manufacturing companies (35.4%), public sectors (16.1%), and others (4.7%). Further, most of them were college graduates (61.4%), while others were high school graduates (11.1%) or postgraduate degree holders (27.5%). Those who have worked with their supervisors for more than one year should have enough experiences to understand felt trust, supervisor behavioral integrity, and non-work relationship with their supervisor.

4 Results

To test the hypotheses, a mixed ANOVA was conducted with felt trust, supervisor behavioral integrity, and non-work relationship with supervisor as within-subject factors; perceived OCB stress as a between-subject factor; and gender, age, and tenure with supervisor as covariates. The results of ANOVA indicated that the main effects of felt trust (F (1, 311) = 47.28, p < 0.001, ηp2 = 0.13), supervisor behavioral integrity (F (1, 311) = 58.03, p < 0.001, ηp2 = 0.16), non-work relationship with supervisor (F (1, 311) = 27.21, p < 0.001, ηp2 = 0.08), and perceived OCB stress (F (1, 311) = 34.48, p < 0.001, ηp2 = 0.10) were all significant. As shown in Fig. 2, the higher the level of felt trust, supervisor behavioral integrity, and non-work relationship with supervisor, the greater the OCB intention. As shown in Fig. 4, the lower the perceived OCB stress, the greater the OCB intention.

Fig. 2
figure 2

The Effects of felt trust, supervisor behavioral integrity, and non-work relationship with supervisor on OCB intention

Full size image

Regarding the moderating effects, the results showed that the interaction effect between felt trust and behavioral integrity was significant (F (1, 311) = 12.46, p < 0.001, ηp2 = 0.04), indicating that H1 is supported. As shown in Fig. 3, the slope of high behavioral integrity curve is greater than that of low behavioral integrity curve; thus, behavioral integrity moderated the relationship between felt trust and OCB intention, and that the relationship was stronger when behavioral integrity was higher.

Fig. 3
figure 3

Interactive effect of felt trust and behavioral integrity on OCB intention

Full size image

On the other hand, the results showed that the interaction effect between felt trust and non-work relationship (F (1, 311) = 0.00, p > 0.90, ηp2 = 0.00) was not significant, which indicates that H2 is not supported. That is, non-work relationship does not moderate the relationship between felt trust and OCB intention.

The results also showed that there was a significant interaction effect between felt trust and perceived OCB stress (F (1, 311) = 9.76, p < 0.001, ηp2 = 0.03). As shown in Fig. 4, the slope of low perceived OCB stress curve is greater than that of high perceived OCB stress curve. This indicates that H3 is supported; that is, the relationship between felt trust and OCB intention was stronger in the low OCB stress task than in the high OCB stress task.

Fig. 4
figure 4

Interactive effect of felt trust and perceived OCB stress on OCB intention

Full size image

In addition, the results demonstrated that the interaction effect between felt trust, behavioral integrity, and perceived OCB stress was significant (F (1, 311) = 12.93, p < 0.001, ηp2 = 0.04). It was found that the moderating effect of behavioral integrity on the relationship between felt trust and OCB intention was stronger in the high perceived OCB stress task than in the low one (see Fig. 5). This result suggests that the increased value derived from supervisor behavioral integrity is more critical to enhancing the effect of felt trust in a high-stress OCB situation than in a low one.

Fig. 5
figure 5

Interactive effect of felt trust, behavioral integrity, and perceived OCB stress on OCB intention. Note BI represents supervisor behavioral integrity

Full size image

5 Discussion

5.1 Theoretical contributions

Interpersonal trust is vital for work relationships. Mutual trust is particularly important in the workplace because it influences cooperation and reduces transaction costs in organizations (Blau 1964; Williams 2016). Extant research mainly concentrated on subordinates’ trust in their supervisors than on their feelings of being trusted by their supervisors, despite the latter being as important as the former (Lau et al. 2007; Brower et al. 2009; Lau et al. 2014; Baer 2015; Skiba and Wildman 2019; Zheng et al. 2019). Based on the trust model (Mayer et al. 1995) and the five mechanisms of felt trust in the literature, we summarized a theoretical model of felt trust, serving as the basis to choose potential moderators of the mechanism of felt trust. Consequently, we make two main contributions. One is filling the knowledge gap on the moderator in the literature of felt trust. The other is providing new evidence for the controversies of OBSE and felt obligation as mediators in the relationship between felt trust and job performance. The former sophisticates the model of felt trust effect on job performance, the latter justifies the underlying dynamics of felt trust effect proposed by previous studies. Both make value-adding contributions to theory development of felt trust. The main findings and theoretical implications are discussed in detail as below.

According to the results of this study, behavioral integrity moderated the relationship between felt trust and OCB intention. We found that the relationship is stronger when behavioral integrity is higher. This is consistent with the self-evaluative perspective (OBSE as a mediator) and the social cognitive perspective (OSE as a mediator). Supervisors are usually regarded as knowledgeable, and those with behavioral integrity are considered credible and trustworthy (Simons 2002). When employees perceive that such significant individuals trust them, their OBSE will be enhanced (Lau et al. 2014). Furthermore, trustees may experience an increased sense of self-efficacy when they receive trust messages from someone knowledgeable and trustworthy (Mellor et al. 2006).

The aforementioned findings contribute to the model of felt trust in two ways. First of all, the findings answer Lau et al. (2014) call for investigating the moderating effect of trustor characteristics. This is the first study demonstrating the importance of the trustor’s characteristics in the relationship between felt trust and OCB intention; that is, feeling trusted by supervisors with behavioral integrity moderated the relationship between felt trust and OCB intention. Given the multi-faceted nature of supervisor characteristics, it is worthy to explore which supervisor characteristics reinforce or weaken the effect of felt trust in future studies. For example, certain leader character strengths (Crossan et al. 2017; Wright et al. 2017) may reinforce the relationship between followers’ felt trust and job performance.

In addition, the present findings provide new evidence for solving the inconsistent findings between the studies of Lau et al. (2014), Zheng et al. (2019). The differences between the two studies are that the mediating effect of OBSE on the relationship between felt trust and task performance was supported by Lau et al. (2014) but not supported by Zheng et al. (2019), and that the mediating effect of OBSE on the relationship between felt trust and OCBO was not supported by Lau et al. (2014) but supported by Zheng et al. (2019). Based on the present findings, the inconsistencies may be due to the characteristics of the supervisor and the type of task. We suggest that the two moderators should be considered in future studies when investigating the relationship between felt trust and job performance.

Unexpectedly, this study found that non-work relationship does not moderate the relationship between felt trust and OCB intention. Our finding that non-work relationship facilitated OCB intention coincides with prior study results which revealed that positive aspects of the supervisor–subordinate relationship facilitate extra-role behaviors (Masterson et al. 2017; Lin and Ho 2010). Prior research also found that close personal ties between supervisors and subordinates facilitate subordinates’ felt obligation and OCB (Shih and Lin 2014; Zhang et al. 2015; Guan and Frenkel 2018). The aforementioned findings suggest that non-work relationship facilitates felt obligation to reciprocate by engaging in OCB. Therefore, the unexpected result in this study reflects that felt obligation to reciprocate may be not a mediator of the relationship between felt trust and OCB intention.

This finding provides new evidence to solve the controversy of whether felt obligation to reciprocate is a mediator in the relationship between felt trust and job performance. The mediating effect of felt obligation was supported by Skiba and Wildman (2019) but not supported by Lau et al. (2018). The present findings are consistent with the finding of Lau et al. (2018). However, there might be another possibility. According to the theoretical model proposed in this study (see Fig. 1), the characteristic of the trustee is a potential moderator of the felt trust mechanism. In line with the self-determination theory (Gagné and Deci 2005), felt trust may evoke external regulation such as the motivation to maintain reputation, introjected regulation such as felt obligation to reciprocate other’s trust, and identified regulation such as being a capable person. We suggest that future studies on the mediating role of felt obligation in the relationship between felt trust and job performance consider the moderating role of the trustee’s characteristics.

The present study also found that perceived OCB stress moderates the relationship between felt trust and OCB intention, and that the relationship is weaker when perceived OCB stress is higher. This result is consistent with the COR perspective (Hobfoll 1989) and the finding in Bolino et al. (2015) study, suggesting that high perceived OCB stress elicits high expected resource investment with uncertain resource gain, which contributes to increased perceived stress and OCB withdrawal.

The above findings contribute to the model of felt trust in two ways. First, Baer et al. (2015) discovered that feeling trusted makes employees perceive more workload, which results in greater emotional exhaustion that negatively influences job performance in the long term. The present study demonstrates that the positive effects of felt trust decline in the situation of high perceived OCB stress. This suggests that employees tend to protect themselves from being exhausted by escaping from high-stress OCB. Second, this study, combined with Baer et al. (2015), demonstrates that the nature of employees’ extra tasks might be a critical factor in understanding the effect of feeling trusted. Bergeron (2007) noted that forms of OCB vary in terms of time and effort needed; some may be time consuming and may require extensive effort, whereas others may not. Bolino et al. (2015) suggested that the relationship between OCB and citizenship fatigue may vary with the specific facet of OCB. Therefore, it is worthwhile to investigate how the nature of extra tasks resulting from supervisor trust plays a role in the relationship between felt trust and job performance in future studies.

Finally, this study found that perceived OCB stress intervenes in the moderating effect of supervisor behavioral integrity on the relationship between felt trust and OCB intention; in that, the moderating effect is stronger when perceived OCB stress is higher. This is in contrast with the finding that perceived OCB stress lessens the positive effect of felt trust on OCB intention. As shown in Fig. 5, low behavioral integrity strongly weakens the relationship between felt trust and OCB intention in the task of high perceived OCB stress. This result entails a theoretical implication that perceived OCB stress, i.e. the cost of being trusted, and supervisor behavioral integrity, i.e. the characteristic of the trustee, can interact to influence the relationship between felt trust and OCB intention. It is worthwhile to explore the interactions in future studies.

5.2 Managerial implications

OCB is an essential aspect that influences organizational functioning, and the employees’ feeling of being trusted is a way to facilitate OCB intention. The effect of felt trust depends on trustor characteristics, and it is reinforced when supervisors behave with integrity, weakened when supervisors behave otherwise—particularly under high perceived OCB stress. Further, showing trust does not mean that workers are expected to take extra work and responsibility indefinitely. Highly exhausting tasks can dampen their willingness to engage in those tasks, even when they feel that they are trusted. Moreover, when compared with establishing non-work relationship, demonstrating behavioral integrity can better facilitate the intention of OCB induced by employees’ felt trust.

Finally, a dilemma emerges in managing the relationship between managers and trusted employees. Managers are inclined to assign significant tasks to trusted employees. However, significant tasks are usually resource consuming, leading to emotional exhaustion and OCB fatigue in the long term. Therefore, eliminating extra-role tasks could be a good option to maintain the effect of felt trust (Baer et al. 2015). However, an even better way to navigate the dilemma is by showing behavioral integrity, which will not only help enhance the positive effect of felt trust, but will also eliminate the negative effect of perceived OCB stress.

5.3 Limitation and future study

Although the research design in this study was rigorous (e.g. tasks were selected from well accepted OCBs, measurements were adapted from important literature), it still has limitations. First, vignette methodology inherits the external validity strengths of survey research and the internal validity strengths of experimental methods, if it is used appropriately (Evans et al. 2015). However, a hypothetical scenario and its elicited behavior might not be sufficiently representative of real-world phenomena, and this raises concerns about the conclusions based on them (Gould 1996, Hughes and Huby 2002). Although the scenarios used in this study happen commonly in the workplace, the conclusion is not warranted by different combination of vignette sample and participant sample. Therefore, divergent combinations of vignette and sample are urged to replicate our conclusion.

Second, this study did not measure the mediators of the relationship between felt trust and OCB intention. Although prior studies have proven the effects of those mediators, such as OSE, OBSE, felt obligation, and perceived workload (e.g., Lau et al. 2014; Baer et al.2015; Skiba and Wildman 2019; Zheng et al. 2019), the theoretical inferences in this study would be more solid if those mediators were measured. Third, this study did not measure the characteristics of the trustee. According to Mayer et al. (1995), the characteristics of the trustee may also influence the effect of felt trust. For example, the influences of behavioral integrity, felt trust, and non-work relationship may involve cultural differences (e.g., Schoorman et al. 2007; Simons et al. 2015). The present findings are based on data from Taiwan, a country characterized by collectivism and high power distance (Oyserman et al. 2002; Huang and Van de Vliert 2003); therefore, future studies should recruit participants from other cultures to obtain a more generalizable result.

Finally, this study found that non-work relationship does not moderate the relationship between felt trust and OCB intention. Workplace relationships have a different set of dimensions such as instrumental work-focused exchanges versus affective and friendship exchanges (Methot et al. 2016). Future studies can examine the moderating effects of other facets of workplace relationship on the relationship between felt trust and OCB intention to confirm the current findings.