Introduction
Supervisor communication has long been regarded as a critical factor in enhancing employee in-role performance (IRP; Boies, Fiset, & Gill, Reference Boies, Fiset and Gill2015), and has great potential to develop beneficial relationships between supervisors and their employees (Brown, Paz-Aparicio, & Revilla, Reference Brown, Paz-Aparicio and Revilla2019). Indeed, feedback is one of the basic supervisor communication strategies (Brown, Kulik, & Lim, Reference Brown, Kulik and Lim2016). Previous research has examined the positive effects of supervisor feedback on employee work behavior (Peng & Lin, Reference Peng and Lin2016), most of which focuses on the impacts of traditional performance feedback toward specific tasks, addressing the discrepancies between an anticipated goal and the actual outcome (Li, Harris, Boswell, & Xie, Reference Li, Harris, Boswell and Xie2011).
Beyond performance feedback, Zhou (Reference Zhou2003) put forward the concept of supervisor developmental feedback (SDF), referring to the extent to which supervisors provide helpful or valuable information to their employees which can support learning, development and improvements on the job. Zheng, Diaz, Jing, and Chiaburu (Reference Zheng, Diaz, Jing and Chiaburu2015) and Li et al. (Reference Li, Harris, Boswell and Xie2011) provided direct evidence of the positive impacts of SDF on employee IRP. However, existing research has thus far focused on the between-person level, thereby omitting the possible fluctuations of both SDF and IRP (Breevaart, Bakker, Demerouti, & Derks, Reference Breevaart, Bakker, Demerouti and Derks2016; de Gieter, Hofmans, & Bakker, Reference de Gieter, Hofmans and Bakker2018).
It is reasonable to assume that supervisor behaviors fluctuate from day to day, and especially previous research has revealed the fluctuations in supervisors' feedbacks (Diebig, Bormann, & Rowold, Reference Diebig, Bormann and Rowold2016). When encountering relevant occasions, the supervisor may provide targeted information to employees to solve work-related problems and enhance employees' ability to learn and improve in the organizations. On another day, such information may be not necessary (Diebig, Bormann, & Rowold, Reference Diebig, Bormann and Rowold2016). Moreover, a within-person approach to supervisor behaviors allows us to examine such behavior as a proximal predictor of followers' outcomes and reduces recall bias, focusing on supervisor managerial behavior in the natural context and offering a comprehensive understanding of the supervisor managerial process (Breevaart et al., Reference Breevaart, Bakker, Demerouti and Derks2016). Knowing that each of our study variables fluctuates over time, this study will add to the SDF and performance literature by uncovering the dynamic interrelationships among our focal variables (de Gieter, Hofmans, & Bakker, Reference de Gieter, Hofmans and Bakker2018).
In this context, the first purpose of the current study is to examine the relationship between SDF and IRP from one episode to another. We draw on social exchange theory (SET; Cropanzano and Mitchell, Reference Cropanzano and Mitchell2005) to develop our framework. Workplace relationships are one of the main reasons why SET is the subject of research (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005). The general assumption of SET is that workers form distinguishable social exchange relationships with their immediate supervisors which depends upon the quality of perceived supervisor support (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005), and that such relationships will have implications for their work behavior thereafter (Park, Sturman, Vanderpool, & Chan, Reference Park, Sturman, Vanderpool and Chan2015). Previous research mainly adopts leader–member exchange (LMX; Wayne, Shore, & Liden, Reference Wayne, Shore and Liden1997) or supervisor–subordinate Guanxi (SSG) to represent the supervisor–subordinate relationship and to explain the mechanism by which resources provided by supervisors shape employee work outcomes (Newman, Schwarz, Cooper, & Sendjaya, Reference Newman, Schwarz, Cooper and Sendjaya2017).
However, in the daily supervisor–subordinate communication context, subordinates usually form their impressions about their relationship with supervisors based on the perceptions of justice and feelings of trust toward those supervisors (Curhan, Elfenbein, & Xu, Reference Curhan, Elfenbein and Xu2006), which is the dynamic foundation on which LMX or SSG is formed (Scandura & Pellegrini, Reference Scandura and Pellegrini2008; Selvarajan, Singh, & Solansky, Reference Selvarajan, Singh and Solansky2018). Yet much less is known about how leadership behavior affects the development of subordinates' immediate perception of the communication quality with the leader and that thereafter enhances work outcomes. We thus introduce the concept of rapport and argue that employees’ perceived rapport with supervisors (PRS) would mediate the effect of SDF on IRP. PRS is the immediate subjective outcome of the communication with supervisors, meaning the perception that one has been heard and treated justly, as well as positive impressions of and trust toward supervisors (Curhan, Elfenbein, & Eisenkraft, Reference Curhan, Elfenbein and Eisenkraft2010). Previous research has examined its positive effects on employees’ ongoing proactive behavior in organizations (Curhan, Elfenbein, & Kilduff, Reference Curhan, Elfenbein and Kilduff2009). Thus, we have chosen employees' PRS as a proximal mediator to outline the underlying mechanism through which day-level SDF impacts day-level employee IRP.
Another critical issue in examining the positive influences of day-level SDF is identifying the instances in which SDF is more or less effective in enhancing employee IRP through the formation of PRS. One underlying assumption of SET is that the formation of an exchange relationship depends on employees' perceptions of the value of the exchanged resources (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005). The current study takes job control – the perceived capability to exert influence on one's work environment to make it less onerous and more rewarding – as a boundary condition for the effect of SDF on PRS. Previous research has revealed the influence of job control on employees' judgments concerning the value of work-related resources (Du, Zhang, & Tekleab, Reference Du, Zhang and Tekleab2018c; Wang & Cheng, Reference Wang and Cheng2010). Consequently, individuals with higher job control are more likely to respond actively to the leader's feedback behaviors and thus generate favorable impressions about their communication process with leaders. The main purpose of high job control for employees is to attain further career growth in organizations, something directly supported by information provided through SDF (Tangirala & Ramanujam, Reference Tangirala and Ramanujam2008). Applying this principle, we propose that the strengths of the rapport mechanism depend on the level of employee job control.
In order to offer an integrative answer to the basic question, ‘How and when does SDF influence employee IRP at the episode level?’, we have developed a multilevel framework to test our research model (see Figure 1). In so doing, the present study aims to make three specific contributions to the literature on SDF and IRP.
Figure 1. Conceptual model
First, we provide direct evidence for the positive influence of SDF on IRP from one episode to another. Previous research has examined SDF trajectories over longer time periods (Zheng et al., Reference Zheng, Diaz, Jing and Chiaburu2015), while leaving short-term trends unexplored (Li et al., Reference Li, Harris, Boswell and Xie2011) and thereby omitting the possible fluctuations of SDF and IRP at the within-person level. In this vein, our research extends the scope of recent research on SDF and IRP by offering new insight into this dynamic process of social exchange.
Second, by introducing the mediating role of PRS, we help to uncover the previously implicit relationship between SDF and IRP. SET addresses the importance of relationships to the transformative influence of SDF on employee IRP (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005). Following this logic, we introduced PRS, being the subjective outcome of SDF (Curhan, Elfenbein, & Xu, Reference Curhan, Elfenbein and Xu2006), as a temporal mediator which can highlight the underlying connection between SDF and IRP in the daily supervisor–subordinate communication context.
Third, by integrating the research on job control into our approach, the current study helps to explain when SDF influences IRP through PRS. Employees will only engage in high-quality relationships with their supervisors and adapt their work behavior when they are in receipt of desirable resources (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005). In accordance with this implicit assumption of SET, we introduce job control as a moderator due to its potential influence in shaping employees' judgments of the value of SDF. In doing so, we explore the boundary condition under which SDF is more or less beneficial for the formation of rapport between supervisors and their employees.
Hypothesis Development
Supervisor developmental feedback and perceived rapport with supervisors
Perceived rapport is the employees' subjective outcome of communication with their supervisors (Curhan, Elfenbein, & Xu, Reference Curhan, Elfenbein and Xu2006). It contains two subdimensions. One involves positive impressions, trust, and a solid foundation for working and communicating with their supervisors in future (Curhan, Elfenbein, & Eisenkraft, Reference Curhan, Elfenbein and Eisenkraft2010). The other is the perception that the employee has been heard and treated justly, and that the communication process with supervisors is efficient (Curhan, Elfenbein, & Eisenkraft, Reference Curhan, Elfenbein and Eisenkraft2010). SDF exerts an influence on employee PRS by enhancing employees' trust in their supervisors and the perception that those supervisors act justly.
According to the definition of SDF provided by Zhou (Reference Zhou2003), we can infer that it has three unique characteristics. First, it provides useful and helpful information for employees. Second, SDF is future-oriented, focusing on employees' development in organizations and on-the-job improvements. Third, it is well-suited to creating an agreeable team climate free from pressure and the fear of making mistakes (Li et al., Reference Li, Harris, Boswell and Xie2011; Zhou, Reference Zhou2003).
We speculate that SDF will improve PRS through engendering employee's trust in the leader. Specifically, employees develop positive impressions of their supervisors when they attribute the supervisors' behavior to selfless and sincere motives, which results in affective trust in those supervisors (Chen, Eberly, Chiang, Farh, & Cheng, Reference Chen, Eberly, Chiang, Farh and Cheng2014). As we argued above, SDF is informational in nature, aiming to foster learning and improvement in employees (Li et al., Reference Li, Harris, Boswell and Xie2011). After receiving SDF, it is the supervisors' genuine care and attention that will nurture the employees' sense of gratitude and indebtedness, thereby ensuring a reciprocal interpersonal relationship and further strengthening the affective bonds of trust (Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014; Chen, Chen, & Xin, Reference Chen, Chen and Xin2004). Furthermore, developmental feedback is an intimate process where personal information is shared, and therefore, engaging in such personal and intimate conversations creates rapport because employees are being vulnerable and honest with each other. This therefore builds a communal, rather than an exchange relationship (Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014).
On the other hand, SDF can also enhance PRS through the employee's perception of being treated justly and efficiently. The provision of feedback to employees is one of the important processes through which supervisors can influence employees' perceptions of justice. Unlike traditional feedback, SDF actively directs employees toward making improvements on the job (Zhou, Reference Zhou2003) and does not come with a rigid set of requirements designed to overcome discrepancies between work outcomes and anticipated goals. Instead, SDF reveals supervisors' sense of care for their employees' development and it actively devotes resources to reward employees' appropriate behavior (Zhou, Reference Zhou2003). Following the logic of SET, these benefits are likely to be perceived by employees as fairly treated and confidence in their leaders. As a result, employees are more likely to experience high-quality exchange relationship with their leaders. Also, developmental feedback is a form of socioemotional resource, rather than an economic one, which therefore forms a social exchange of rapport building. Thus, when receiving SDF, employees are inclined to improve PRS by forming a perception of interpersonal justice in their relationship.
On this basis, we can infer that SDF enhances employees' PRS by promoting employees' trust in supervisors and perceptions of justice. We therefore hypothesize the following:
Hypothesis 1 Day-level SDF positively relates to day-level PRS.
Perceived rapport with supervisors and in-role performance
PRS involves two issues regarding participants' feelings about the communication process and the relationship (Curhan, Elfenbein, & Xu, Reference Curhan, Elfenbein and Xu2006). In the supervisor–subordinate interaction, the communication process is affected by the sense of being treated justly by supervisors, while the relationship itself is affected by the sense of trust in supervisors.
Previous research has revealed the positive relationship between justice perception and IRP (Edwards & Kudret, Reference Edwards and Kudret2017; He, Fehr, Yam, Long, & Hao, Reference He, Fehr, Yam, Long and Hao2017). The feeling of interpersonal justice is associated with more positive emotions and fewer negative emotions (Zapata-Phelan, Colquitt, Scott, & Livingston, Reference Zapata-Phelan, Colquitt, Scott and Livingston2009). From the broaden and build perspective (Fredrickson, Reference Fredrickson2001), positive emotions broaden individual thinking, attention, and behavioral repertoires. This helps employees to develop their social and psychological resources (Fredrickson, Tugade, Waugh, & Larkin, Reference Fredrickson, Tugade, Waugh and Larkin2003) and enables them to enhance their task performance (Meneghel, Salanova, & Martínez, Reference Meneghel, Salanova and Martínez2016).
Trust in supervisors has also been shown to induce high IRP (de Jong, Dirks, & Gillespie, Reference de Jong, Dirks and Gillespie2016; Mo & Shi, Reference Mo and Shi2017). When employees maintain a trust-based relationship with their supervisors, they often have a stronger sense of psychological identification both with supervisors and with the organization (Edú-Valsania, Moriano, & Molero, Reference Edú-Valsania, Moriano and Molero2016). Thus, their intrinsic motivations are triggered to exert greater effort and achieve better performance levels, and the previous meta-analysis has shown that trust in supervisors is a positive indicator of employees' performance outcomes (de Gieter, Hofmans, & Bakker, Reference de Gieter, Hofmans and Bakker2016). In accordance with these perspectives, we hypothesize the following:
Hypothesis 2 Day-level PRS positively relates to day-level IRP.
Within the framework of SET, daily SDF offers helpful and useful information to employees, nurturing their PRS, including increasing justice perception and trust in supervisors. In turn, high PRS could foster employees' IRP and thereby enhance the reciprocal quality of their relationship with supervisors. Thus, we hypothesize the following:
Hypothesis 3 Day-level PRS mediates the positive relationship between day-level SDF and day-level IRP.
Moderating role of job control
SET assumes that social exchange relationships evolve when employees are taken care of by employers, which thereby engenders beneficial consequences (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005). Employees will have different preferences regarding the resources exchanged with their supervisors (Halbesleben, Neveu, Paustian-Underdahl, & Westman, Reference Halbesleben, Neveu, Paustian-Underdahl and Westman2014). In order to better understand the formation of supervisor–subordinate relationships in the context of day-to-day workplace relations, this study adopts employee job control as an indicator of these personal preferences for resources. In this conceptual model, job control will thereby be used to examine the boundary condition under which SDF is more or less beneficial for the formation of exchange relationships.
Job control is considered to be a predictor of an individual's beliefs about the causes of important work outcomes, the number of impacts that the individual has over work-related events, and the resources available to the individual to achieve their goals in the workplace (Lee, Ashford, & Bobko, Reference Lee, Ashford and Bobko1990). Tangirala and Ramanujam (Reference Tangirala and Ramanujam2008) suggested that high job control enables employees not only to personally initiate action at work, but also provides a clear insight into the connection between such action and important organizational outcomes. They argued that employees with high job control are driven by an expectancy-based perspective which leads them to actively attend to work-related affairs in order to enhance their interpersonal relationship in the workplace (Tangirala & Ramanujam, Reference Tangirala and Ramanujam2008). As mentioned earlier, SDF provides employees with general rather than task-specific information, focusing on employees' improvement and development within an organization (Li et al., Reference Li, Harris, Boswell and Xie2011; Zhou, Reference Zhou2003). Individuals with high job control perceive themselves of being capable of utilizing these developmental feedbacks, thus increasing employees' satisfaction and resulting in positive evaluation of their communication quality with leaders. Therefore, at high levels of job control, we expect that employees are more likely to experience favorable effects of SDF and thereby develop high-quality rapport with leaders. Previous research also identified a positive relationship between the usefulness of the information and the extent to which employees are satisfied with their immediate communication with the supervisors (Chan & Lai, Reference Chan and Lai2017), which then facilitates the formation of rapport between supervisors and employees (Thomas, Zolin, & Hartman, Reference Thomas, Zolin and Hartman2009). Thus, SDF will have a stronger positive effect on PRS for employees with high job control.
In contrast, employees with low job control are driven by a ‘dissatisfaction-based perspective’ and are motivated to increase their sense of job control (Tangirala & Ramanujam, Reference Tangirala and Ramanujam2008). To that end, they prefer feedback which focuses on past outcomes and which provides them with specific information to help them decrease the discrepancies between work outcomes and intended goals (Earley, Northcraft, Lee, & Lituchy, Reference Earley, Northcraft, Lee and Lituchy1990). Even if supervisors communicate with employees in a sincere and respectful way, we argue that the positive effects of SDF on PRS will be undermined or negated for employees with low job control. Due to the irrelevant information contained in SDF, employees tend to recognize their supervisors as not understanding their needs, which do harms to the establishment of beneficial social exchange between supervisors and employees (Colquitt, Lepine, Piccolo, Zapata, & Rich, Reference Colquitt, Lepine, Piccolo, Zapata and Rich2012; Meyer, Ohana, & Stinglhamber, Reference Meyer, Ohana and Stinglhamber2018). Moreover, mismatched information provisions will undermine the effectiveness of communication, which will also decrease the quality of rapport (Stuhlmacher & Champagne, Reference Stuhlmacher and Champagne2000). Integrating the above arguments, we put forward the following hypothesis:
Hypothesis 4 Job control moderates the positive relationship between day-level SDF and day-level PRS, such that the relationship is positive for high job control employees and nonsignificant for low job control employees.
In accordance with SET, SDF enhances employees' PRS by providing them with helpful and useful information to foster their development and improvement within organizations. Further, increasing PRS could in turn lead employees to enhance the performance as a means to ‘repay’ their supervisors. However, for high job control employees, they are likely to achieve a higher PRS and, for them, the positive indirect effect of SDF on IRP will be stronger. By contrast, for low job control employees, SDF will not help them to enhance their job control over specific tasks and the positive indirect effect of SDF on IRP through PRS will be negated.
Hypothesis 5 Job control moderates the indirect effect of day-level SDF on day-level IRP through PRS. Specifically, PRS mediates the SDF–IRP relationship when job control is high, but not when it is low.
Methods
Participants and procedure
We developed our conceptual model within the daily supervisor–subordinate communication context and applied the experience sampling method (ESM) to collect daily data. We chose 100 alumni of a Chinese university in Beijing randomly from those who had updated their contact information within the past 2 years. We then contacted the 100 samples through email, social apps, and phone to invite them to take part in our survey. We explained the purpose of the research and the proposed procedure. Sixty-nine participants confirmed their participation. We formed a research group on WeChat (a widely-used messaging app in China) and sent them weblinks to the questionnaires. On Sunday, we sent an initial questionnaire, which ascertained their code, gender, education, tenure, and job control. Thereafter, from Monday to Friday, participants received two questionnaires at 10:00 (SDF) and 17:00 (PRS and IRP), respectively. Participants were requested to finish the questionnaires within 30 min. As soon as the participants finished their surveys over the course of 5 days, they would receive 25 RMB (≈3.55 USD) as compensation for their participation.
A total of 290 matched surveys nested in 58 valid questionnaires were returned for analysis. The effective response rate was 84.1%. Study participants worked in a variety of industries in mainland China, including Internet companies, financial services, and manufacturing, ensuring the representativeness of the sample. Male participants represented 49.3%; 88.1% of the respondents held a bachelor's degrees or higher; the mean duration of work experience was 8.448 years (sd = 7.467).
Measures
Considering the requirements of ESM, we only adopted several items from the original scales (Donald, Atkins, Parker, Christie, & Ryan, Reference Donald, Atkins, Parker, Christie and Ryan2016). To ensure the validity of our selected items, we asked two professors of organizational behavior, three MBA students (one human resource manager and two grassroots leaders) to ask for their suggestions. All the items in the current study adopted a 5-point Likert scale and were employee self-reported, with 1 = ‘strongly disagree’ and 5 = ‘totally agree.’ In order to ensure accurate translation into Chinese, all items were subject to the double translation procedure (Brislin, Reference Brislin1970).
Job control was measured using the scale developed by Lee, Ashford, and Bobko (Reference Lee, Ashford and Bobko1990). The scale comprised of three items; a sample item was: ‘I have enough power in this organization to control events that may affect my job.’ The Cronbach's α of this scale was .735.
SDF was assessed using the three items developed by Zhou (Reference Zhou2003). A sample item was: ‘Today, while giving me feedback, my supervisor focused on helping me to learn and improve.’ The Cronbach's α of this scale was .939.
To measure PRS, we adapted four items from the subjective value inventory developed by Curhan, Elfenbein, and Xu (Reference Curhan, Elfenbein and Xu2006). The items were: ‘Do you feel your supervisor listened to your concerns today?’; ‘Did your supervisor consider your wishes, opinions, or needs today?’; ‘Did the communication today make you trust your supervisor?’; and ‘How satisfied are you with your relationship with your supervisor as a result of the communication today?’. The Cronbach's α of this scale was .937.
IRP was assessed using the four items developed by Williams and Anderson (Reference Williams and Anderson1991). An example item was: ‘Today I adequately completed all of my assigned duties.’ The Cronbach's α of this scale was .928. The current study adopted self-reported IRP, which has often been used in ESM (Parke, Weinhardt, Brodsky, Tangirala, & DeVoe, Reference Parke, Weinhardt, Brodsky, Tangirala and DeVoe2018).
Control variables were included: gender (0 = male, 1 = female); education (1 = college or below, 2 = bachelor, 3 = master or above); and work experience (years), as we considered these variables might potentially affect our research results (Ding & Chang, Reference Ding and Chang2019; Duan, Bao, Huang, & Brinsfield, Reference Duan, Bao, Huang and Brinsfield2018; Li et al., Reference Li, Harris, Boswell and Xie2011). Because age is always correlated with job experience (Duan, Li, Xu, & Wu, Reference Duan, Li, Xu and Wu2017; Holland, Cooper, & Sheehan, Reference Holland, Cooper and Sheehan2017), to control the multicollinearity, we only adopted work experience in the equation model.
Results
Confirmatory factor analysis
The current study adopted a multilevel confirmatory factor analysis (MCFA) to test the validity of the conceptual four-factor construct. Results showed the hypothesized four-factor model fitted better for the data compared to other measurement constructs (χ2[41] = 118.51, RMSEA = .08, SRMRWithin = .04, CFI = .97). The result of the MCFA is shown in Table 1, and reveals the range of discrimination validity in different factor constructs.
Table 1. Results of multilevel confirmed factor analysis
Note. N = 290.
JC, job control; PRS, perceived rapport with supervisors; SDF, supervisor developmental feedback; IRP, in-role performance.
*p < .05; **p < .01.
Regression results
Table 2 provides the descriptive statistics, intra-correlations, and correlations among all study variables. IRP is positively correlated with SDF (r = .20, p < .01) and PRS (r = .20, p < .01). SDF is positively related to PRS (r = .44, p < .01).
Table 2. Means, standard deviations, and correlations
*p < .05; **p < .01.
The first stage of statistical analysis investigated systematic within- and between-person variance for the daily variables. The ICC(1) for SDF, PRS, and IRP are .79, .70, and .59, justifying the use of multilevel analysis.
The current study used HLM software (Version 6.08) to employ a multilevel moderated mediation model to test the hypotheses. Before we ran hierarchical linear regression models, within-person variables (SDF, IRP, and PRS) were all group centered, while a part of between-person variables (job experience and job control) was grand centered.
Table 3 shows that the result in Model 1 indicated that SDF positively affected PRS (γ = .21, p < .05). These results support Hypothesis 1. When both PRS and SDF were simultaneously entered into the equation (Model 3), PRS significantly affected IRP (γ = .22, p < .01), whereas the effect of SDF on IRP was not significant (γ = .11, n.s.). Further, we tested the indirect effect of SDF on IRP through PRS. Using the Monte Carlo method, the result indicated that the indirect effect was significant (CI = [.01, .09]). This outcome supported Hypotheses 2 and 3.
Table 3. Results of the hieratical linear model
Note. N = 290 observations nested within 58 individuals.
Pseudo R 2 is calculated on the basis of the formula from Snijders and Bosker (Reference Snijders and Bosker1999).
JC, job control; PRS, perceived rapport with supervisors; SDF, supervisor developmental feedback; IRP, in-role performance.
*p < .05; **p < .01.
In Model 1, the interactive item of SDF with job control influenced PRS (γ = .22, p < .05). This result therefore supported Hypothesis 4. Furthermore, to directly elaborate the moderating role of job control, we developed Figure 2 based on the method proposed by Aiken and West (Reference Aiken and West1991). Figure 2 indicates that when job control was high, the effect of SDF on PRS was positive, but this relationship was not significant under the condition of low job control. This revealed the amplifying effect of job control on this relationship.
Figure 2. Moderated effect of positive affect
Table 4 showed the indirect effects and their Monte Carlo confidence intervals at different combinations of higher and lower values of the moderators, calculated using R. SDF had a positive indirect effect on IRP through PRS when job control was high (low job control CI = [−.01, .04]; high job control CI = [.03, .12]). Moreover, the positive indirect effect only emerged when there was high job control rather than low (difference = .060, CI = [.01, .12]). This outcome supports Hypothesis 5.
Table 4. Moderated mediation model test
Discussion
Managerial coaching is a process of feedback provision, behavioral modeling, and goal setting with subordinates to improve their performance and facilitate their growth in organizations (Dahling, Taylor, Chau, & Dwight, Reference Dahling, Taylor, Chau and Dwight2016). Supervisor feedback is a key ingredient of managerial coaching which is positively associated with organizational effectiveness and employees' career growth (Beattie, Kim, Hagen, Egan, Ellinger, & Hamlin, Reference Beattie, Kim, Hagen, Egan, Ellinger and Hamlin2014; Huang & Hsieh, Reference Huang and Hsieh2015; Yuan, Wang, Huang, & Zhu, Reference Yuan, Wang, Huang and Zhu2019). Adopting ESM, our research collected data for five consecutive days and examined the positive effect of day-level SDF on day-level IRP. Further, the current study found that day-level PRS mediated this relationship and that levels of job control moderated the indirect positive effect. When job control is high, day-level SDF enhances day-level PRS and thereafter promotes day-level IRP. Under the condition of low job control, the positive effect of day-level SDF on day-level PRS disappears and therefore cannot facilitate day-level IRP.
Theoretical implications
The current study has three potential theoretical contributions to existing IRP research. First, we provided direct episodic evidence of the positive effect of day-level SDF on IRP. Zheng et al. (Reference Zheng, Diaz, Jing and Chiaburu2015) and Li et al. (Reference Li, Harris, Boswell and Xie2011) have examined the positive effect of SDF on IRP, adopting between-person variance and omitting the possible fluctuations of SDF and IRP at the within-person level. Li et al. (Reference Li, Harris, Boswell and Xie2011) emphasized the importance of future research of examining how employees responded to changing SDF over time in order to gain further insights into the implicit relationship between feedback and performance. To address this issue directly, the current study adopted ESM and used data collected over five consecutive days. In so doing, we found that SDF could enhance the reciprocal relationships between supervisors and employees, engendering beneficial consequences – namely, higher employee IRP. Our research revealed the dynamic social exchange process and thereby offers a new perspective from which to uncover the relationship between feedback and IRP at the episode level.
Second, we discovered the underlying mechanism by which SDF exerts influence on IRP by examining the mediated role of PRS. Social exchange is a process of relationship formation (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005). Given the specific daily communication context, we adopted PRS rather than LMX to represent employees' perceptions of their relationship with their immediate supervisors. PRS is the subjective outcome of communication with supervisors and includes two essential elements: trust in supervisors and perceptions of justice in the communication between the two parties (Curhan, Elfenbein, & Xu, Reference Curhan, Elfenbein and Xu2006). This is enhanced by the informational and future-oriented nature of SDF (Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014; Lee & Jensen, Reference Lee and Jensen2014; Li et al., Reference Li, Harris, Boswell and Xie2011; Zhou, Reference Zhou2003). Because it also nurtures positive emotions and increases their intrinsic motivation (Selvarajan, Singh, & Solansky, Reference Selvarajan, Singh and Solansky2018), this also plays an important role in increasing IRP (Gabriel, Cheshin, Moran, & van Kleef, Reference Gabriel, Cheshin, Moran and van Kleef2016; Van Iddekinge, Aguinis, Mackey, & DeOrtentiis, Reference Van Iddekinge, Aguinis, Mackey and DeOrtentiis2018). The current study deepens our knowledge of how day-level SDF affects day-level IRP, and it thereby offers a specific perspective through which to develop this relationship in the daily supervisor–subordinate communication context.
Third, by examining the moderating role of job control, we identified the influence of employees' preferences regarding SDF on the relationship between SDF and IRP. Although SET (Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005) indicates that day-level SDF should engender high-level IRP by enhancing PRS, the existing literature on employment relationships suggests that the nature of such associations may vary depending on employees' preferences for the specific resources they exchange with their supervisors (Bartol, Liu, Zeng, & Wu, Reference Bartol, Liu, Zeng and Wu2009). Our research introduced job control as a means by which to reveal employees' preferences for specific resources. High job control employees already have a high level of autonomy; for them, SDF provides general principles concerning how to behave in order to achieve career growth within their organizations, catering to their desire to enhance their influence in the workplace, rather than on specific roles or tasks (Tangirala & Ramanujam, Reference Tangirala and Ramanujam2008). In contrast, low job control employees prefer specific performance feedback aimed at developing their mastery of specific skills and improving performance in their current roles (Tangirala & Ramanujam, Reference Tangirala and Ramanujam2008). Thus, given these different feedback preferences, the positive indirect effect of SDF on IRP through PRS only occurs for high job control employees. Our research therefore highlights the importance of taking employees' preferences into consideration when researching this dynamic social exchange process.
Practical implications
This study yields several important management implications to consider. Supervisor coaching is important in shaping employee psychological state and work behavior. Also, it is a key role in facilitating employee career growth (Huang & Hsieh, Reference Huang and Hsieh2015). Feedback is one of the basic strategies of supervisor coaching (Dahling et al., Reference Dahling, Taylor, Chau and Dwight2016). Consistent with Feedback Intervention Theory (FIT; Kluger and DeNisi, Reference Kluger and DeNisi1996), the effective managerial coaching behaviors, such as developmental feedback, can resolve ambiguity to clarify for subordinates why performance-related tasks are important and how best to complete them, resulting in an improved understanding of one's responsibilities and objectives (Dahling et al., Reference Dahling, Taylor, Chau and Dwight2016). Our research verifies the benefits of SDF as a possible means by which to enhance IRP. However, the characteristics of SDF are themselves very important: messy information can undermine employees' motivation to devote themselves to their current tasks, so managers should therefore provide clear targets if employees are to improve and to achieve career growth. Feedback messages are better to focus on providing objective information about performance rather than evaluative information (Booth-Butterfield, Reference Booth-Butterfield1989; King, Reference King2016). It is useful to direct the locus of attention toward the task and the goal to ensure the success of feedback interventions (Kluger & DeNisi, Reference Kluger and DeNisi1996, Reference Kluger and DeNisi1998).
The other managerial implication is the importance of taking employees' level of job control into consideration. SDF is not always an effective means for supervisors to build reciprocal relationships with their employees and is dependent upon the particular employee's level of job control. For high job control employees, SDF caters to their desires to achieve further career growth in an organization and thereby enhances supervisors' managerial effectiveness. However, for low job control employees, managers should provide feedback centered on specific information which can help those employees evolve from their passive positions within the organization and enhance their autonomy and ability within their current jobs. Based on the FIT, politeness strategies can also be utilized to reduce the threats to ego and self-image for low job control subordinates, and the application of the face threat mitigation tactics has the impact on the individual perceptions of both the quality and usefulness of feedback (Trees, Kerssen-Griep, & Hess, Reference Trees, Kerssen-Griep and Hess2009).
Limitations and future research
Our research has several limitations. First, we cannot establish firm causal relationships between our focal variables. We tried to infer the causal effect of day-level SDF on day-level PRS and IRP by adopting multi-wave ESM for five consecutive days. However, we still need to adopt an experimental design or a cross-lagged design to rule out interference factors and develop firm causal results. Furthermore, the daily survey of the present study lasts for five consecutive days, which may underscore the fluctuations of our within-person variables (Gabriel, Koopman, Rosen, & Johnson, Reference Gabriel, Koopman, Rosen and Johnson2018). Although there are some studies adopting 5-day surveys (Du, Derks, & Bakker, Reference Du, Derks and Bakker2018a; Du, Derks, Bakker, & Lu, Reference Du, Derks, Bakker and Lu2018b; Ouweneel, Le Blanc, Schaufeli, & van Wijhe, Reference Ouweneel, Le Blanc, Schaufeli and van Wijhe2012; Xanthopoulou, Bakker, Demerouti, & Schaufeli, Reference Xanthopoulou, Bakker, Demerouti and Schaufeli2012), it would be better for future research to adopt a longer period to explore the episodic relationships between focal variables (Beal, Trougakos, Weiss, & Dalal, Reference Beal, Trougakos, Weiss and Dalal2013; Gabriel, Frantz, Levy, & Hilliard, Reference Gabriel, Frantz, Levy and Hilliard2014).
Second, all the focal variables in our research were collected through self-reported questionnaires, which leads to potential common method variance (CMV; Fuller, Simmering, Atinc, Atinc, & Babin, Reference Fuller, Simmering, Atinc, Atinc and Babin2016). Although we highlighted the reasons why we chose to use self-reported evaluations, it is still important to collect multi-source data to rule out CMV and thereby test the robustness of our results. For instance, we can use supervisor-rated or colleague-rated IRP as an outcome. Furthermore, the investigation procedure and research data will have a higher validity with the participation of supervisors and colleagues, because it would decrease the bias induced by our convenience sampling and self-reported questionnaires. A multi-source design may be adopted by future research to increase the validity of the research design.
Third, our research is performed in the Chinese culture context. Previous research has revealed the formation of supervisor–subordinate relationship varies from culture to culture (Chong et al., Reference Chong, Peng, Fu, Richards, Muethel, Caldas and Shang2015). Scholars suggest to use Chinese Guanxi rather than LMX to represent the leader–member relationship (Chen, Friedman, Yu, Fang, & Lu, Reference Chen, Friedman, Yu, Fang and Lu2009). The present study adopts PRS to indicate the immediate subjective outcome of the communication with supervisors, which are the episodic foundation of both Guanxi and LMX. We further collected data from 168 Chinese full-time workers at the between-person level. Among them, 56% were males; 82.1% held bachelor certificates or above; the mean age was 32.06 ± 9.3 years and mean job experience was 8.87 ± 8.78 years. We found high and significant relationships among Guanxi, PRS, and LMX, ranging from .80 to .87. However, there are also differences among the three concepts. Future research may collect data of these three different concepts, and further compare their mediating effects to offer a comprehensive sight into the relational path of SET.
Fourth, we explore the relational path through which SDF impacts IRP, which may ignore other potential mechanisms. We base our research within the framework of SET, which addresses the importance of relationship in the social exchange process. Lawler (Reference Lawler2001; Lawler and Thye Reference Lawler and Thye1999) pointed out affects also play an important role in SET. Recently, both literature reviews and empirical studies begin to explore the theoretical position of effects in SET and provide evidence of the importance of effects in the social exchange process (Porath, Gerbasi, & Schorch, Reference Porath, Gerbasi and Schorch2015; Troth, Lawrence, Jordan, & Ashkanasy, Reference Troth, Lawrence, Jordan and Ashkanasy2018; Tse, Troth, Ashkanasy, & Collins, Reference Tse, Troth, Ashkanasy and Collins2018). Future research may examine both the relational and affective paths in future research and further explore the theoretical positions of these two paths.
Last but not least, we adopt SDF as an antecedent of IRP and PRS due to its developmental nature. Besides developmental feedback, performance feedback is the other stream of managerial feedback in organizations, and Steelman, Levy, and Snell (Reference Steelman, Levy and Snell2004) developed a comprehensive scale of feedback environment to deeply reflect the contextual aspects surrounding the transmission on job performance feedback on the recurrent or daily basis. Prior research also indicated its influences on enhancing employee empowerment (Gabriel et al., Reference Gabriel, Frantz, Levy and Hilliard2014). Future research could adopt feedback environment, especially the supervisor feedback environment, into consideration to compare its similar and different impacts on the social exchange process with SDF.
Acknowledgements
Our research is supported by the National Natural Science Foundation of China (71772052, 71802045, 71701083).
Conflict of Interest
The authors have declared that no conflict of interest exists.
