The influence of voice quality,voice content,and managers’ mood on their evaluations of voice: An experimental investigation

The role of voice quality in managers’ evaluations of voice

Employees who express voice are generally rewarded with higher performance and promotion evaluations by managers (Chamberlin et al., 2017; Zhang et al., 2022). Managers do so to acknowledge and reciprocate employees’ efforts to improve organizational functioning and their own position as a valuable source of information and ideas in the organization (Blau, 1964). However, in certain situations, employees may also face negative consequences for speaking up, such as receiving lower promotion ratings (Seibert et al., 2001).The distribution of power within organizations offers an explanation for these varied outcomes (Georgesen and Harris, 1998). Due to their hierarchical position, managers possess authority and power that enable them to influence others and have control over valued resources, information, and decisions, including the implementation of employee voice (Anderson and Brion, 2014; Glauser, 1984; Kipnis, 1972; Magee and Galinsky, 2008). Once managers acquire power, they strive to maintain it, as it can lead to an increase in their social status and influence (Magee and Galinsky, 2008). As a result, managers are particularly attentive to situations that help them preserve their hierarchical position and achieve their personal goals (DeCelles et al., 2012; Keltner et al., 2003). Simultaneously, they seek to avoid situations they perceive as threats to their power and status (Anderson and Galinsky, 2006). In order to maintain their power, managers often, for instance, justify the status quo when organizational changes involve adjustments to the hierarchical structure (Jost et al., 2004; Jost and Banaji, 1994). Regarding employee voice, managers evaluate the risk associated with the potential lack of success of new ideas, as well as the implications of their implementation for themselves and their position within the organization (Lonergan et al., 2004; Urbach and Fay, 2021). If managers perceive an idea as a threat to their own position, they tend to act in a self-interested manner, such as by rejecting the idea (Urbach and Fay, 2018). However, when an idea contributes to organizational effectiveness and aligns with managers’ own goals, such as ensuring stable processes, managers are more likely to reciprocate by providing a positive evaluation (Li et al., 2023).

In the case of high-quality voice, managers might feel a particularly strong obligation to reciprocate due to the amount of time, effort, and potential risk that employees willingly invest in developing and communicating rational, feasible, organization-focused, and novel proposals (Brykman and Raver, 2021). Furthermore, managers are likely to attribute a high probability of success, and therefore, the achievement of personal goals, to ideas that demonstrate high quality. Therefore, the first hypothesis is derived:

The role of voice content in managers’ evaluations of voice

This section outlines the influence of high-quality voice’s content—meaning whether it is promotive or prohibitive—on managers’ evaluations.

Voice quality is composed of four dimensions: rationality, feasibility, organizational-focus, and novelty. A glance at the literature suggests that rationality—that is, the role of objective facts and logical arguments (Brykman and Raver, 2021)—may differ in managers’ evaluations of voice depending on the message’s content (promotive vs prohibitive). Prohibitive voice refers to problems that are or could be harmful to the organization (Liang et al., 2012), so solving the issue addressed can be urgent to avoid process errors, accidents, and organizational losses. Based on the previous reasoning, managers may have a heightened interest in resolving identified problems promptly to prevent any negative impact on their reputation and power within the organization (Burris, 2012; Morrison, 2023). Other members of the organization may interpret prohibitive voice as an indication of managerial mistakes (Isaakyan et al., 2021). In contrast, promotive voice usually presents issues that are deferrable, as they often refer to improving the status quo, usually in the long term. Research demonstrates that managers prefer that, whatever the voice content, it be expressed with sufficient time to react (Whiting et al., 2012).

Emotions like anger, sadness, and enthusiasm, refer to the dimension of voice quality that is opposite to rationality (Carr, 2001). Such emotions may be shown when employees talk about a problem by which they are personally affected (Lin and Johnson, 2015), so they may shift managers’ attention from the problem to the employee. Managers’ thinking about how to deal with prohibitive voice may be interrupted by the employees’ emotional state and the need to, for instance, calm and comfort the employee. Managers tend to evaluate voice expressed in a negative emotional state as of less value because they are likely to interpret these proposals as complaints or criticism instead of constructive input (Grant, 2013). In addition, individuals who hold positions of power tend to exhibit lower levels of empathetic concern (Woltin et al., 2011) and engage less in perspective-taking (Galinsky et al., 2006). These tendencies may further contribute to the negative attitudes toward prohibitive voice. With promotive voice, whose endorsement tends to be less urgent but resource-intensive (Liang et al., 2012), managers may tolerate the positive emotions that employees tend to show when talking about ways to improve the status quo (Lin and Johnson, 2015) that may help managers to achieve their own goals. Therefore, rationality may be more important for managers’ evaluations of prohibitive than for promotive voice.

Brykman and Raver (2021) see the next dimension of voice quality, feasibility, as relating to whether the proposal can be implemented, considering the effort that would be required. Burris et al. (2017) demonstrate that managers are more likely to endorse ideas when the ideas require limited resources (i.e. financing, personnel, time, effort, and interdependencies). Managers might not be concerned about the resources necessary to solving the problem, as “the ends may justify the means”, particularly when it comes to prohibitive voice, where the aim is to resolve a problem before it becomes widely known within the organization and jeopardizes managers’ image as a competent leader (Isaakyan et al., 2021). In contrast, the literature on idea management demonstrates that feasibility is usually one of the main selection criteria for choosing ideas for implementation (Gerlach and Brem, 2017) due to the high risk of failure (e.g. Dean et al., 2006; Rietzschel et al., 2019). In many studies, feasibility was rated as the most important factor in participants’ selection of ideas, compared to, for example, the originality of an idea (e.g. Rietzschel et al., 2010). Therefore, managers may usually pay more attention to the feasibility of promotive voice.

The next dimension of voice quality is organizational focus, that is, whether voice is in line with the organization’s goals and values (Brykman and Raver, 2021). Dutton et al. (2001) demonstrate that issue-selling, a subset of voice (Morrison, 2011), is more likely to be successful when it supports the organization’s strategy. Furthermore, according to the literature on innovation management (e.g. Annosi et al., 2020; Herfert and Arbige, 2008), alignment with organizational goals is considered as one of the main metrics of product and R&D portfolios as a basis for managers’ decisions on the pursuit of ideas. When an issue is congruent with what the top management has already decided, managers are likely to evaluate it as important and prioritize its’ implementation. However, solving problems that are addressed by prohibitive voice may be of more importance to managers than solving problems that are addressed by promotive voice because they may perceive not endorsing promotive voice and maintaining the status quo as less consequential. As a result, organization-focus may be more important for managers’ evaluations of promotive voice.

The fourth dimension of voice quality is novelty, which refers to proposals that are unique and new to the organization (Brykman and Raver, 2021). These ideas are more likely than those that are not novel to be considered legitimate and then implemented (Lu et al., 2019). This is because novel ideas tend to receive more attention in organizations and more cognitive resources are devoted to process them (Johnston et al., 1990; Schulz, 2001). The increased effort to think about and understand novel ideas increases the likelihood that further steps will be taken (Li et al., 2013). Thus, an idea’s novelty serves as a selection criterion for line managers’ willingness to seek the necessary resources and support of others. Further, these ideas also tend to receive attention from the organizations’ top management (Dutton and Ashford, 1993), as today’s dynamic environment requires organizations to improve existing processes and to generate new ideas in order to gain competitive advantages. However, novelty increases the chances of an idea’s implementation only up to a point because they are often risky and could have negative consequences for their sponsors’ careers if they fail (Yuan and Woodman, 2010). Highly creative ideas usually require a large number of resources and carry a high risk of failure (Škerlavaj et al., 2014). Novelty in prohibitive voice may be less important for managers, as implementing novel solutions are likely to consume more resources (Reiter-Palmon and Illies, 2004), especially considering that problems expressed with prohibitive voice tend to be urgent. Thus, managers are unlikely to require novelty to endorse prohibitive voice.

In sum, rationality may be more important for managers’ evaluations of prohibitive voice, whereas feasibility, organization focus, and novelty may be more important for their evaluations of promotive voice. Therefore, voice’s quality is overall more important for managers’ perceptions of promotive voice.

To demonstrate whether existing findings on managers’ evaluations of promotive and prohibitive voice (e.g. Chamberlin et al., 2017; Huang et al., 2018) can be supported, also the direct effect of voice’s content on managers’ evaluations will be investigated based on two hypotheses:

The role of managers’ mood in their evaluations of voice

The literature reveals that managers’ mood influences both employees’ expression of voice (Liu et al., 2017) and managers’ evaluations of voice (Xu et al., 2020). Individuals who possess power are more likely to experience positive moods, as they have access to resources and autonomy, whereas individuals without power often experience negative emotions such as fear, grief, and embarrassment (Berdahl and Martorana, 2006; Keltner et al., 2003; Langner and Keltner, 2008). A possible explanation for the influence of mood on the evaluation of voice may be derived from the persuasion literature, as persuasion’s mechanisms are similar to those of the evaluations of voice (Howell et al., 2015; Huang et al., 2018).

According to the ELM, which describes the mechanisms behind persuasion, how an individual processes information depends on the person’s ability and motivation to do so (Petty and Cacioppo, 1986).When an individual is able to process information because of, for example, prior knowledge and is also motivated because, for instance, the topic is personally relevant, the likelihood that the person will elaborate on and respond to the information is high (Petty et al., 1983). As a consequence, information is processed via the central route, which leads to critical thinking, recall of past associations from memory, and persistent attitudes (Haugtvedt et al., 1992; Petty and Cacioppo, 1986). When an individual is neither able nor motivated to process information because of, for example, the topic is not of interest or the person is distracted, the likelihood that the person will elaborate on the information is low (Heesacker et al., 1983). In this case, information is processed via the peripheral route, where cognitive effort is reduced, and attitudes are less stable (Petty and Cacioppo, 1986).

Individuals who possess power tend to align their actions with internal desires because they have higher self-confidence and are less dependent on others (Briñol et al., 2007; Magee and Smith, 2013; Rucker et al., 2012). When powerful individuals have a high elaboration likelihood, for instance, if the message’s topic may facilitate or endanger the achievement of their personal goals (Guinote, 2007), they are more likely to rely on their own opinions, making them resistant to persuasion and conformity pressures (Eaton et al., 2009; Galinsky et al., 2008). In contrast, when the elaboration likehood is low, powerful individuals are more easily persuaded compared to powerless individuals, as power serves as a cue that triggers positive feelings (Briñol et al., 2017).

Additionally, power influences social cognition. High-power individuals tend to engage in automatic social cognition processes such as stereotyping and abstract thinking, while low-power individuals scrutinize others’ behavior more carefully (Keltner et al., 2003; Smith and Trope, 2006).

Research shows that mood influences both the central and the peripheral routes of information processing but in different ways (Petty et al., 1988). When an individual is neither able nor motivated to process a message (low likelihood of elaboration), his or her attitude is influenced by mood via a peripheral process that is consistent with the message’s valence (Petty et al., 2003). Consequently, a positive mood tends to result in a favorable attitude toward the message. Individuals use their mood as a cue to decide how to react to a message instead of evaluating the message’s merits, so they reduce cognitive effort and agree with the message when they feel good and refuse it when they do not (Petty et al., 2003).

When the likelihood of elaboration is high, that is, the individual is able and motivated to process a message, the person scrutinizes the message for its merits. Under this condition, mood can cause individuals to retrieve information from memory; that is, depending on whether they are happy or sad (positive or negative mood) positively or negatively associated with the message (Bower, 1981; Petty et al., 1988).

Another explanation for the effect of mood on attitude change may be derived from the Hedonic Contingency Model (Wegener and Petty, 1994), according to which individuals’ behavior is influenced by how the behavior may affect their mood. If individuals are happy, there is a high risk that their actions make them feel worse, while those who are sad have a high chance to feel better after. When a task is likely to worsen their positive mood, individuals will try to avoid it (Kuykendall and Keating, 1990; Wegener and Petty, 1994). In this regard, Wegener et al. (1995) and van ’t Riet et al. (2010) show that a message is more likely to be processed by happy individuals when it is positively framed and pro-attitudinal because both have uplifting effects. Following this reasoning, positive mood makes individuals more sensitive to potential losses than they are to potential gains (Isen et al., 1988). Individuals who are in a positive mood perceive potential losses as especially unpleasant, so they behave as though they have much to lose—not only the potential loss itself but also their feeling of being happy. Consequently, in a situation of uncertainty, individuals place greater importance on the probability of an outcome than on the outcome’s potential value (Isen et al., 1988), and they consider how they may feel when the result is a loss. In contrast, sad individuals may care less about their behaviors’ potential consequences for their mood because it is already negative, and most behavior would make them feel better (Wegener et al., 1995; Wegener and Petty, 1994). Therefore, messages are processed equally, whether they are uplifting or depressing.

Regarding voice, managers who are in a positive mood may be unmotivated to process a proposal when it is of low quality because they don’t want to waste their time processing voice that is, for example, unfeasible (Brykman and Raver, 2021). Their ability to do so may also be limited when the proposal is, for instance, communicated not rationally but emotionally. However, since the persuasion literature shows that individuals’ attitudes are changed by their mood, whether they have a low or high likelihood of elaborating a message (Petty et al., 1993), happy managers may evaluate voice more positively. Therefore, the following hypothesis is derived:

The anticipated effect of managers’ positive mood on their evaluations of voice may be stronger for high-quality voice than it is for low-quality voice. Since happy individuals are more sensitive to potential gains and place greater importance on the probability of an outcome in a situation of uncertainty (Isen et al., 1988), just as managers do to keep their hierarchial position and power within the organization (DeCelles et al., 2012; Keltner et al., 2003), managers might value high-quality voice even more because they assign it a higher probability of success.

Managers who are in a negative mood might evaluate voice in general more negatively, as they tend to refuse messages (Petty et al., 2003) or retrieve information from memory that is negatively associated with employees’ input (Petty et al., 1988). Therefore, the following hypothesis is derived:

Moreover, the effect of managers’ positive mood on their perceptions of voice may be stronger for promotive than prohibitive voice. As prohibitive voice means dealing with a problem, happy managers may see its processing as a potential risk to their positive mood (Wegener and Petty, 1994). In addition, the potential loss associated with not implementing prohibitive voice may be interpreted as more unpleasant than maintaining the status quo (Isen et al., 1988). Therefore, the following hypothesis is derived:

The hypotheses are visualized in Figure 1.

OPEN IN VIEWER

Figure 1.

Research model.

Procedure and sample

To test the hypotheses, a 2 (voice quality: high or low) × 2 (voice content: promotive or prohibitive) × 2 (managers’ mood: positive or negative) factorial, between-subjects experimental design was employed to test the hypotheses. This method has been used several times to investigate managers’ evaluations of voice (e.g. McClean et al., 2022; Whiting et al., 2012) because of its potential to establish causation by manipulating relevant factors and excluding others that might be confounding (Aguinis and Bradley, 2014).

Before testing the hypotheses, an extensive pilot study was conducted online to ensure the success of the research design and test the implemented manipulations. In this way, also the most appropriate approach to manipulate mood for the present study should be identified (Joseph et al., 2020; Westermann et al., 1996). The manipulations were tested among German employees and students who were recruited through personal networks and social media platforms such as Xing and LinkedIn. Participants were randomly assigned to one of eight scenarios adapted from previous experimental studies that investigated the evaluations of (promotive and prohibitive) voice (Weiss and Morrison, 2019; Welsh et al., 2022). Participants were asked to imagine that they lead a team of nine employees at the airline TravelAir. The airline is facing an increasing number of customer complaints related to two central issues: overbooked flights during rush hours and rude flight attendants. On average, participants took 9.57 minutes to complete the questionnaire in the pilot study.

The final sample consisted of 188 participants, with an average age of 28.38 years and 63.9% holding a university degree. Among the participants, 69.1% were female. Employees and students were selected as participants because existing literature does not suggest differences in the manipulations based on participants’ managerial responsibilities (e.g. Weiss and Morrison, 2019; Westermann et al., 1996).

Manipulations

Measures

Procedure and sample

The scenarios for the main study were slightly adjusted based on the results of the pilot study. To enhance the effectiveness of the manipulation, the voice content (promotive or prohibitive) was reiterated at the end of each scenario. In the case of promotive voice, participants read that “the employee came forward with an idea and solution to improve TravelAir’s operations.” For prohibitive voice, participants read that “the employee has come forward with concerns about the plan developed by management, which he believes could hurt TravelAir. He has not expressed any ideas for improving operations.” The final versions of the scenarios can be found in the appendix of this study.

The experiment was again conducted online, and participants, now German managers, were recruited in May 2022 via the online panel bilendi & respondi ¹ and randomly assigned to one of the eight scenarios. These online panels recruit participants online and provide compensation for completing questionnaires (Behrend et al., 2011). Participation is voluntary and based on personal interests and compensation (Landers and Behrend, 2015). Data obtained from online panels offer several advantages, including demographic diversity and reliability, compared to traditionally collected data (Buhrmester et al., 2011). To enhance the validity of the results, recommendations from the literature, such as implementing attention checks and concealing the study’s purpose, were followed (Cheung et al., 2017). Managers were asked to participate in this study as they are the primary recipient of employee voice (Burris et al., 2013; Wåhlin-Jacobsen, 2020).

It took the participants 7.12 minutes on average to complete the questionnaire in the main study. The main study’s sample consisted of 384 managers from various industries. 49.70% of whom were women. The sample averaged 44.13 years of age, 48.30% had a degree from a university, the average organizational tenure was 13.51 years, the average span of control was 17.18 employees, and the average managerial experience was 10.50 years.

The same items were used to measure the independent and dependent variables in the main study as were used in the pilot study, and the same attention and instructional manipulation-check items were included. Again, the reliability coefficients of the used scales were good (α = 0.83 for managers’ valuations of voice, α = 0.93 for managers endorsements’ of voice). Because of the larger differences in mood induced by the short story in comparison to the film clip, mood was manipulated via the short stories used in the pilot study.

The main study controlled for the managers’ openness to experience using the validated German five-item scale from Rammstedt and John (2005; α = 0.78), as openness to experience may influence managers’ evaluation of voice (Huang et al., 2018). An example item was “I’m curious about many different things” (1 = “strongly disagree”, 7 = “strongly agree”). In addition, managerial self-efficacy was measured using the validated German 10-item scale from Jerusalem and Schwarzer (1999; α = 0.91), as the influence of managerial self-efficacy is also demonstrated (Fast et al., 2014). An example item was “I have no difficulty realizing my intentions and goals” (1 = “strongly disagree”, 7 = “strongly agree”).

Manipulation checks

Hypotheses testing

A three-way multivariate analysis of variance (MANOVA) was conducted to test the hypotheses, which is a common way to analyze factorial experimental data (Cole et al., 1993). A MANOVA allows to test for differences between groups with several dependent variables (Bray and Maxwell, 1985). The results revealed that voice quality has a significant effect on managers’ valuation (F[1, 384] = 89.302; p < 0.001; η² = 0.192) and endorsement of voice (F[1, 384] = 109.489; p < 0.001; η² = 0.226), supporting hypothesis 1. The content of voice has a significant effect on both valuation (F[1, 384] = 26.186; p < 0.001; η² = 0.065) and endorsement (F[1, 384] = 12.580; p < 0.01; η² = 0.032), supporting hypothesis 2b. However, hypothesis 2a, which proposes that voice content moderates the effect of voice quality on managers’ evaluations of voice is not supported, as the MANOVA showed no significant effects for the valuation of voice (F[1, 384] = 0.208; p = 0.649; η² = 0.001) or the endorsement of voice (F[1, 384] = 1.103; p = 0.294; η² = 0.003). The direct effect of managers’ mood on their evaluation of voice was also not supported for the valuation of voice (F[1, 384] = 0.000; p = 0.986; η² = 0.000) or the endorsement of voice (F[1, 384] = 0.085; p = 0.770; η² = 0.000), so hypothesis 3a is not supported. Further, the moderating effect of managers’ mood on the relationship between voice quality and the valuation of voice (F[1, 384] = 0.000; p = 0.992; η² = 0.001) and between voice quality and the endorsement of voice (F[1, 384] = 0.203; p = 0.652; η² = 0.001) or on the relationship between voice content and the valuation of voice (F[1, 384] = 0.018; p = 0.894; η² = 0.000) and between voice content and the endorsement of voice (F[1, 384] = 0.000; p = 0.986; η² = 0.000) was not significant, so hypothesis 3b und 3c are also not supported.

The overall model explains 23.6% of the variance in the valuation of voice (R² = 0.236; corr. R² = 0.222) and 25% of the variance in the endorsement of voice (R² = 0.250; corr. R² = 0.236). Including the control variables showed that openness to experience has a significant effect on managers’ valuation of voice (F[1, 384] = 3.981; p < 0.05; η² = 0.011) but not on their endorsement of voice F[1, 384] = 1.691; p = 0.194; η² = 0.005). As for managerial self-efficacy, significant effects were revealed for managers’ valuation of voice (F[1, 384] = 15.534; p < 0.001; η² = 0.041) and endorsement of voice (F[1, 384] = 21.236; p = < 0.001; η² = 0.055). Other than that, only organizational tenure significantly influenced managers’ endorsement of voice (F[1, 384] = 5.417; p = < 0.05; η² = 0.015).

The means and standard deviations of the valuations and endorsements of voice across each condition are shown in Table 1, and the results of the three-way MANOVA are shown in Table 2.

OPEN IN VIEWER

Table 1.

Means and standard deviations across each condition.

Scenario	111	110	101	100	011	010	001	000
Quality	low	high	low	high	low	high	low	high
Content	promotive	promotive	prohibitive	prohibitive	promotive	promotive	prohibitive	prohibitive
Mood	negative	negative	negative	negative	positive	positive	positive	positive
Valuation M	4.81	6.03	4.24	5.32	4.81	6.00	4.24	5.35
Valuation SD	1.31	1.05	1.57	1.11	1.14	0.93	1.22	1.12
Endorsement M	4.71	5.76	4.07	5.58	4.63	5.78	4.19	5.38
Endorsement SD	1.32	1.07	1.44	0.93	1.28	0.86	1.21	1.01
Totals (N = 384)	44	50	48	52	47	47	48	48

OPEN IN VIEWER

Table 2.

Three-way MANOVA for the valuation of voice (and endorsement of voice).

Source of variation	Type III SS	df	MS	F	p	ηp²
Quality	127.083 (144.847)	1 (1)	127.083 (144.847)	89.302 (109.489)	0.000 (0.000)	0.192 (0.226)
Content	37.264 (16.642)	1 (1)	37.264 (16.642)	26.186 (12.580)	0.000 (0.000)	0.065 (0.032)
Mood	0.000 (0.113)	1 (1)	0.000 (0.113)	0.000 (0.085)	0.986 (0.770)	0.000 (0.000)
Quality × Content	0.296 (1.460)	1 (1)	0.296 (1.460)	0.208 (1.103)	0.649 (0.294)	0.001 (0.003)
Quality × Mood	0.000 (0.269)	1 (1)	0.000 (0.269)	0.000 (0.203)	0.992 (0.652)	0.001 (0.001)
Content × Mood	0.025 (0.000)	1 (1)	0.025 (0.000)	0.018 (0.000)	0.894 (0.986)	0.000 (0.000)
Quality × Content × Mood	0.028 (1.041)	1 (1)	0.028 (1.041)	0.029 (0.787)	0.888 (0.376)	0.000 (0.002)
Error	535.073 (497.425)	376 (376)	1.423 (1.323)
Total	10,720.250 (10,355.600)	384 (384)
Corrected total	700.765 (663.380)	383 (383)

Dependent variable: participants’ valuation; R² = 0.236 (corr. R² = 0.222) (dependent variable: participants’ endorsement; R² = 0.250 (corr. R² = 0.236).