Why Didn’t You Tell Me? Voicing Concerns Over Objective Information About a Project’s Flaws

Abstract

Innovation contributes to firm performance. An important task of effective innovation management is to terminate poorly performing projects and reallocate their resources to other, more promising projects. Despite the challenges of such a task, some actors quickly terminate flawed projects while others persist. To investigate decisions about a project’s flaws, we build on theoretical insights from the voice literature to offer a model of voicing concerns over a project’s flaws based on the amount of information available to project team members. We test the model using 3,760 decisions nested within 235 project team members from the research and development (R&D) departments of large firms operating in innovative industries. We find that more information about a project’s flaw increases project team members’ willingness to voice concerns and that this positive effect is stronger for project team members who believe that they have more open-minded supervisors and who themselves are more prosocially motivated. This theorizing and set of findings provides new insights into our understanding of facilitating voice in organizations and managing innovation and entrepreneurial projects.

Keywords

corporate venturing entrepreneurial cognition/psychology innovation management employee voice entrepreneurship

Innovation is critical for firms to achieve high performance (Hull & Rothenberg, 2008; Li & Atuahene-Gima, 2001; Thornhill, 2006). Not surprisingly, scholars have been interested in understanding the innovation process. Typically the innovation process involves projects progressing through a number of stages (Bakker & Shepherd, in press; Cooper & Sommer, 2016). As a project progresses through the stages of the innovation process, more information becomes available and the required investment of resources also increases (Bakker & Shepherd, in press; Bowman & Hurry, 1993). An important task of effective innovation management is to terminate poorly performing projects and reallocate their resources to other, more promising projects (Klingebiel & Adner, 2015). However, research has found that the decision to rapidly terminate a poorly performing project is highly challenging to make.

First, although the costs are designed to be limited, there are some sunk costs into the project as well as other factors that encourage persistence (Barsky & Zyphur, 2016; Boulding, Guha, & Staelin, 2016; Roy & Sarkar, 2015). Second, an antifailure bias is pervasive in organizations and society (Decker & Mellewigt, 2007; Lee, Peng, & Barney, 2007; Peng, Yamakawa, & Lee, 2010), which encourages persistence despite poor performance (for a review, see Sleesman, Conlon, McNamara, & Miles, 2012). Finally, information about problems with the project may be asymmetrical such that some know of the problems while others, perhaps also including the supervisor, remain ignorant of them (Booth & Schulz, 2004). Despite these challenges, some actors quickly terminate flawed projects while others persist. Persisting despite poor performance creates a serious “drag” for the organization as resources are invested with little hope of a sufficient return (Drummond, 2014) and/or advancing the organization’s strategic goals (Unger, Kock, Gemünden, & Jonas, 2012).

The above leads to the research questions: To what extent does the amount of information on a project’s flaws influence project team members’ willingness to share this information with the supervisor responsible for the persistence/termination decision, and to what extent does this relationship vary across project team members? To address these questions, we build on theoretical insights from the voice literature (Detert & Burris, 2007; Detert & Edmondson, 2011; Fast, Burris, & Bartel, 2014; Milliken, Morrison, & Hewlin, 2003; Morrison, 2014; Van Dyne, Cummings, & Parks, 1995) to offer a model of voicing project concerns. Voice is “speaking up in ways that seek to challenge or replace the established practices, policies, or strategic directions that comprise the status quo among these individuals who created or otherwise sustain those aspects of the organization” (Fast et al., 2014: 1015).¹

Voice is particularly relevant in the project termination context because (a) once created, a project has sustained momentum, including the commitment of people who volunteered for the project and its project champion (Royer, 2003); (b) to communicate flaws may disrupt the status quo (Graham & Van Dyne, 2006)—namely, to change direction by terminating the project (Janney & Dess, 2004); and (c) people like the status quo (Frese & Fay, 2001), especially for a project they are working on, and are likely unhappy with the person who “rocks the boat”; but (d) “rocking the boat” in the case of a flawed project benefits the organization (Adner & Levinthal, 2004; McGrath, 1999; Royer, 2003). Therefore, although voicing concerns about a project’s flaws to the supervisor may benefit the organization, the individual who voices these concerns often faces personal costs (Ashford & Gibbs, 1990; Detert & Burris, 2007; Milliken et al., 2003) such that the decision to use voice has been explained as either being fear-driven or a cognitive calculation (Detert & Edmondson, 2011; Liang, Farh, & Farh, 2012).

In this article, we develop a model of voicing project concerns and test it using 3,760 decisions nested within 235 project team members from the research and development (R&D) departments of large firms operating in innovative industries. Specifically, we investigate the attributes of perceived feasibility (namely, amount of information on project concerns and perception of supervisor’s managerial openness) and perceived desirability (namely, prosocial motivation and affective commitment) in explaining the willingness to voice project concerns. In doing so, we make four primary contributions. First, although voice research on the notion of psychological safety finds that an open-minded manager encourages employees to engage in voice (e.g., Detert & Burris, 2007; Dutton & Ashford, 1993; Nembhard & Edmondson, 2006; Takeuchi, Chen, & Cheung, 2012), we theorize and find that this relationship is more complicated—perceived managerial openness only plays a role in promoting team members’ voice when considered in conjunction with the amount of information underlying the concerns. An interesting insight is that efforts to train and recruit open-minded managers (Ribeiro-Soriano & Urbano, 2010) will have little impact on voice unless the project team also develops and collects considerable data on a project’s performance; only then is negative project performance information likely voiced. Therefore, instead of emphasizing the overall benefits of managerial openness for triggering voice (Detert & Burris, 2007; Mowbray, Wilkinson, & Tse, 2015), our study highlights the boundary conditions of these benefits.

Second, research on prosocial motivation highlights people’s willingness to take actions to benefit others (Grant, Parker, & Collins, 2009) and/or the organization (Grant & Mayer, 2009), and voicing concerns to superiors is one such prosocially motivated action (Fuller, Marler, & Hester, 2006; Liang et al., 2012; Tangirala, Kamdar, Venkataramani, & Parke, 2013). While it appears that prosocial motivation enhances employees’ tendency to take action (i.e., engage in voice) to prevent undesirable outcomes (direct effect of prosocial motivation) (Burris, Detert, & Romney, 2013; Detert & Burris, 2007; Van Dyne & LePine, 1998), we theorize and find that a high tendency to act based on prosocial motivation seems to require considerable information about the nondesirability of the outcome. The danger is that prosocially motivated employees do not engage in voice when they detect weak or ambiguous signals about threats of undesirable outcomes, thereby increasing the likelihood of persistence and the further “wasting” of resources by continued investments in failing projects. Therefore, the counterintuitive insight is that those most motivated to help others require more evidence to decide to act in a way that reduces the costs of failure for the organization, the supervisor, and/or the team.

Third, voice is typically investigated using survey scales (Burris et al., 2013; Detert & Burris, 2007; Detert, Burris, Harrison, & Martin, 2013). While most survey-based measures capture voice more in terms of a general behavior at work (across many situations), the use of conjoint analysis allows the researcher to focus on a specific situation, in this case a team member’s willingness to engage in voice in the detection of project flaws. Furthermore, conjoint analysis is well-suited to explore theories-in-use (Priem, 1992, 1994) such as implicit voice theories (Detert & Edmondson, 2011). We are not suggesting that surveys are inappropriate for investigating voice; rather, we offer conjoint analysis as a methodological approach for understanding the decision making underlying voice as a means of opening up new terrain for future research.

Finally, the innovation and corporate entrepreneurship literatures highlight the importance of quickly noticing, interpreting, and terminating poorly performing projects and redeploying resources that show promise but also acknowledge heterogeneity in firms’ ability to do so (Bakker & Shepherd, in press; Brown & Eisenhardt, 1997; McGrath, Ferrier, & Mendelow, 2004). Although much has been made of why it is important to collect evidence of a project’s performance (Cooper, 2008; Griffin, 1997) and how disconfirming evidence is either ignored or discounted (Ng, Westgren, & Sonka, 2009; Tripsas & Gavetti, 2000), an implication of the current study is that the extent to which information is shared with decision makers is critical and depends on the project team member’s (i.e., information holder’s) perception of his or her supervisor as well as the project team member’s prosocial behavior.

A Model of Voicing Project Concerns

Voice is critical to challenging the status quo of projects. Voice is a voluntary behavior (Van Dyne et al., 1995), and encouraging project team members to voice concerns over the status quo of poorly performing projects can be important to organizations (Farson & Keyes, 2002; Royer, 2003). Indeed, employees’ voice has been associated with improving business performance (Morrison & Milliken, 2000) by quickly revealing work-related problems (Tangirala & Ramanujam, 2008), preventing crisis (Schwartz & Wald, 2003), and facilitating organizational learning (Tangirala & Ramanujam, 2008). A project team member’s voice is typically expressed to his or her supervisor (e.g., Burris, 2012; Detert & Burris, 2007; Detert et al., 2013; Edmondson, 2003; McClean, Burris, & Detert, 2013; Tangirala & Ramanujam, 2012). Voicing concerns to a supervisor has the greatest potential to enhance performance because supervisors serve as a lynchpin between project team members close to operations of the status quo and managers responsible for major change (Marchington & Suter, 2013; Townsend, Wilkinson, & Burgess, 2013). As a result, supervisors are typically tasked with encouraging (Detert & Burris, 2007; Detert & Trevino, 2010) and managing member voice (Marchington & Suter, 2013).

Given the benefits to the organization of project team members’ voice, the prescription for managers has been to encourage members to engage in voicing behavior (Edmondson, 1999; Morrison & Milliken, 2000). However, project team members often do not engage in voice even when given the opportunity to do so (Detert & Burris, 2007). Although benefits may accrue to the employees who engage in voicing concerns with current operations (Allen & Rush, 1998; Griffin, Neal, & Parker, 2007; Johnson, Erez, Kiker, & Motowidlo, 2002)—for example, higher supervisor performance evaluations (Grant et al., 2009)—engaging in voice to disrupt the status quo of an existing project may not be an unambiguous blessing for the project team member expressing concerns about the project. Indeed, many managers have shown reluctance and even an aversion to seek and implement solutions for issues raised through voice (Milliken et al., 2003).

First, voicing concerns about the viability of a project—regardless of its organizational benefit—may be perceived as a criticism of the leader himself or herself. That is, managers can find negative feedback personally threatening and, as a result, act defensively to protect their ego (Ashford & Cummings, 1981; Detert & Edmondson, 2011; Kluger & DeNisi, 1996; Morrison & Milliken, 2000), especially when it comes from subordinates (Korsgaard, Roberson, & Rymph, 1998). Second, other project team members, including the supervisor, have made contributions to the project and thus may look unfavorably on the individual who “rocks the boat.” Indeed, proactive behavior, such as voice over project flaws, “is not always welcomed by supervisors. … Initiative ‘rocks the boat’ and makes changes. Since people tend not to like changes, they often greet initiative with skepticism” (Frese & Fay, 2001: 141; see also Burris, 2012). Finally, a project team member voicing concerns over a project can sometimes be seen by others as engaging in self-serving rather than benevolent behavior. To the extent that voice is attributed to self-serving motives, there will be few rewards and possibly sanctions (Chiu, Morris, Hong, & Menon, 2000). For example, individuals who speak up for change have been described as manipulative (Ashford & Gibbs, 1990) and troublemakers (Redding, 1985).

Thus, project team members considering voicing concerns are faced with a number of benefits and costs. Morrison’s (2014) model of the antecedents of employee voice highlights a number of mechanisms by which project team members decide whether to voice concerns when presented with a latent voice opportunity (e.g., the opportunity to voice concerns about project flaws). One such mechanism is the subjective expected utility calculus of voicing concerns (Morrison, 2014; see also Ashford, Rothbard, Piderit, & Dutton, 1998; Milliken et al., 2003). In the present project team context, we will term this cognitive calculation “perceived feasibility.” It is the team member’s subjective perceived likelihood that her or his voice will have an effect such that the supervisor will heed the concerns voiced. Morrison’s (2014) model further suggests that motivators influence the perceived feasibility of voice. As shown in our decision-making model of voicing project concerns in Figure 1, we propose two motivators that influence the perceived feasibility of voice. The first is reflected in the amount of information supporting the concern voiced; the more supporting information, the more feasible voice will seem. However, we theorize that the impact of the perceived feasibility to engage in voice also depends on the project team member’s perception of the supervisor’s openness to the new information; the more open the supervisor is perceived to be, the more feasible voice will seem.

Figure 1

A Decision-Making Model of Voicing Project Concerns

A second motivator in Morrison’s (2014) model is prosocial motivation, such that an increased desire to expend effort for others’ benefit can make team members more willing to engage in voice.² As shown in Figure 1, we suggest that two team member factors influence the “perceived desirability” of voicing project concerns, namely, team members’ affective commitment (desire to help the organization; Jacobsen, 2000; Van Dyne et al., 1995) and trait prosocial motivation (desire to help others; LePine & Van Dyne, 1998; Van Dyne, Ang, & Botero, 2003). We build on these motivations to gain a deeper understanding of how the desirability of voice for the team member impacts the emphasis on the amount of information in determining the willingness to engage in voice.

Amount of Information and Voicing Project Concerns

Assessment of the feasibility of voicing concerns about a project is likely enhanced when the project team member’s voice is based on objective information to back up concerns about the current project. First, considerable information on the project’s flaws may not fully alleviate a manager’s feelings of threat, but it likely makes him or her more inclined to react positively to the information because it helps him or her avoid a more costly error that arises from continued investment in a flawed project. The longer a manager’s commitment to a flawed project, the greater the accumulation of losses, and the more vulnerable he or she is to an unfavorable assessment by top management. For example, middle managers are more hesitant to approach top management when there is the potential for larger than for smaller failures (Dutton, Ashford, O’Neill, Hayes, & Wierba, 1997; Schilit, 1987). Thus, a project team member’s voice can help supervisors (as well as the organization) avoid larger failures (Farson & Keyes, 2002; Royer, 2003).

Second, while others involved in the project may enjoy the status quo, they are more likely to accept someone challenging that status quo when there is considerable information about the project’s flaws. That is, even when those enjoying the status quo approach information suggesting the need for change with skepticism, considerable information is more likely to overcome this skepticism to raise doubts about the viability of the project than if there is lesser information. For example, research on beliefs in cognition and decision making has argued that individuals form beliefs based on information collected from their environment (e.g., Hedstrom & Swedberg, 1998: 128). With more information on a project’s performance, project team members are likely to feel better prepared to shape their manager’s belief that the project is indeed flawed, and they are more likely to receive a positive reaction for voicing their concerns than if they try to challenge the status quo with less information at hand.

Finally, and consistent with the above, voicing concerns based on considerable information on the project’s flaws will likely alleviate the attribution of self-serving motivations to the project team member voicing concerns. That is, while voice may still be unwelcomed, a greater amount of information likely alleviates suspicions that the voicing project team member is simply trying to manipulate the situation or cause others trouble. The above reasoning leads to the following hypothesis, which forms our baseline for subsequent theorizing:

Hypothesis 1: The greater the amount of information on a project’s flaws, the more willing the project team member is to voice concerns over that project.

Decision-Maker Heterogeneity and the Role of the Amount of Information

Our subsequent theorizing explains variance in the nature of the relationship between the amount of information and voice based on the extent to which the project team member believes voice will have an effect and that the effect is desirable. For voice to have an effect, the project team member’s manager needs to be open to being convinced of the project’s flaws. That is, not all managers feel threatened by voice, are attached to the status quo, and/or assume voice represents self-serving behaviors. By opening themselves to project team members who voice concerns on a focal project, managers encourage project team members to speak up about ways to improve the current or subsequent projects. Indeed, managerial openness has been found to be positively related to employees engaging in voice (Detert & Burris, 2007; Mowbray et al., 2015). As managers are more open to voice, they can generate higher follower motivation (Zapata-Phelan, Colquitt, Scott, & Livingston, 2009), decrease turnover (McClean et al., 2013), create a climate of psychological safety (Fast et al., 2014; Nembhard & Edmondson, 2006), and thereby become a better manager (Morrison, 2011). However, encouraging more voice may create some problems for managers and project effectiveness. That is, such encouragement of voice may generate more implicit or explicit criticism of managers because they perceive voicing concerns as personally threatening (Burris, 2012) and/or a distraction from more important operational tasks (Detert & Trevino, 2010).

In anticipating the benefits and costs of voice for the more open-minded manager, project team members are likely to be guided by the amount of information backing up their concerns. If there is considerable evidence of the project’s flaws, then it is less likely that voicing concerns will be viewed by the open-minded manager as a distraction to the project process; however, this is less likely to be the case when the evidence is minimal. In contrast to the more open-minded manager, a more closed-minded manager is less receptive to new information about the project’s flaws, particularly information that challenges the status quo. This manager will remain largely unaffected by the amount of information about a project’s flaw. Based on the above reasoning, we offer the following:

Hypothesis 2: The positive relationship between the amount of objective information about a project’s flaws and the project team member’s willingness to voice concerns over that project is greater for project team members who believe that their supervising manager is more open-minded than for project team members who believe that their supervising manager is less open-minded.

Motivating Voice About Project Concerns

Although we have theorized that the ratio of benefits to costs is enhanced when the voiced concerns are backed up by considerable information on the project’s flaws, this calculation of the “value” of the quality of evidence prompting voice likely varies based on employees’ motivation—prosocial motivation and affective commitment.

Prosocial motivation refers to “the desire to expend effort in order to benefit other people” (Grant & Sumanth, 2009: 928). Prosocial motivation likely influences employees’ emphasis on the evidence necessary to voice project concerns to the manager. Individuals who are more prosocially motivated focus more of their attention on external sources of information than on internal sources (Grant & Mayer, 2009; Meglino & Korsgaard, 2004). Therefore, while many gain intrinsic benefits from engaging in the project (Zhang & Bartol, 2010), employees focused outwardly (e.g., those who are more prosocially motivated) are more likely to recognize, interpret, accumulate, and not quickly discount external information indicating a flaw with the project. With greater attention to this external evidence, they are more likely to be guided by it when deciding whether or not to voice their concerns.

Furthermore, individuals who are more prosocially motivated are willing to subjugate their personal interests to contribute to the organization (Bolino & Turnley, 2005; Grant & Mayer, 2009; Meglino & Korsgaard, 2004). This suggests that the more information such project team members have on a project’s flaws, the more willing they are to take the personal risks of voice to ensure that others become aware of the issue. That is, project team members who are prosocially motivated place greater value on improving the welfare of others and the organization (Grant, 2007, 2008; Ilies, Scott, & Judge, 2006) than those who are less prosocially motivated. In calculating whether their voicing concerns will help (i.e., the extent to which it will help others and the organization), prosocially motivated persons tend to avoid suggesting changes that might not lead to positive outcomes or that might give the appearance of them complaining. Therefore, they are more likely to voice a concern when they are more certain that it will bring about such positive change/improvement. More information makes them more certain that voicing a concern will bring about positive change relative to when they have less information and might just be working off of a hunch. In contrast, less prosocially motivated persons are less concerned about whether their suggestions for change will lead to positive outcomes or help others, so they are more likely to voice concerns even if they do not have much information to back them up. Based on the above reasoning, we offer the following:

Hypothesis 3: The positive relationship between the amount of information on a project’s flaws and the project team member’s willingness to voice concerns over that project is greater when the project team member is more prosocially motivated than less prosocially motivated.

Project team members can also be motivated to “do good” for the organization when they are affectively committed to the organization. Affective commitment refers to the “extent that an individual perceives his or her affective or emotional attachment to the organization” (Allen & Meyer, 1990) such that they experience “positive feelings of identification with, attachment to, and involvement in, the work organization” (Meyer & Allen, 1984: 375). This affective commitment likely influences the extent to which a project team member relies on the amount of evidence regarding a project’s flaws in voicing concerns.

First, affective commitment indicates a motivation to engage in actions that help and do not harm the organization (Meyer & Allen, 1997). On the one hand, voicing concerns about the project based on minimal evidence can harm the organization by distracting the team from the task at hand (DeNisi & Kluger, 2000); by questioning project team members’ self-efficacy and collective efficacy, thus diminishing perceptions of the feasibility of the current project and undermining performance (Bandura & Locke, 2003); and by undermining project team member motivation (Klein, 1990). On the other hand, by not voicing concerns that eventually lead to project failure, the individual’s lack of voice may cause greater harm to the organization by delaying the termination of the project—the cost of failure may be greater than it needs to be (Sharp & Salter, 1997), and the organization may experience greater opportunity costs with this project (Drummond, 2014; Elfenbein & Knott, 2015).

Second, affective commitment involves positive feelings through identification with the organization (Allen & Meyer, 1990). These positive feelings can open the project team member’s mind to evidence of what a project’s flaws can mean for the organization. Indeed, positive emotions have been associated with undoing negative emotions (e.g., the negative emotions from anticipating the grief over the realization that the project will not be a success [Shepherd, Wiklund, & Haynie, 2009]) and thereby expanding the scope of attention and cognition (Fredrickson, 1998, 2001). Individuals with high positive emotions are more open to evidence and are more likely to use it in their decision making (Fredrickson, 1998, 2001). In the absence of these positive emotions, negative emotions generated from being confronted with evidence of the project’s flaws can lead to information being ignored and discounted or, if interpreted, leading to threat rigidity (Staw & Ross, 1989; Tripsas & Gavetti, 2000), which in turn maintains the status quo largely regardless of the amount of evidence about the project’s potential flaws.

Finally, project team members affectively committed to the organization are more prepared to engage in extrarole behaviors to benefit the organization (Meyer & Allen, 1991; Mowday, Porter, & Steers, 1982), particularly those that have greater potential for problem solving (Luchak, 2003). More evidence of the flaws of a project provides greater potential for problem solving even if it means terminating the current project and going back to the “drawing board.” Based on the above, we offer the following:

Hypothesis 4: The positive relationship between the amount of information on a project’s flaws and the project team member’s willingness to voice concerns over that project is greater when the project team member is more affectively committed to the organization than less affectively committed.

Research Methods

Sample

We explored the willingness to voice project concerns with a sample of project team members from R&D departments of large German corporations operating in a variety of industries. To generate a list of potential firms relevant for our empirical study, we drew on the Hoppenstedt database, which is a comprehensive database of German corporations. We applied several selection criteria to derive the final sample. First, we focused on firms with more than 500 employees because large firms tend to have more continuous innovation processes over time and more resources to engage in entrepreneurial and potentially failing projects (Covin, Garrett, Kuratko, & Shepherd, 2015). Furthermore, we eliminated all firms that did not have dedicated budgets for R&D activities. Finally, we focused on firms for which the Hoppenstedt database provided the name of the senior-level R&D manager so that we could contact them to encourage participation. Overall, the list contained 324 potentially relevant firms.

In the next step, we randomly called the senior-level R&D managers of these firms and asked them for participation in our study. However, we soon realized that most of the managers contacted considered the time investment of one hour for several (we asked for at least 10) R&D project team members to be too substantial and costly to commit to our study. Therefore, in addition to cold calling firms on our list randomly, we leveraged personal contacts into firms on the list. Overall, we made contact with 51 firms. When we contacted them by phone, we explained that we were conducting a study about failing R&D projects and asked for email addresses and names of R&D project team members as potential participants. To encourage participation, we promised the firms a detailed report about the study’s results. Altogether, 16 firms participated in our study (11 through cold calling and 5 through leveraging personal contacts). The 16 firms had on average 47,000 (SD = 96,000) employees and annual revenues of 11,898 (SD = 21,896) million euros. These firms provided names and email addresses of 283 R&D project team members, who received an email link to our research instrument. In the same email, we assured project team members full anonymity. A reminder was sent to those who did not participate after two weeks. This approach yielded 254 participant responses, 173 from the cold-called firms (15.7 participants per firm), and 81 from those in which we had personal contacts (16.2 participants per firm). Thus, our response rate was 31.3% in terms of firms contacted, and 90.0% in terms of project team members invited to our study. The firms’ sectors were machinery and instruments (six firms), electronics (three firms), materials (two firms), chemicals (two firms), manufacturing (two firms), and transportation (one firm).

In terms of demographic characteristics and background, the project team members were on average 42.6 (SD = 9.9) years old, 88.6% were male, 87% were German nationals, 26.4% held a PhD degree, and 75.6% had a background in engineering. Their mean work experience was 17.8 (SD = 10.6) years, average industry experience was 13.5 (SD = 9.8) years, average firm tenure was 11.2 (SD = 10.1) years, and the time they held their current position was 7.3 (SD = 7.2) years. Further, 40.2% of participants were in management positions, and they had substantial project experience—on average, they had worked on 15.8 (SD = 21.9) projects for their current firms. For the projects in their firms, the participants reported a mean R&D budget of 0.93 (SD = 2.13) million euros, a team size of 12.3 (SD = 20.1) project team members, and a failure rate of 21.7% (SD = 24.6%). We did not find a significant difference between early (first 25%) and late (last 25%) respondents to our survey with respect to these individual-level variables. Since late respondents can serve as a proxy for nonrespondents (Armstrong & Overton, 1977), this finding suggests that there is unlikely to be nonresponse bias in our dataset. Furthermore, there was not a significant difference between the responses generated through cold calls and those generated through personal contacts.

Conjoint Analysis

To investigate a project team member’s decision to voice concerns about a project, we used a metric conjoint experiment. This methodology decomposes decision makers’ evaluations into underlying structures (Louviere, 1988), which allows for studying how the amount of data on a project (as a decision cue) impacts participants’ decision policies to voice project concerns and enables us to explain variance across project team members’ decision policies in terms of their manager’s open-mindedness, their prosocial motivation, and their affective commitment to the firm. The metric conjoint study method has been widely used to understand decision making in the disciplines of strategic management (e.g., Pablo, 1994; Priem, 1992), marketing (Green, Krieger, & Wind, 2001; Green & Srinivasan, 1990), and entrepreneurship (e.g., Shepherd, Patzelt, & Baron, 2013) and therefore is highly appropriate for investigating the decision making of project team members with regards to voicing concerns over a project to their supervisor.

In a metric conjoint experiment, participants evaluate a number of hypothetical scenarios (i.e., profiles), which are described by a combination of decision attributes. Specifically, in the current study, the scenarios described different situations for underperforming projects, and participants evaluated their likelihood of voicing their concerns in response. With each of the attributes being described by two levels (“high” and “low”), this study used a set of unique scenarios with different combination levels for the three decision attributes (independent variable: amount of data; control variables: manager identification with the project and participant identification with the project; see below). We used an orthogonal fractional factorial design to determine that eight scenarios would enable us to test the hypothesized relationships (Hahn & Shapiro, 1966). The eight original profiles were fully replicated as a test-retest of participants’ reliability of evaluations, and an additional profile was added at the beginning of the experimental task for practice (but was not included in the analysis). Therefore, each participant evaluated a total of 17 hypothetical project scenarios. In the statistical analysis (which we performed using hierarchical linear modeling [HLM]; see below), the attributes and the participants’ decisions were individual-level variables at Level 1, whereas the variables capturing variance across individuals were treated as Level 2 variables. Therefore, our hypotheses explore cross-level interaction effects—that is, how Level 2 variables (their manager’s open-mindedness, their prosocial motivation, and their affective commitment to the firm) impact the effect of the Level 1 variable (amount of data) on project team members’ decision to voice concerns over the project.

To ensure that all participants were “on the same page,” we explained a common context before they evaluated the hypothetical project scenarios. At the beginning of the experiment, participants received the following instructions for the general situation:

As an employee of an established technology firm, you are asked to evaluate hypothetical R&D projects in your company context. You are part of the project team and have serious concerns about whether the successful completion of the project is still possible. In the following, the task is to judge your willingness to communicate your concerns about the continuation of the project to your supervisor.

Research Variables

Dependent variable: Willingness to voice concerns about a flawed project

Following similar metric conjoint studies (e.g., McKelvie, Haynie, & Gustavsson, 2011), team members assessed their willingness to voice concerns about an underperforming project to their supervisor on a 7-point Likert-type scale, which was anchored by the endpoints of 1 (very unlikely to voice concerns about the project) to 7 (very likely to voice concerns about the project). It is important to note that this conceptualization of the dependent variable differs from established scales used to operationalize voice. Specifically, taking a decision-making perspective and an experimental approach, we focus on team members’ willingness to engage in voice in a specific situation (the detection of project flaws), while most survey-based measures capture voice more in terms of a general behavior at work (across many situations) and draw on items such as “I give suggestions [to the General Manager] about how to make this restaurant better” (Detert & Burris, 2007; see also Burris et al., 2013) or the assessment of how often (on a scale from seldom to always) employees speak up to a specific target manager (Detert et al., 2013). Despite this difference, the dependent variable in the present study and established voice scales both focus on the same behavioral domain—voicing concerns to supervisors.

Independent variable at Level 1: Amount of data on a project’s flaws

We operationalized this variable in the profiles describing the decision scenarios at two levels: high (“There is substantial objective information supporting your hunch that the project will likely fail due to a fatal flaw”) and low (“There is little objective information supporting your hunch that the project will likely fail due to a fatal flaw”).

Control variables at Level 1

Project team members differ in their identification with projects such that the anticipation of losing a project with which they strongly identify may generate negative emotions and delay actions that lead to its termination (Shepherd, Covin, & Kuratko, 2009). Therefore, based on the organizational identification scale developed by Smidts, Pruyn, and Van Riel (2001), we offered two additional decision cues. (1) Personal identification with the project as high (“You have strong ties with the project, experience a strong sense of belonging to the project, and feel a great deal of pride from working on the project”) or low (“You have weak ties with the project, experience a weak sense of belonging to the project, and feel little pride from working on the project”). (2) Given that voicing concerns about a project flaw might upset the supervisor (see our earlier arguments), this effect (and the supervisor’s experience of negative emotions) might be stronger the more she or he identifies with the project, so we included the following decision cue: supervisor’s identification with the project as high (“The supervisor has strong ties with the project, experiences a strong sense of belonging to the project, and feels a great deal of pride from working on the project”) or low (“The manager has weak ties with the project, experiences a weak sense of belonging to the project, and feels little pride from working on the project”).

In addition to instructions described above, participants received an explanation of the hypothetical project. Specifically, to control for different causes of the project flaw, we crafted two different versions of project scenarios (second version indicated in parentheses):

Your R&D project is structured into four phases: feasibility analysis, development, market testing, and commercialization. Currently, the project is in Phase 2 of development. As part of the project team, 100% of your time at work is dedicated to the project and has been since the project started. Meanwhile, you believe that the project leader’s plans were much too ambitious with respect to both timing and the budget, which cannot be corrected at this point in time such that the achievement of the intended goals is heavily threatened. (Meanwhile you believe that you and the project team do not have the necessary knowhow and experience to realize the project within the required time and financial budget such that the achievement of the intended goals is heavily threatened.) You believe that the project should be terminated in order to avoid further investments of resources, energy, and time. The relevant person to voice these concerns to is your supervisor, who is in the position to terminate the project. You are considering voicing your opinion in the next team meeting or in a personal conversation with the supervisor.

Therefore, our profiles contained three decision cues representing variables at Level 1. To test for order effects of decision cues and profiles, we created four versions representing a 2 × 2 matrix (two random orders of decision cues and two random orders of profiles). Since we also manipulated causes of project failure in the instructions of the experiment (see above), our final experimental design yielded eight versions in total. There were no statistically significant differences in participants’ responses across these eight versions (p > .1).

Moderating variable at Level 2: Managerial openness

Following Tröster and van Knippenberg (2012), we assessed project team members’ perceptions of their supervisors’ openness using a three-item, 7-point Likert-type scale anchored by the endpoints I do not agree at all (1) and I totally agree (7). Exemplary items include “My supervisor is open to suggestions” and “Good ideas get serious consideration from my supervisor.” The Cronbach’s alpha of this variable was .94 in our study, indicating high reliability (Hair, 2010).

Moderating variable at Level 2: Prosocial motivation

Following Grant and Sumanth (2009), we assessed each respondent’s prosocial motivation using a five-item, 7-point Likert-type scale anchored by the endpoints I do not agree at all (1) and I totally agree (7). Exemplary items include “I get energized by working on tasks that have the potential to benefit others” and “It is important to me to have the opportunity to use my abilities to benefit others.” The Cronbach’s alpha of this variable was .86, indicating high reliability (Hair, 2010).

Moderating variable at Level 2: Affective commitment

Following Allen and Meyer (1990), we assessed each respondent’s affective commitment to his or her organization using an eight-item, 7-point Likert-type scale anchored by the endpoints I do not agree at all (1) and I totally agree (7). Exemplary items include “I enjoy discussing my organization with people outside it” and “I really feel as if this organization’s problems are my own.” The Cronbach’s alpha of this variable was .84 in our study, indicating high reliability (Hair, 2010).

Control variables at Level 2

In addition to control variables at Level 1, we also included two control variables at the individual level of analysis. First, we controlled for participants’ age since older employees have been found to engage in more voicing behaviors (Detert & Burris, 2007). Second, we controlled for participants’ gender since there is an indication of a positive correlation between being male and voicing behaviors (Van Dyne & LePine, 1998). We also wanted to rule out the possibility that firm or industry characteristics did not confound our results. We reran the analysis reported below with, first, the addition of dummy variables for firm; and second, the addition of dummy variables for industry. Both analyses resulted in findings that did not substantively differ from those reported below.

Potential Limitations of Conjoint Analysis

While metric conjoint analysis has many advantages for our research (e.g., the possibility to explore cross-level interactions between variables as posited in Hypotheses 2–4 and the avoidance of self-report biases [cf. Lohrke, Holloway, & Woolley, 2010], which are likely in studies of voicing concerns), one potential criticism is that the hypothetical scenarios might suffer from low face validity. We took several steps to minimize this concern. First, we derived the decision attributes presented in the scenarios based on strong theoretical arguments and prior research (see above). Second, we pretested the experimental design with two R&D managers and two R&D project team members of large corporations. We also interviewed these individuals with respect to their opinion on the experiment and descriptions offered. We received only minor suggestions for improvement of the introduction to the experiment, which we incorporated in the final conjoint instrument. All interviewees reported that the described situations were realistic and the questions answerable. Typical comments we received included the following: “If there are substantial data on the upcoming failure, the willingness to voice concerns is very high. It is even a necessity [to voice concerns]”; “The description [of the situation] is practical”; and “The information [provided at the beginning of the experiment] is quite interesting because this is exactly what we are facing at the moment [in our company].” Finally, we asked participants to self-assess the importance of the decision attributes using a Likert-type scale ranging from very unimportant (1) to very important (7). The self-report average values were 3.55 (supervisors’ identification with the project), 4.23 (personal identification with the project), and 6.2 (amount of objective information on a project’s flaws). Thus, all decision cues were reported to have at least some espoused importance, with the independent variable of our study (amount of objective information on a project’s flaws) being reported as highly important.

Statistical Procedure

Since in the experimental conjoint task each participating project team member made 16 assessments of hypothetical project profiles, the data points are not independent of each other but nested within individual participants. To account for this nested nature of our data, we applied HLM for statistical analysis (Bryk & Raudenbush, 1992). HLM explains the total variance in participants’ assessments in terms of variance between decisions (Level 1, based on the decision cues presented in the profiles) and in terms of variance across individuals (Level 2, based on differences between decision makers). HLM also allows for the analysis of cross-level effects—that is, how attributes of the decision makers (here in terms of participants’ perceptions of managerial openness, prosocial motivation, and affective commitment) explain the strength of the Level 1 relationships (here between the amount of objective information and the likelihood of engaging in voicing behaviors).

Results

In total, 254 R&D project team members completed the conjoint experiment and postexperiment questionnaire. As detailed above, for each respondent, the original set of profiles was replicated, allowing for a test-retest calculation of reliability. In all, 17 of the respondents had unreliable responses (below .40) and, consistent with other conjoint studies (Hitt & Middlemist, 1979; Keats, 1991; Shepherd, Patzelt, & Baron, 2013; Tyler & Steensma, 1995), these nonreliable responses were dropped from the study. After dropping the unreliable respondents, the mean test-retest correlation was .84. Because of missing data, an additional two project team members were dropped. Therefore, the analysis was based on 3,760 decisions nested in 235 project team members.

In Table 1, we present the descriptive statistics and the intercorrelation matrix for all Level 2 variables. We do not display the Level 1 variables because, given the orthogonal design, there is zero correlation between them. Although most correlations between the Level 2 variables are low, there are some moderate-level correlations. To determine if these correlations created issues of multicollinearity, we calculated the value inflation factors (VIFs) and found that the highest VIF was 1.192, and we calculated the condition index and found that the highest index was 23.281. These scores are well below the recommended limits of 10.0 (Kutner, Nachtsheim, & Neter, 2004) and 30 (Belsley, Kuh, & Welsch, 1980), respectively.

Table 1

Means/Frequencies, Standard Deviations, and Correlations for Level 2 Variables

Variables	M	SD	1.	2.	3.	4.
1. Age	42.45	9.81	1
2. Gender (share female)	12%	n.a.	−.24**	1
3. Managerial openness	17.61	3.60	−.03	−.09	1
4. Prosocial motivation	28.60	4.56	−.07	−.00	.30**	1
5. Affective commitment	38.42	8.73	−.02	−.11	.33**	.28**

Note: n = 235.

p < .01.

In Table 2, we report the results for Level 1 only. There is a significant and positive coefficient for the amount of information on a project’s flaws (coefficient = 1.964; p < .001), which indicates that the greater the information on a project’s flaws, the more willing the project team member is to voice concerns about the project. This finding provides support for Hypothesis 1. The results also indicate a significant and negative coefficient for supervisor’s identification with the project (coefficient = –.258; p < .001). That is, the more the supervisor identifies with the project, the less willing the project team member is to voice concerns about the project. The coefficients for the cause of the project’s flaws and the project team member’s identification with the project are not statistically significant (p > .10). Now we turn to the results for the cross-level relationships, which test Hypotheses 2 through 4.

Table 2

Hierarchical Linear Modeling Results of Willingness to Voice Project Concerns (Level 1)

Variables	Coefficients(Std. Errors)
Intercept	4.54***
Intercept	(.06)
Cause of project’s flaws	−.02
Cause of project’s flaws	(.02)
Manager’s identification with the project	−.26***
Manager’s identification with the project	(.05)
Personal identification with the project	−.15^†
Personal identification with the project	(.08)
Amount of information on project’s fatal flaws	1.96***
Amount of information on project’s fatal flaws	(.10)
Model χ²	1,929.10***

Note: n = 3,760 decisions nested in 235 R&D project team members.

†

p < .10.

***

p < .001.

In Table 3, we report the results for the impact of Level 2 variables on the decision policy as captured in Level 1. In the first set of rows, we report the overall intercept and the intercepts for each of the variables at Level 1. In the second set of rows, we report on the moderating role of the Level 2 independent variable—amount of information on project’s flaws. Each model is represented by a control-variables-only model and a full model. We calculated the Pseudo R² at Level 1 and Level 2 as described by Snijders and Bosker (1999). For the base model, the Pseudo R² at Level 1 was 59.1%, and at Level 2 it was 1.0%. For the full model, the Pseudo R² at Level 1 was 60.4%, and at Level 2 it was 6.1%. The results of the hypotheses testing are in the last column. The Level 2 variables explain variance in the coefficient for each of these attributes in the project team members’ willingness to voice concerns over the project.

Table 3

Results of Explaining Variance Across Project Team Members in Decision Making on Voicing Project Concerns

	Overall Intercept		Amount of Information on Project’s Flaws
Variables	BaseCoefficient (Std. Errors)	FullCoefficient(Std. Errors)	BaseCoefficient(Std. Errors)	FullCoefficient(Std. Errors)
Overall intercept	4.17***	2.34***	4.17***	2.34***
	(.21)	(.70)	(.21)	(.70)
Cause of project’s flaws	−.01	−.34	−.01	−.34
	(.13)	(.23)	(.13)	(.23)
Supervisor’s identification with the project	−.46	.25	−.46	.25
	(.24)	(.41)	(.24)	(.41)
Personal identification with the project	−.47	.01	−.47	.01
	(.40)	(.83)	(.40)	(.83)
Amount of information on project’s fatal flaws	2.30***	−1.06	2.30***	−1.06
	(.45)	(.81)	(.45)	(.81)
Age	.01	.01	.01	−.02
	(.01)	(.01)	(.01)	(.01)
Gender	.39*	.46*	−.37	−.24
	(.19)	(.19)	(.39)	(.37)
Managerial open-mindedness		.03		.07*
		(.03)		(.03)
Prosocial motivation		.03		.05*
		(.02)		(.02)
Affective commitment		.01		.01
		(.01)		(.01)

Note: n = 3,760 decisions nested in 235 project team members.

p < .05.

***

p < .001.

The results for the “amount of information on project’s flaw” model indicate a significant and positive coefficient for supervisor open-mindedness (coefficient = .068; p < .05) and prosocial motivation (coefficient = .053; p < .05). The result for open-mindedness indicates that the positive relationship between the amount of objective information about a project’s flaws and the project team member’s willingness to voice concerns over that project is greater when the supervising manager is perceived as more open-minded than less open-minded (see Figure 2). This finding supports Hypothesis 2. The result for prosocial motivation indicates that the positive relationship between the amount of objective information about a project’s flaws and the project team member’s willingness to voice concerns over that project is greater when the project team member is more prosocially motivated than less prosocially motivated (see Figure 3). This finding supports Hypothesis 3. There is a nonsignificant coefficient for affective commitment, and thus Hypothesis 4 is not supported.

Figure 2

The Moderating Role of Managerial Openness on the Relationship Between the Amount of Objective Information About a Project’s Flaws and the Project Team Member’s Willingness to Voice Concerns Over That Project

Figure 3

The Moderating Role of Prosocial Motivation on the Relationship Between the Amount of Objective Information About a Project’s Flaws and the Project Team Member’s Willingness to Voice Concerns Over That Project

Discussion

In this study, we explored project team members’ decisions to voice concerns over the viability of an innovation project to the supervising manager. Using a conjoint experiment and data on 3,760 decisions nested within 235 individuals, we found that the amount of objective information on the project’s flaw was positively related to the willingness to voice concerns, but that this effect depended on the managers’ open-mindedness and the team member’s prosocial motivation. These findings have implications for the literatures on voice, prosocial motivation, and the management of innovation and corporate entrepreneurship projects.

Implications for the Literature on Voice

Recent reviews of the voice literature have lamented the lack of research on the role of the information employees possess because the likelihood of engaging in voice and the effectiveness of voice are likely dependent on the nature of the information employees possess (Morrison, 2011, 2014). We show that the more objective information about the project flaws that the team member possesses, the more likely she or he will be to engage in voice. Thus, we contribute to the voice literature by demonstrating an important new factor in employees’ likelihood to engage in voice.

The present study’s results advance our understanding not only of what causes employees to engage in voice, but also about when these factors have a stronger or weaker influence on the decision to engage in voice (and why). Specifically, although team members are more likely to voice concerns when they possess more objective information about a project’s flaws, the amount of information interacts with both the individual’s attributes (prosocial motivation) and his or her perception of the supervisor’s attributes (managerial openness) to influence the likelihood of voice. These cross-level moderating relationships answer calls to attend to the effects of information on the likelihood of engaging in voice (Morrison, 2014) and highlight a need for greater consideration of the concomitant (rather than independently) role of information, decision maker, and supervisor characteristics in future studies on voice.

Although care must be taken not to make too much of nonsignificant findings, we do point out that despite expectations based on our theorizing, affective commitment did not have a significant main-effect relationship with voice or change the emphasis on the amount of information in determining voice. This finding of a nonsignificant main-effect relationship for affective commitment on voice is consistent with Burris, Detert, and Chiaburu’s (2008) finding that psychological safety—measured as affective commitment—was not a direct predictor of a project team member’s voice to supervisors. It could be that the personal calculation in determining voice (Detert & Edmondson, 2011; Liang et al., 2012) is not influenced by affect or by an affective connection to the organization because, for example, there are more proximal affective connections (e.g., to the project and/or the team) that overpower the affective commitment to the organization. Of course, this proposition requires further theorizing and testing.

Further, we found a nonsignificant main-effect relationship between the project team members’ perception of their supervisor’s open-mindedness and voice. While a simple application of psychological safety for the project team member would be that an open-minded manager would lower the bar for voicing concerns (e.g., Detert & Burris, 2007; Dutton & Ashford, 1993; Nembhard & Edmondson, 2006; Takeuchi et al., 2012), this argument was not supported by our results. However, importantly, perceived managerial openness does play a role in project team members’ voicing concerns over a project’s flaws when we consider it in conjunction with the amount of information about the project’s flaws. Interestingly, the importance of the amount of information in encouraging voice is greater for those who perceive their managers as open-minded. It seems like the psychological safety to voice concerns requires both evidence and an open-minded manager. Indeed, the amount of information has little effect on voice for those who perceive their managers as close-minded. An implication is that efforts to train and recruit open-minded managers (which itself is a challenging task [Ribeiro-Soriano & Urbano, 2010]) will have little impact unless the project team also develops and collects considerable data on a project’s performance; only then is negative project performance information likely to be shared. The flipside of the same implication is that efforts to develop and collect information about a project’s (poor) performance (Cooper, 2008; Griffin, 1997) will not do much good for the organization if its managers are perceived by project team members as close-minded because that critical information is unlikely to be shared. Therefore, instead of emphasizing the overall benefits of managerial openness for triggering voice (Detert & Burris, 2007; Mowbray et al., 2015), our study highlights the boundary conditions of these benefits.

Finally, voice is typically investigated using a survey. For example, Detert and Burris (2007) use a three-item measure (from Van Dyne’s six-item prosocial voice measure) with a 5-point frequency scale (1 = almost never; 5 = almost always), for example, “I give suggestions [to the General Manager] about how to make this restaurant better” (see also Burris et al., 2016). By being the first to investigate voice using a metric conjoint approach, we provide a methodological approach to exploring the decision criteria underlying voice and the ways the “weights” of those decision criteria can vary across individuals and contexts. Although conjoint analysis has its own limitations (as detailed in the methods section), such an approach avoids some of the issues associated with self-reporting and retrospective data (Priem & Harrison, 1994). The benefits of conjoint analysis may be particularly salient in investigating implicit voice theories. Implicit voice theories are “taken for granted beliefs about when and why speaking up at work is risky or inappropriate” (Detert & Edmondson, 2011: 461) and because they “often operate below consciousness” (Detert & Edmondson, 2011: 463) there are advantages in capturing theories-in-action rather than relying on espoused theories. Conjoint analysis is well-suited to capturing theories in action (Priem, 1992, 1994). For example, one implicit voice theory identified by Detert and Edmondson (2011: 467) was “the perceived need to have solid data, polished ideas, or complete solutions before it is safe to speak.” The present study used conjoint analysis to investigate the degree to which this idiosyncratic belief about requiring objective information about project flaws influenced willingness to voice project concerns, and we demonstrated that this relationship depends on factors such as managerial openness. Future research could use conjoint analysis to systematically vary attributes related to other implicit voice theories (see Detert & Edmonson, 2011: acceptability of speaking up to bosses in the presence of others and/or the negative career consequences for speaking up). We hope this different methodological approach continues to open up new research questions for advancing our knowledge of the when, how, and why of voice in general, and in voicing project concerns specifically.

Implications for the Prosocial Motivation Literature

The finding of a significant moderating effect of prosocial motivation also offers new insights on the role of prosocial motivation in the context of innovation. The management literature has demonstrated that prosocial motivation triggers the development of novel ideas either directly (Bechtold, De Dreu, Nijstad, & Choi, 2010) or by either channeling employees’ intrinsic motivation toward creating new ideas that benefit others within or outside the organization (Grant & Berry, 2011) or the creation of social ventures (Miller, Grimes, McMullen, & Vogus, 2012). Therefore, these studies have focused on the role of prosocial motivation in promoting desirable outcomes (e.g., creating new ideas or initiating social ventures). Our results extend this focus to the role of prosocial motivation for preventing undesirable outcomes—namely, minimizing the costs of potentially flawed projects for the organization by voicing concerns about these projects such that they are terminated sooner.

Indeed, the focus on cost minimization suggests a potential downside of prosocial behavior. While it appears that prosocial motivation enhances employees’ tendency to take action (i.e., engage in voice) to prevent undesirable outcomes (direct effect of prosocial motivation), a high tendency to act based on prosocial motivation seems to require considerable information about the nondesirability of the outcome. The danger is that prosocially motivated employees do not engage in voice when they detect weak or ambiguous signals about threats of undesirable outcomes, thereby increasing the likelihood of persistence and the further “wasting” of resources by continued investments in failing projects.

Implications for the Innovation and Corporate Entrepreneurship Literatures

The innovation and corporate entrepreneurship literatures highlight the importance of quickly noticing, interpreting, and terminating poorly performing projects and redeploying resources that show promise but also acknowledge heterogeneity in firms’ ability to do so (Bakker & Shepherd, in press; Brown & Eisenhardt, 1997; McGrath et al., 2004). Our study offers a number of new insights into the termination of poorly performing projects by focusing on project team members’ willingness to voice concerns over their project’s flaws.

First, for firms to notice and act on problems (e.g., a project’s flaws), team members must both detect and report these problems to the managers responsible for decision making (Ocasio, 1997, 2011). In this study, we focus on project team members’ reporting of concerns over projects to their supervisors. Although much has been made of why it is important to collect evidence of an innovation project’s performance (Cooper, 2008; Griffin, 1997) and how disconfirming evidence is often either ignored or discounted (Ng et al., 2009; Tripsas & Gavetti, 2000), our results show that the sharing of information is critical and depends on the project team member’s (i.e., information holder’s) perception of her or his supervisor as well as her or his prosocial motivation.

Second, from a managerial perspective studies in the corporate entrepreneurship literature exploring project termination decisions have focused on information content (e.g., Behrens & Patzelt, 2016; Green, Welsh, & Dehler, 2003) and how managerial characteristics (e.g., Corbett, Neck, & DeTienne, 2007) explain the incorporation of this content into their decision policies. With an investigation of team members’ voicing of project performance concerns, we shift the focus away from how managers decide and act on information about poorly performing projects toward their access to information about project performance from those who know it best (i.e., the project team members). While we did not find that managers’ higher identification with the project substantially diminished team members’ willingness to voice their concerns, our results identified managers’ open-mindedness as a moderator in the relationship between available information about a project’s flaw and team members’ willingness to voice concerns. That is, the way managers are perceived by project team members appears to have an important impact on the project termination decision by either facilitating or diminishing the gathering of information about project flaws (i.e., through voice).

Finally, there is also a conversation in the corporate entrepreneurship literature about the role of project team members during project failures. In this context, studies have focused on the emotional experiences of failure, finding that team members often experience negative emotions from project failure, which can be a substantial obstacle to learning from the experience (Edmondson, 1999; Moenkemeyer, Hoegl, & Weiss, 2012; Shepherd, Covin, & Kuratko, 2009; Shepherd, Patzelt, & Wolfe, 2011; Shepherd, Patzelt, Williams, & Warnecke, 2014). While these studies have provided important insights into how team members experience project failure during or after the termination event, it was not their purpose to investigate the role of team members before the termination event. In the current study we theorize and find that team members are more or less inclined to contribute to the initiation of project termination by voicing their concerns about the project based on available evidence for flaws, and that this heterogeneity is (partly) explained by team members’ personal characteristics (prosocial motivation) and their perception of the project supervisor (open-mindedness). These results emphasize the need to study the role of team members in project termination before it actually takes place and, in particular, their role in anticipating failure and informing the project manager about it.

Limitations and Future Research

This study also highlights a number of remaining questions that provide opportunities for future voice research. First, there is a distinction between promotive (i.e., making suggestions for improvement) and prohibitive (i.e., pointing out problems) voice (e.g., Liang et al., 2012). The present study clearly focuses on prohibitive voice, which begs the question of whether results generalize to promotive voice. We suggest results are likely to be similar for promotive voice as having more objective information about areas for improvement would likely make employees more confident (i.e., perceive less riskiness and greater instrumentality) about providing suggestions. Of course, this is an empirical question for future research.

Second, the present study focuses on what explains engagement in voice but does not directly address the effectiveness of voice. Employees are probably more apt to engage in voice when they possess more objective information about project flaws because they perceive voice as more likely to be effective in such instances. However, future research can investigate whether voicing concerns about a project based on more information really is more effective (e.g., better outcomes for the employee, team, project, organization). Given research demonstrating that voice is better received when the message is framed positively rather than negatively (Whiting, Maynes, Podsakoff, & Podsakoff, 2012), it may be fruitful to consider whether this framing effect might interact with the amount of information about the project’s flaws. For instance, perhaps with more information the team member can frame his or her concerns more positively (versus negative framing), thereby enhancing the effectiveness of voice (especially prohibitive voice).

Finally, there are some limitations to the current study that provide opportunities for future research. Specifically, we explored the situation of voicing concerns to the supervisor at a team meeting or in person. However, it is possible that the decision policy for voicing concerns about a project to the supervisor may be different in the presence or absence of others. Future research can explore this “social” aspect to the decision to voice project concerns. Furthermore, although we somewhat control for the project team member’s investment in the project through the decision attribute of “personal identification with the project” (at Level 1), future research could further explore the extent to which personal investments (perhaps contributing to the sunk cost or endowment effect) and/or differences across investment types (time, money, and other resources) influence the likelihood of voicing concerns.

Conclusion

Project team members’ voicing of concerns about a project’s flaw can help terminate underperforming projects and thus help entrepreneurial firms save valuable resources. However, there is considerable variance in project team members’ likelihood to voice their concerns. In this article, we developed and empirically tested a model of voicing concerns over a project’s flaws based on the amount of information available to project team members. We found that more information about a project’s flaw increases project team members’ willingness to voice concerns and that this positive effect is stronger for project team members who believe that they have more open-minded supervisors and who themselves are more prosocially motivated. This theorizing and set of findings provides new insights into our understanding of managing innovation and entrepreneurial projects and facilitating voice in organizations.

Footnotes

Acknowledgements

This article was accepted under the editorship of Patrick M. Wright. The authors would like to thank Bill Schulze (editor) and three anonymous reviewers for their valuable comments.

Notes

References

Adner

Levinthal

D. A.

2004. What is not a real option: Considering boundaries for the application of real options to business strategy. Academy of Management Review, 29: 74-85.

Allen

N. J.

Meyer

J. P.

1990. The measurement and antecedents of affective, continuance and normative commitment to the organization. Journal of Occupational Psychology, 63: 1-18.

Allen

T. D.

Rush

M. C.

1998. The effects of organizational citizenship behavior on performance judgments: A field study and a laboratory experiment. Journal of Applied Psychology, 83: 247-260.

Armstrong

J. S.

Overton

T. S.

1977. Estimating nonresponse bias in mail surveys. Journal of Marketing Research, 14: 396-402.

Ashford

S. J.

Cummings

L. L.

1981. Strategies for knowing: When and from where do individuals seek feedback? Academy of Management Proceedings 1: 161-165.

Ashford

B. E.

Gibbs

B. W.

1990. The double-edge of organizational legitimation. Organization Science, 1: 177-194.

Ashford

S. J.

Rothbard

N. P.

Piderit

S. K.

Dutton

J. E.

1998. Out on a limb: The role of context and impression management in selling gender-equity issues. Administrative Science Quarterly, 43: 23-57.

Bakker

R. M.

Shepherd

D. A.

in press. Pull the plug or take the plunge: Multiple opportunities and the speed of venturing decisions in the Australian mining industry. Academy of Management Journal.

Bandura

Locke

E. A.

2003. Negative self-efficacy and goal effects revisited. Journal of Applied Psychology, 88: 87-99.

10.

Barsky

A. P.

Zyphur

M. J.

2016. Disentangling sunk-costs and completion proximity: The role of regulatory focus. Journal of Experimental Social Psychology, 65: 105-108.

11.

Bechtoldt

M. N.

De Dreu

C. K.

Nijstad

B. A.

Choi

H. S.

2010. Motivated information processing, social tuning, and group creativity. Journal of Personality and Social Psychology, 99: 622-637.

12.

Behrens

Patzelt

2016. Corporate entrepreneurship managers’ project terminations: Integrating portfolio-level, individual-level, and firm-level effects. Entrepreneurship: Theory and Practice, 40: 815-842.

13.

Belsley

D. A.

Kuh

Welsch

R. E.

1980. Regression diagnostics. New York, NY: John Wiley and Sons.

14.

Bolino

M. C.

Turnley

W. H.

2005. The personal costs of citizenship behavior: The relationship between individual initiative and role overload, job stress, and work-family conflict. Journal of Applied Psychology, 90: 740-748.

15.

Booth

Schulz

A. K. D.

2004. The impact of an ethical environment on managers’ project evaluation judgments under agency problem conditions. Accounting, Organizations and Society, 29: 473-488.

16.

Boulding

Guha

Staelin

2016, July 12. Do we really need to change the decision maker? Counterintuitive escalation of commitment results in real options contexts. Management Science.

17.

Bowman

E. H.

Hurry

1993. Strategy through the option lens: An integrated view of resource investments and the incremental-choice process. Academy of Management Review, 18: 760-782.

18.

Brown

S. L.

Eisenhardt

K. M.

1997. The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42: 1-34.

19.

Bryk

A. S.

Raudenbush

S. W.

1992. Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: SAGE Publications.

20.

Burris

E. R.

2012. The risks and rewards of speaking up: Managerial responses to employee voice. Academy of Management Journal, 55: 851-875.

21.

Burris

E. R.

Detert

J. R.

Chiaburu

D. S.

2008. Quitting before leaving: The mediating effects of psychological attachment and detachment on voice. Journal of Applied Psychology, 93: 912-922.

22.

Burris

E. R.

Detert

J. R.

Romney

A. C.

2013. Speaking up vs. being heard: The disagreement around and outcomes of employee voice. Organization Science, 24: 22-38.

23.

Chiaburu

D. S.

Berry

C. M.

Gardner

R. G.

2011. The relationship between Big Five personality traits and citizenship behaviors: A meta-analysis. Journal of Applied Psychology, 96: 1140-1166.

24.

Chiu

C. Y.

Morris

M. W.

Hong

Y. Y.

Menon

2000. Motivated cultural cognition: The impact of implicit cultural theories on dispositional attribution varies as a function of need for closure. Journal of Personality and Social Psychology, 78: 247-259.

25.

Cooper

R. G.

2008. Perspective: The Stage-Gate® idea-to-launch process—Update, what’s new, and NexGen systems. Journal of Product Innovation Management, 25: 213-232.

26.

Cooper

R. G.

Sommer

A. F.

2016. The Agile–Stage-Gate hybrid model: A promising new approach and a new research opportunity. Journal of Product Innovation Management, 33: 513-526.

27.

Corbett

A. C.

Neck

H. M.

DeTienne

D. R.

2007. How corporate entrepreneurs learn from fledgling innovation initiatives: Cognition and the development of a termination script. Entrepreneurship Theory and Practice, 31: 829-852.

28.

Covin

J. G.

Garrett

R. P.

Kuratko

D. F.

Shepherd

D. A.

2015. Value proposition evolution and the performance of internal corporate ventures. Journal of Business Venturing, 30: 749-774.

29.

Decker

Mellewigt

2007. Thirty years after Michael E. Porter: What do we know about business exit? Academy of Management Perspectives, 21: 41-55.

30.

DeNisi

A. S.

Kluger

A. N.

2000. Feedback effectiveness: Can 360-degree appraisals be improved? Academy of Management Executive, 14: 129-139.

31.

Detert

J. R.

Burris

E. R.

2007. Leadership behavior and employee voice: Is the door really open? Academy of Management Journal, 50: 869-884.

32.

Detert

J. R.

Burris

E. R.

Harrison

D. A.

Martin

S. R.

2013. Voice flows to and around leaders understanding when units are helped or hurt by employee voice. Administrative Science Quarterly, 58: 624-668.

33.

Detert

J. R.

Edmondson

A. C.

2011. Implicit voice theories: Taken-for-granted rules of self-censorship at work. Academy of Management Journal, 54: 461-488.

34.

Detert

J. R.

Trevino

L. K.

2010. Speaking up to higher-ups: How supervisors and skip-level leaders influence employee voice. Organization Science, 21: 249-270.

35.

Drummond

2014. Escalation of commitment: When to stay the course? Academy of Management Perspectives, 28: 430-446.

36.

Dutton

J. E.

Ashford

S. J.

1993. Selling issues to top management. Academy of Management Review, 18: 397-428.

37.

Dutton

J. E.

Ashford

S. J.

O’Neill

R. M.

Hayes

Wierba

E. E.

1997. Reading the wind: How middle managers assess the context for selling issues to top managers. Strategic Management Journal, 18: 407-425.

38.

Edmondson

1999. Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44: 350-383.

39.

Edmondson

A. C.

2003. Speaking up in the operating room: How team leaders promote learning in interdisciplinary action teams. Journal of Management Studies, 40: 1419-1452.

40.

Elfenbein

D. W.

Knott

A. M.

2015. Time to exit: Rational, behavioral, and organizational delays. Strategic Management Journal, 36: 957-975.

41.

Farson

Keyes

2002. The failure-tolerant leader. Harvard Business Review, 80 (8): 64-71.

42.

Fast

N. J.

Burris

E. R.

Bartel

C. A.

2014. Managing to stay in the dark: Managerial self-efficacy, ego defensiveness, and the aversion to employee voice. Academy of Management Journal, 57: 1013-1034.

43.

Fredrickson

B. L.

1998. What good are positive emotions? Review of General Psychology, 2: 300-319.

44.

Fredrickson

B. L.

2001. The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions. The American Psychologist, 56: 218.

45.

Frese

Fay

2001. Personal initiative: An active performance concept for work in the 21st century. Research in Organizational Behavior, 23: 133-187.

46.

Fuller

J. B.

Marler

L. E.

Hester

2006. Promoting felt responsibility for constructive change and proactive behavior: Exploring aspects of an elaborated model of work design. Journal of Organizational Behavior, 27: 1089-1120.

47.

Graham

J. W.

Van Dyne

2006. Gathering information and exercising influence: Two forms of civic virtue organizational citizenship behavior. Employee Responsibilities and Rights Journal, 18: 89-109.

48.

Grant

A. M.

2007. Relational job design and the motivation to make a prosocial difference. Academy of Management Review, 32: 393-417.

49.

Grant

A. M.

2008. The significance of task significance: Job performance effects, relational mechanisms, and boundary conditions. Journal of Applied Psychology, 93: 48-58.

50.

Grant

A. M.

Berry

J. W.

2011. The necessity of others is the mother of invention: Intrinsic and prosocial motivations, perspective taking, and creativity. Academy of Management Journal, 54: 73-96.

51.

Grant

A. M.

Mayer

D. M.

2009. Good soldiers and good actors: Prosocial and impression management motives as interactive predictors of affiliative citizenship behaviors. Journal of Applied Psychology, 94: 900-912.

52.

Grant

A. M.

Parker

Collins

2009. Getting credit for proactive behavior: Supervisor reactions depend on what you value and how you feel. Personnel Psychology, 62: 31-55.

53.

Grant

A. M.

Sumanth

J. J.

2009. Mission possible? The performance of prosocially motivated employees depends on manager trustworthiness. Journal of Applied Psychology, 94: 927-944.

54.

Green

P. E.

Krieger

A. M.

Wind

2001. Thirty years of conjoint analysis: Reflections and prospects. Interfaces, 31 (3): S56-S73.

55.

Green

P. E.

Srinivasan

1990. Conjoint analysis in marketing: New developments with implications for research and practice. Journal of Marketing, 54 (4): 3-19.

56.

Green

Welsh

Dehler

2003. Advocacy, performance, and threshold influences on decisions to terminate new product development. Academy of Management Journal, 46: 419-434.

57.

Griffin

1997. The effect of project and process characteristics on product development cycle time. Journal of Marketing Research, 34: 24-35.

58.

Griffin

M. A.

Neal

Parker

S. K.

2007. A new model of work role performance: Positive behavior in uncertain and interdependent contexts. Academy of Management Journal, 50: 327-347.

59.

Hahn

G. J.

Shapiro

S. S.

1966, August 15. A cataloging and computer program for the design and analysis of orthogonal symmetric and asymmetric fractional factorial experiments (General Electric Research and Development Center). Presented at the 1966 Annual Convention of the American Statistical Association, Los Angeles, CA.

60.

Hair

J. F.

2010. Multivariate data analysis: A global perspective, 7th ed. Upper Saddle River, NJ: Prentice Hall.

61.

Hedstrom

Swedberg

1998. Social mechanisms: An analytic approach to social theory. New York, NY: Cambridge University Press.

62.

Hitt

M. A.

Middlemist

R. D.

1979. A methodology to develop the criteria and criteria weightings for assessing subunit effectiveness in organizations. Academy of Management Journal, 22: 356-374.

63.

Hull

C. E.

Rothenberg

2008. Firm performance: The interactions of corporate social performance with innovation and industry differentiation. Strategic Management Journal, 29: 781-789.

64.

Ilies

Scott

B. A.

Judge

T. A.

2006. The interactive effects of personal traits and experienced states on intraindividual patterns of citizenship behavior. Academy of Management Journal, 49: 561-575.

65.

Jacobsen

D. I.

2000. Managing increased part-time: Does part-time work imply part-time commitment? Managing Service Quality, 10: 187-201.

66.

Janney

J. J.

Dess

G. G.

2004. Can real-options analysis improve decision-making? Promises and pitfalls. The Academy of Management Executive, 18 (4): 60-75.

67.

Johnson

D. E.

Erez

Kiker

D. S.

Motowidlo

S. J.

2002. Liking and attributions of motives as mediators of the relationships between individuals’ reputations, helpful behaviors and raters’ reward decisions. Journal of Applied Psychology, 87: 808-815.

68.

Keats

B. W.

1991. An empirical investigation of strategic investment decision models. Strategic Management Journal, 12: 243-250.

69.

Klein

J. I.

1990. Feasibility theory: A resource-munificence model of work motivation and behavior. Academy of Management Review, 15: 646-665.

70.

Klingebiel

Adner

2015. Real options logic revisited: The performance effects of alternative resource allocation regimes. Academy of Management Journal, 58: 221-241.

71.

Kluger

A. N.

DeNisi

1996. The effects of feedback interventions on performance: A historical review, a meta-analysis, and a preliminary feedback intervention theory. Psychological Bulletin, 119: 254-284.

72.

Korsgaard

M. A.

Roberson

Rymph

R. D.

1998. What motivates fairness? The role of subordinate assertive behavior on manager’s interactional fairness. Journal of Applied Psychology, 83: 731.

73.

Kutner

M. H.

Nachtsheim

Neter

2004. Applied linear regression models. Chicago, IL: McGraw-Hill/Irwin.

74.

Lee

S. H.

Peng

M. W.

Barney

J. B.

2007. Bankruptcy law and entrepreneurship development: A real options perspective. Academy of Management Review, 32: 257-272.

75.

LePine

J. A.

Van Dyne

1998. Predicting voice behavior in work groups. Journal of Applied Psychology, 83: 853.

76.

Atuahene-Gima

2001. Product innovation strategy and the performance of new technology ventures in China. Academy of Management Journal, 44: 1123-1134.

77.

Liang

Farh

C. I.

Farh

J. L.

2012. Psychological antecedents of promotive and prohibitive voice: A two-wave examination. Academy of Management Journal, 55: 71-92.

78.

Lohrke

F. T.

Holloway

B. B.

Woolley

T. W.

2010. Conjoint analysis in entrepreneurship research: A review and research agenda. Organizational Research Methods, 13: 16-30.

79.

Louviere

J. J.

1988. Analyzing decision making: Metric conjoint analysis (No. 67). Newbury Park, CA: SAGE Publications.

80.

Luchak

A. A.

2003. What kind of voice do loyal employees use? British Journal of Industrial Relations, 41: 115-134.

81.

Marchington

Suter

2013. Where informality really matters: Patterns of employee involvement and participation (EIP) in a non-union firm. Industrial Relations: A Journal of Economy and Society, 52: 284-313.

82.

McClean

E. J.

Burris

E. R.

Detert

J. R.

2013. When does voice lead to exit? It depends on leadership. Academy of Management Journal, 56: 525-548.

83.

McGrath

1999. Falling forward: Real options reasoning and entrepreneurial failure. Academy of Management Review, 24: 13-30.

84.

McGrath

R. G.

Ferrier

W. J.

Mendelow

A. L.

2004. Real options as engines of choice and heterogeneity. Academy of Management Review, 29: 86-101.

85.

McKelvie

Haynie

J. M.

Gustavsson

2011. Unpacking the uncertainty construct: Implications for entrepreneurial action. Journal of Business Venturing, 26: 273-292.

86.

Meglino

B. M.

Korsgaard

2004. Considering rational self-interest as a disposition: Organizational implications of other orientation. Journal of Applied Psychology, 89: 946-959.

87.

Meyer

J. P.

Allen

N. J.

1984. Testing the “side-bet theory” of organizational commitment: Some methodological considerations. Journal of Applied Psychology, 69: 372-378.

88.

Meyer

J. P.

Allen

N. J.

1991. A three-component conceptualization of organizational commitment. Human Resource Management Review, 1: 61-89.

89.

Meyer

J. P.

Allen

N. J.

1997. Commitment in the workplace: Theory, research, and application. Thousand Oaks, CA: SAGE Publications.

90.

Miller

T. L.

Grimes

M. G.

McMullen

J. S.

Vogus

T. J.

2012. Venturing for others with heart and head: How compassion encourages social entrepreneurship. Academy of Management Review, 37: 616-640.

91.

Milliken

F. J.

Morrison

E. W.

Hewlin

P. F.

2003. An exploratory study of employee silence: Issues that employees don’t communicate upward and why. Journal of Management Studies, 40: 1453-1476.

92.

Moenkemeyer

Hoegl

Weiss

2012. Innovator resilience potential: A process perspective of individual resilience as influenced by innovation project termination. Human Relations, 65: 627-655.

93.

Morrison

E. W.

2011. Voice and silence within organizations: Literature review and directions for future research. Academy of Management Annals, 5: 373-412.

94.

Morrison

E. W.

2014. Employee voice and silence. Annual Review of Organizational Psychology and Organizational Behavior, 1: 173-197.

95.

Morrison

E. W.

Milliken

F. J.

2000. Organizational silence: A barrier to change and development in a pluralistic world. Academy of Management Review, 25: 706-725.

96.

Mowbray

P. K.

Wilkinson

Tse

H. H.

2015. An integrative review of employee voice: Identifying a common conceptualization and research agenda. International Journal of Management Reviews, 17: 382-400.

97.

Mowday

R. T.

Porter

L. W.

Steers

1982. Organizational linkages: The psychology of commitment, absenteeism, and turnover. San Diego, CA: Academic Press.

98.

Nembhard

I. M.

Edmondson

A. C.

2006. Making it safe: The effects of leader inclusiveness and professional status on psychological safety and improvement efforts in health care teams. Journal of Organizational Behavior, 27: 941-966.

99.

Westgren

Sonka

2009. Competitive blind spots in an institutional field. Strategic Management Journal, 30: 349-369.

100.

Ocasio

1997. Towards an attention-based view of the firm. Strategic Management Journal, 18: 187-206.

101.

Ocasio

2011. Attention to attention. Organization Science, 22: 1286-1296.

102.

Pablo

A. L.

1994. Determinants of acquisition integration level: A decision-making perspective. Academy of Management Journal, 37: 803-836.

103.

Peng

M. W.

Yamakawa

Lee

S. H.

2010. Bankruptcy laws and entrepreneur-friendliness. Entrepreneurship Theory and Practice, 34: 517-530.

104.

Priem

R. L.

1992. An application of metric conjoint analysis for the evaluation of top managers’ individual strategic decision making processes. Strategic Management Journal, 13: 143-151.

105.

Priem

R. L.

1994. Executive judgment, organizational congruence, and firm performance. Organization Science, 5: 421-437.

106.

Priem

R. L.

Harrison

D. A.

1994. Exploring strategic judgment: Methods for testing the assumptions of prescriptive contingency theories. Strategic Management Journal, 15: 311-324.

107.

Redding

W. C.

1985. Rocking boats, blowing whistles, and teaching speech communication. Communication Education, 34: 245-258.

108.

Ribeiro-Soriano

Urbano

2010. Employee-organization relationship in collective entrepreneurship: An overview. Journal of Organizational Change Management, 23: 349-359.

109.

Roy

Sarkar

M. B.

2015. Knowledge, firm boundaries, and innovation: Mitigating the incumbent’s curse during radical technological change. Strategic Management Journal, 37: 835-854.

110.

Royer

2003. Why bad projects are so hard to kill. Harvard Business Review, 81 (2): 48-56.

111.

Schilit

W. K.

1987. An examination of the influence of middle-level managers in formulating and implementing strategic decisions. Journal of Management Studies, 24: 271-293.

112.

Schwartz

Wald

M. L.

2003, March 9. NASA’s curse? Groupthink is 30 years old, and still going strong. New York Times.

113.

Sharp

D. J.

Salter

S. B.

1997. Project escalation and sunk costs: A test of the international generalizability of agency and prospect theories. Journal of International Business Studies, 28: 101-121.

114.

Shepherd

D. A.

Covin

J. G.

Kuratko

D. F.

2009. Project failure from corporate entrepreneurship: Managing the grief process. Journal of Business Venturing, 24: 588-600.

115.

Shepherd

D. A.

Patzelt

Baron

R. A.

2013. “I care about nature, but…”: Disengaging values in assessing opportunities that cause harm. Academy of Management Journal, 56: 1251-1273.

116.

Shepherd

D. A.

Patzelt

Williams

T. A.

Warnecke

2014. How does project termination impact project team members? Rapid termination, “creeping death,” and learning from failure. Journal of Management Studies, 51: 513-546.

117.

Shepherd

D. A.

Patzelt

Wolfe

2011. Moving forward from project failure: Negative emotions, affective commitment, and learning from the experience. Academy of Management Journal, 54: 1229-1259.

118.

Shepherd

D. A.

Wiklund

Haynie

J. M.

2009. Moving forward: Balancing the financial and emotional costs of business failure. Journal of Business Venturing, 24: 134-148.

119.

Sleesman

D. J.

Conlon

D. E.

McNamara

Miles

J. E.

2012. Cleaning up the big muddy: A meta-analytic review of the determinants of escalation of commitment. Academy of Management Journal, 55: 541-562.

120.

Smidts

Pruyn

A. T. H.

Van Riel

C. B.

2001. The impact of employee communication and perceived external prestige on organizational identification. Academy of Management Journal, 44: 1051-1062.

121.

Snijders

T. B.

Bosker

R. R.

1999. Multilevel analysis: An introduction to basic and advanced multilevel modeling. London: SAGE Publications.

122.

Staw

B. M.

Ross

1989. Understanding behavior in escalation situations. Science, 246: 216-220.

123.

Takeuchi

Chen

Cheung

S. Y.

2012. Applying uncertainty management theory to employee voice behavior: An integrative investigation. Personnel Psychology, 65: 283-323.

124.

Tangirala

Kamdar

Venkataramani

Parke

2013. Doing right versus getting ahead: The effects of duty and achievement orientation on employees’ voice. Journal of Applied Psychology, 98: 1040-1050.

125.

Tangirala

Ramanujam

2008. Employee silence on critical work issues: The cross level effects of procedural justice climate. Personnel Psychology, 61: 37-68.

126.

Tangirala

Ramanujam

2012. Ask and you shall hear (but not always): Examining the relationship between manager consultation and employee voice. Personnel Psychology, 65: 251-282.

127.

Thornhill

2006. Knowledge, innovation and firm performance in high- and low-technology regimes. Journal of Business Venturing, 21: 687-703.

128.

Townsend

Wilkinson

Burgess

2013. Filling the gaps: Patterns of formal and informal voice. Economic and Industrial Democracy, 34: 337-354.

129.

Tripsas

Gavetti

2000. Capabilities, cognition, and inertia: Evidence from digital imaging. Strategic Management Journal, 21: 1147-1161.

130.

Tröster

van Knippenberg

2012. Leader openness, nationality dissimilarity, and voice in multinational management teams. Journal of International Business Studies, 43: 591-613.

131.

Tyler

B. B.

Steensma

K. H.

1995. Evaluating technological collaborative opportunities: A cognitive modeling perspective. Strategic Management Journal, 16 (S1): 43-70.

132.

Unger

B. N.

Kock

Gemünden

H. G.

Jonas

2012. Enforcing strategic fit of project portfolios by project termination: An empirical study on senior management involvement. International Journal of Project Management, 30: 675-685.

133.

Van Dyne

Ang

Botero

I. C.

2003. Conceptualizing employee silence and employee voice as multidimensional constructs. Journal of Management Studies, 40: 1359-1392.

134.

Van Dyne

Cummings

L. L.

Parks

J. M.

1995. Extra-role behaviors: In pursuit of construct and deﬁnitional clarity (a bridge over muddied waters). In Cummings

L. L.

Staw

B. M.

(Eds.), Research in Organizational Behavior, 17: 215-285. Greenwich, CT: JAI Press.

135.

Van Dyne

LePine

J. A.

1998. Helping and voice extra-role behaviors: Evidence of construct and predictive validity. Academy of Management Journal, 41: 108-119.

136.

Whiting

S. W.

Maynes

T. D.

Podsakoff

N. P.

Podsakoff

P. M.

2012. Effects of message, source, and context on evaluations of employee voice behavior. Journal of Applied Psychology, 97: 159-182.

137.

Zapata-Phelan

C. P.

Colquitt

J. A.

Scott

B. A.

Livingston

2009. Procedural justice, interactional justice, and task performance: The mediating role of intrinsic motivation. Organizational Behavior and Human Decision Processes, 108: 93-105.

138.

Zhang

Bartol

K. M.

2010. Linking empowering leadership and employee creativity: The influence of psychological empowerment, intrinsic motivation, and creative process engagement. Academy of Management Journal, 53: 107-128.