UNIVERSITEIT VAN AMSTERDAM
Voice Behavior
The Influence of Social Anxiety and Attention Control
Cijfer: 8
Marlijn Kruitwagen Student number: 10215417
Supervisor: Inge Wolsink Word Count: 8995
Abstract
Previous studies address that attention seems to have no influence on voice versus silence (Morrison, 2014; Wolsink, 2014; Kish-Gephart, Detert, Trevino and Edmondson, 2009) but could have an influence on voice frequency (Parker, Bindl & Strauss, 2010; Grant & Ashford, 2008). Implicit voice theories account that social anxiety would negatively influence voice (Detert and Edmonson, 2011). Thereby, low attention control could strengthen this effect (Baddeley & Hitch, 2003; Moriya and Sugiura, 2013). To investigate what the relationship between social anxiety, attention control and voice versus silence and voice frequency is, an empirical and experimental study with mixed repeated measures ANOVA is conducted. Students (N=49, 60% male, 38% female) were asked to generate creative ideas for improvements of the faculty. Participants got the opportunity to voluntary voice their ideas towards the leader, while their attention control was manipulated. The manipulation was created through two different counting down tasks, which entailed two levels of difficulty. As expected, attention control did not predict voice versus silence. However, attention control also did not predict voice frequency, showing that voice frequency might also be automatically processed (Wolsink, 2014). Social anxiety seems not to show substantial differences over both
conditions. The strengthen effect of low attention control on the voice frequency of social anxious employees was also not confirmed. This shows that apparently it does not matter how social anxious or distracted you are, as long as you are in a pro-social climate.
1 Introduction
A couple of years ago, I worked for the notary office MG. De Jong in the east of the
Netherlands. This company was growing for years and years, resulting in highly specialized departments. Hierarchy was an influential factor in this company, because managers felt predominate over employees. Although they expected employees to show initiative, they often expressed themselves in negative ways. We expressed our ideas, but the negative reactions influenced our fear. This company ended up bankrupt, because employees became ignorant of their speaking behavior. Current literature about upward communication supports this view: employees are expected to show initiative, express their ideas and make
suggestions for changes. However, if supervisors are judgmental and employees are sensitive for negative reactions, employees will remain silent (Morrison & Milliken, 2008). Detert, Burris, Harrison and Martin (2000) examined this speaking up behavior and found a positive association between voice and organization’s effectiveness and financial performance in businesses today. Furthermore, recent research emphasizes that withholding of input can be harmful and dysfunctional for organizations (Crant, 2000; Campbell, 2000; LePine & Van Dyne, 1998, 2001; Morrison, 2014). In literature, silence is defined as the withholding of speaking up towards a leader. It is not simply the lack of speech, but more the lack of sharing suggestions, concerns and information that could be useful or relevant to share (Milliken, Morrison & Hewlin, 2003; Van Dyne et al. 2003). So when an employee is silent it is assumed that the employee knows something but does not share this. Different factors and motivational processes influence this silence, such as feeling responsible for constructive change, sense of obligation and conscientiousness (Morrison, 2014). This study focuses on the factor of social fear and aims to find out how social fear influences speaking up behavior towards leaders.
Previous research suggests that expressing ideas and making efforts for changes is a form of proactive behavior (Grant & Ashford, 2008). Whereby, voice is seen as type of proactive behavior and described as talking constructively about ideas, suggestions or opinions of others with the intent to bring improvement and challenge the status quo (Morrison, 2011; Van Dyne & LePine, 1998; Van Dyne, Ang & Botero, 2003). The action component of voice, also known as voice frequency, defines how often people voice (Wolsink, 2014). When employees decide to remain silent, you can wonder if they fail to voice their suggestions. Because, how can you know for sure that employees withhold input or simply do not possess this input? That is why this study explores if employees can generate ideas and decide to voice them or not.
Previous work examined social anxiety as a personal characteristic. Moriya and Sugiura (2013) suggest that social anxiety influences attention control. Attention control depends on the capacity to keep information mentally active; temporarily storage information and manipulating this in a systematic manner (Baddeley & Hitch, 2003). If you are highly capable of controlling your attention and for that manner keep goals active, you will have a higher focus and filter out distractions. Moriya and Sugiura (2013) found that socially anxious people get easily distracted, unless they have a high capacity to control their attention. The ability of an employee to keep certain tasks relevant and active in your memory is regulated through their Working Memory Capacity (WMC) (Baddeley & Hitch, 2003). Moriya and Sugiura (2013) divided WMC in high and low and found that socially anxious people with low WMC are more distracted to irrelevant non-emotional stimuli, such as visual information. Thereby, earlier suggestions explaining that employees who are scared to obtain a negative evaluation of their judgmental supervisor or not feeling responsible or obligated to voice, will remain silent (Morrison, 2014). Also, according to Moriya and Sugiura (2013), if the
of social anxiety. If you distract attention, which creates low attention control, social anxiety should be of maximum influence. This could result in the remaining of silence or lower frequency of voicing. That is why I examine if differences in attention control influences social anxious employees in their voice behavior (Moriya & Sugiura 2013; Wolsink, 2014; Morrison, 2014).
I divide my study in two ways to examine if attention control will influence voice versus silence or voice frequency. First, I examine the relationship between social anxiety, attention control and voice versus silence. I wonder if the implicit voice theory of Detert and Edmonson (2011) will be confirmed. According to this theory it seems likely that employees, who are socially anxious or have implicit theories about speaking up in public, will voice less than people who do not own this. This theory is based on the assumption that speaking up could result in undesired outcomes, such as damaging of reputation, reduced self-esteem or negative work evaluation, which would result in the remaining of silence of social anxious employees. The employee may automatically retreat without any careful consideration of pros and cons of speaking up (Morrison, 2014). Furthermore, Wolsink (2014) indicates that
attention control does not influence the relationship between voice versus silence, because voice versus silence is unconsciously decided. Also Kish-Gephart, Detert, Trevino and Edmondson (2009) confirm this view that intentionally not voicing (being silent) might be caused by non-deliberative and schema driven processes that are triggered by basic emotions such as fear.
Besides, I wonder if attention control influences the amount of suggestions (voice frequency). So, once an employee decides to speak up, whether distraction influences the quantity of speaking up. Two contradicting views are found in the literature. Some scholars suggest that voice frequency is also an unconsciously process and attention control will not influence the frequency of speaking (Wolsink, 2014; Detert & Edmonson, 2011). On the other
hand, scholars suggest that voice frequency and proactive action is part of your conscious, slow, deliberative and goal direct process (Parker, Bindl & Strauss, 2010) and it is voluntarily chosen and initiated from within (Grant & Ashford, 2008). From that point of view, it can be assumed that voice frequency is positively influenced through a higher attention control and negatively influenced through a lower attention control. Especially in the case of a social anxious employee, who already fears the reaction of their supervisor. If you distract attention they are less capable to cope with the negative consequences of social anxiety and it
maximizes the effect of social anxiety. This is why this study aims to find:
What is the relationship between social anxiety, attention control and voice versus silence and voice frequency?
First, I investigate the influence of social anxiety on expressing voice. Secondly, I address how cognitive ability plays a role in both social anxiety and the communication of suggestions and at the same time active goal maintenance. My contribution to organizations is embedded in the experimental study. So far, attention control and its influence on voice behavior is mainly researched in theoretical manner, leaving a gap for empirical research. This paper aims to show that keeping information active in memory with controlled focus could benefit employees in such a way that they will voice and their voice frequency will increase, especially when considering social anxious employees. This is done through a mixed design of empirical en experimental research, which contributes to the theoretical research of the influence of attention control on voice behavior. This can help organizations to improve human capital effectiveness and increase the knowledge of creating a good environment for different employees in which they feel comfortable enough to express themselves in upward communication towards a leader.
2 Theoretical Framework
Proactive Voice Behavior
Nowadays, organizations hire more employees who take initiative in changing and improving current or future situations rather than just adapt or react to present conditions (Crant, 2000). This proactive behavior (Campbell, 2000) is fundamental for the success and effectiveness of businesses today (Crant, 2000). Extensive research provides information about how proactive behaviors are practiced, for example, through expressing voice. Other relating proactive behaviors are taking charge, issue selling and personal initiative (LePine & Van Dyne, 1998, 2001). Van Dyne et al. (2003) explain voice as an expression of ideas, but there are different conceptualizations of the word voice. I use the definition of LePine and Van Dyne (1998); ‘employees who show initiative and make an effort to change situations and challenging the status quo’. The study of Wolsink (2014) contributes through stating that speaking up behavior is directed at a direct leader with the power to change a decision.
To understand the dynamics of voice, Morrison (2011) proposed three types of voice; 1) suggestion- focused, providing ideas and suggestions for improvement 2) problem-focused voice, communicating concerns and potential hazards to the organization and 3) opinion-focused voice, which concentrates on the dissent-aspect of voice. This study merely focuses on suggestion-focused voice in voicing up situations, meaning voice behavior towards a leader or person higher up in the hierarchy. Suggestion-focused voice is often related to ideas or suggestion of employees towards a leader. But why do employees engage in voice behavior towards leaders?
The question why people speak up or not does not result in an easy answer. A core argument found in research of voice is the important role of pro-social as an underlying motivation. This means that employees would engage in voice behavior if they are motivated through processes such as the desire to bring a important change to the company and feel the obligation to do so (Morrison, 2011). There should be a sense of urgency to challenge the status quo and make a contribution when speaking up. The intent is to bring positive change or improvement. Also, the employee should be aware of the limitations in the current way of doing business and see possibilities for alternatives in suggestion-focused voice behavior (Morrison, 2014).
Earlier research suggests that employees tend to speak up when they prioritize group interest (Tangirala, Kamdar, Venkataramani & Parke, 2013) and the creation of caring climates stimulating pro-social behavior (Wang & Hsieh, 2013). However, the main focus of current research covers the factors a) efficacy, which includes the desired result of voice and b) safety or risk, which includes the potential negative consequences of voicing. A central assumption returning in voice literature is how safe it is to voice; whether voice would upset others or have negative consequences for the employee, such as being assigned to
unreasonable projects, not being considered for promotion or even get fired (Morrison, 2014). Employees constantly weight cost, benefits, risks and likeliness, to assess whether to engage in voice or remain silent (Morrison & Milliken, 2003). Morrison and Milliken (2000 ; 2003) studied voice through scrutinizing the paradox between employees knowing the truth about certain issues and problems in the organization, but not communicating this. They believe that this remaining silent results from certain beliefs such as employees being self-interested and untrustworthy. But they also believe that an important factor is the fear of negative feedback. This fear especially emerges by employees possessing social anxious characteristics
employees voice or not and when they voice what the frequency of voicing is, I examine the relationship of social anxiety towards voice behavior
Social Anxiety of Employees in Voice Behavior
As highlighted in the preceding section, whether employees engage in voice or remain silent does not result in an easy answer. However, it is an important element for leaders or
employers to know which environment and distractions create the optimal and comfortable environment for an employee in which they will voice (Morrison, 2014). Certain factors are believed to influence this voice and silence behavior.
Besides the assumption that employees constantly weight cost, benefits and risks, there is a growing understanding and recognition of other factors. If an employee experiences high levels of fear, they tend to remain silent. Employees may remain silent because of the riskiness of speaking up in hierarchy, the fear of negative evaluation includes the distress arising from concerns about being judged. The employee may automatically retreat without any careful consideration of pros and cons of speaking up (Morrison, 2014). The implicit voice theory of Detert and Edmonson (2011) explains this automatically retreatment through the assumption that employees posses the implicit theory that speaking up results in undesired outcomes, such as damaging of reputation, lower changes for promotion or emotional
wellbeing. This could result in the remaining of silence of especially social anxious employees. Based on these ideas, the following hypotheses are formulated;
Hypothesis 1a; If employees are socially anxious I expect they remain silent more often, because they fear the negative feedback of employers. Therefore, I hypothesize that voice versus silence and social anxiety are negatively related.
Hypothesis 1b: If employees are socially anxious, I expect a lower voice frequency, because they fear the negative feedback of employers. Therefore, I hypothesize that voice frequency and social anxiety are negatively related.
On one hand, negative emotions can result in ‘short-circuiting of systematic processing’ and automatically retreatment of speaking behavior (Kish-Gephart, Detert & Treviño, 2009). On the other hand, it is generally assumed that voice is planned and outcomes are consciously calculated before one chooses to voice or not (Morisson, 2011). One of the factors that might influence this short-circuiting and automatic fear reaction or the consciously calculation is the central executive control. This is an individual’s ability to control his of her attention while doing different tasks (Kane & Engle, 2002; Kane & Engle, 2003). Social anxiety leads to a short-circuiting of systematic processing, however a higher attention control leads to
consciousness. This results in the wondering if a higher attention control could mitigates the negative influence of social anxious employees to voice, so whether attention control
mitigates, strengthen or has no effect on the negative relation between social anxiety and voice frequency.
Attention Control of Employees
The central executive system is assumed to be the center for complex cognitive activities, for example, language comprehension, learning and reasoning (Baddeley, 1992). It is capable of temporarily storage information and manipulating this in a systematic manner. The function consists of keeping task-relevant information active in the system. Research in the 1960s argued the question whether WM should be seen as one single system or a system consisting of multiple components. Different results led Baddeley and Hitch (2003) eliminating the idea of a single system and propose a system where the central executive is supported by two
the distinction of manipulating and storing visual and spatial information (visuo-spatial) and manipulating and storing speech-based information (phonological).
This study focuses on the central executive part of WMC. Putting a load on WM by distracting attention can influence one’s ability to control attention. The higher the WM load, the lower one’s attention control. De Dreu, Baas, Nijstad, Wolsink and Roskes (2012) found that participants with a low load on their WM performed better on a creative task than participants with a high load on their WM. You would expect that low attention control (compiled in a high WM load) and high attention control (compiled in a low WM load) creates a difference for employees’ voice behavior.
This is explained through the central factor of attention control called the individual’s Working Memory Capacity (WMC) (Wolsink, 2014). Especially, Visual Working Memory Capacity (VWMC) emphasizes the individual differences reflecting spatial attention to distractions and is therefore used in this study. At the same time, it is responsible for
‘maintenance and storing visual and spatial task-relevant information, while manipulating and new incoming stimuli to perform a task’ (Baddeley, 1992; Luck & Vogel, 2013; Wolsink, 2014 ). This refers to number of items a person can represent in an on-line state (Moriya & Sugiura, 2013). Visual features flow through the unconscious component visuo-spatial sketch-pad to the conscious central executive. The WM is important in the process of keeping certain tasks of information alive in the memory and not getting lost in the overload of other
distractions during the day.
According to Kane and Engle (2003), attention control is central to WM. They found that, if a goal was maintained, people with high WMC learn faster than people with low WMC. The cognitive and social psychology typically assigned to reasoning, decision-making and social-judgment have in common the distinction of cognitive processes in those that are a) fast, automatic and unconscious b) slow, deliberative, and conscious (Evans, 2008; Tugade &
Engle, 2004). Wolsink (2014) proposed that this systematic and persistence route facilitated by a high WMC would stimulate proactive behavior, where risks and rewards are constantly considered. However, Wolsink (2014) also reviewed that voice is not that often consciously considered but moreover related to stable trait-driven behaviors, such as proactive personality and personal initiative. Based upon previous (emotional) experience and situation-schema’s that voice or remaining of silence flows through a more unconscious and automatic route, the following hypothesis is created;
Hypothesis 2: Attention control regulates the ability for systematic and persistent attention focus on the situation. However, engaging in voice versus silence flows through the automatic and unconscious path. In that case, the level of attention control will not influence whether to voice or not. Therefore, I hypothesize that attention control has no effect on voice versus silence.
When an employee voices, the frequency of voicing can be considered. Here is where the partition in literature arises. Some scholars suggest that voice frequency is also part of the automatic process and attention control will not influence the frequency of speaking up (Wolsink, 2014; Detert & Edmonson, 2011). On the other hand, scholars suggest that voice frequency and proactive action is part of your conscious, slow, deliberative and goal direct process (Parker, Bindl & Strauss, 2010) and it is voluntarily chosen and initiated from within (Grant & Ashford, 2008). Thereby, proactive behavior is stimulated through a systematic and persistence route. From that point of view, it can be assumed that voice frequency is
positively influenced through a higher attention control and negatively influenced through a lower attention control. Based on these views, two contradicting hypotheses (systematic versus automatic) are formulated;
Hypothesis 3a: Attention control regulates the ability for systematic and persistent attention focus. If voice frequency flows through the conscious and systematic path, a higher attention control can positively influence voice frequency and a lower attention control can negatively influence voice frequency. Therefore, I hypothesize that the difference of attention control influences voice frequency.
Hypothesis 3b: Attention control regulates the ability for systematic and persistent attention focus. However, if voice frequency flows through the automatic and unconscious path, the difference in attention control will not influence voice frequency. Therefore, I hypothesize that the difference in attention control has no effect on voice frequency.
Figure 1. Conceptual Framework 1
Figure 2. Conceptual Framework 2a
Attention control
Voice versus Silence Social Anxiety
1a: -
However, when employees experience high levels of fear, they tend to voice less, because negative emotions can results in ‘short-circuiting of systematic processing’ (Kish-Gephart et al. 2009) and the awareness of riskiness of speaking up in hierarchy. Fear is a distraction of attention, so if you are distracted by your own fearful thoughts, you will not voice a high frequency. Anxious and social anxious people tend to be more easily distracted by irrelevant stimuli, so I expect if a person is socially anxious, they fear the reaction of other people and their voice frequency will be diminished. Attention control regulates the ability for systematic and persistent attention focus. If voice frequency flows through the conscious and systematic path, a higher attention control can mitigates the negative influence of social anxiety on voice frequency, because employees are more capable of filtering out irrelevant information and remembering the task-relevant information. Besides, they are less distracted and do not automatically get into fear reaction. Hence they voice just as often as people who are not socially anxious. On the other hand, a lower attention control can give social anxiety an extra negative impulse. In this matter, social anxious employees will even more automatically get into fear reaction and voice less, hence it strengthens the relationship between social anxiety and voice frequency. This leads to the last hypothesis.
Hypothesis 4: Social anxiety of employees is expressed as the sensitivity of employers’ negative feedback. Highly social anxious employees are sensitive for negative feedback and
Attention Control Voice Frequency Social Anxiety 3a: + or - 3b: 0 1b: -
for that reason will remain silent more often than low social anxious employees, who are less influenced through this feeling. If employees have high attention control I expect they are more capable of filtering out distractions of irrelevant stimuli and have a better focus, for that reason mitigates the negative influence of social anxiety on voice frequency. On the other side, if employees have lower attention control I expect they will voice a lower frequency, because they are too distracted. This only applies when voice frequency flows through the conscious and systematic path. In that case, I hypothesize that the difference in attention control influences voice frequency of highly social anxious employees and has no influence on low social anxious employees.
Figure 3. Conceptual Framework 2b
3 Multitasking experiment
Method
Design and Participants
Participants were students of the University of Amsterdam (N= 71, 46.4% female) following a study at the faculty of Economics and Business or followed courses of at least 30 European
WM load
Voice Frequency Social Anxiety
Credit Transfer System (ECTS) at this faculty. This is done to pursue a high feeling of commitment to the faculty. The average age of the participants is 21.87 years (SD = 1.392). The final sample consists of 46.4% female and 53.6% male. This perfectly represents the partition of sexes within the Dutch work force, which consists of approximately 46.5% female and 53.5% male employees (CBS Statline, 2013, June 25). Dutch language proficiency was needed to fulfill the questionnaire and experiment. Most participants were recruited through personal contact. However, to accomplish the participants needed other techniques such as; approaching 3rd year bachelor thesis students in lectures, emails to different volumes at the faculty and social media, were used.
The participants were tested in two different manners, resulting in the use of a Mixed Repeated Measures ANOVA in a 2x2x2 design. First, the independent variable social anxiety was tested between subjects. To create a between subjects factor, I created a mean split, meaning I compared the social anxiety of one participant with another participant. The mean was created, resulting in the distinction; below the mean (1= low social anxiety) and above the mean (2= high social anxiety). The dependent variable of model 1; voice versus silence was tested under a condition of low WMC load and high WMC load. Participant could either remain silent (= 0) or voice (=1). The dependent variable of model 2; voice frequency was also tested under a condition of low WMC load and high WMC load and consisted of the quantity of ideas communicated (≥ 1). I manipulated attention control through creating two conditions. These conditions consisted of two different counting down tasks in which one was extremely heavier than the other one.
The load task in this experiment was the primary task participants had to do: they would receive their salary for the experiment based on this task. Performing well on this task was considered to be in-role behavior of the participant. Their ex-role behavior (or secondary voluntary task) was to voice ideas if they wanted to. Voice was not requested, it was simply
listed as an option, which could either upgrade or reduce the current salary for their primary (in-role) load-task.
Measures and Manipulations
Dependent Variable Model 1: Voice versus Silence
Voice was measured through a brainstorm task where participants were asked to generate 10 ideas in 10 minutes to improve the education system at the faculty of Economics and Business of the University of Amsterdam. After the generation task, participants received two
opportunities; to voice or remain silent. In the first condition, a random selection of 5 ideas would be presented to them for 30 seconds, during which they could select ideas to
communicate to the experimenter. Following this opportunity, in the second condition I presented the other 5 ideas. Voice versus silence was operationalized through the communication of ideas of the participant towards the experimenter or not.
Dependent Variable Model 2: Voice Frequency
Participants who chose the opportunity to voice, could voice between 1-5 ideas per condition towards the leader. A random selection of 5 ideas would be presented to them for 30 seconds, during which they could select ideas to communicate to the experimenter. Following this opportunity, I presented the other 5 ideas. Voice frequency was operationalized through the quantity of ideas communicated towards the experimenter
Independent Variable Social Anxiety
Participants filled in a questionnaire on qualtrics.com one week before the experiment. Social anxiety was measured with 11 items, developed by Carleton, Collimore and Asmundson (2006). Each item was rated on a 5-point Likert scale, ranging from 0 (Not at all characteristic
of me) to 4 (Extremely characteristic of me). An example is; 1) I worry about what other people will think of me even when I know it doesn’t make any difference. A high score represented high social anxiety, meaning the participant feared other people’s opinion and judgment. This variable was tested between subjects and possessed a high internal
consistency (Crohnbach’s alpha between .90 and .91)
Experimental Manipulations: Attention Control
Attention control was manipulated through a counting down task. Two different conditions were created; high and low attention control. This was done to see whether distraction of attention would influence the communication of ideas. The first condition consisted of high attention control (thus in the low WM load condition). The participants had to count down vocally in steps of one, starting at 100 (BEEP 100, BEEP 99, BEEP 98). The second condition consisted of low attention control (thus in the high WM load condition), counting down vocally from 107 with steps of 3 (BEEP 107, BEEP 104, BEEP 101). This variable was manipulated within subjects. To check our manipulation, participants practiced the WM load task before generating the 10 ideas, where they counted back from 100 in steps of two. It was required that the participants filled in the correct number at the end of the task.
Control Variable Working Memory Capacity
WMC was measured for each participant separately. It was measured through an app designed by Ilja Sligte on the Ipad. This app consisted of a ‘sport-test’ where remembering was test. The remembering part took about 20 minutes and consisted of visual manipulations. For example, a rectangle divided in 16 sub-rectangles where in some sub-rectangles tennis rackets were shown. After the tennis rackets disappeared, a red tennis racket was shown to distract the participant. Subsequently, the participant had to decide in which sub-rectangle the
first tennis rackets were shown. Scores were summed up and demonstrated the score in WMC.
Control Variable Perceptions of the Leader
After the experiment, participants filled in a questionnaire about their perception of the experiment leader. In this way we could measure whether the hierarchy was an influential factor and if the participant felt comfortable or safe enough to engage in voice behavior or not. The exit questions were measured with 4 items, developed by Wolsink (2014). Each item was rated on a 7-point Likert scale, ranging from 1 (Completely disagree) to 7 (Completely agree). The 4 separate items were 1) the experiment leader creates a safe environment to take initiative 2) the experiment leader seems an honest person 3) the experiment leader
appreciates initiative and 4) I trust the experiment leader.
Procedure, Tasks and Manipulations
The present experiment was designed to test whether a) social anxiety predicts voice behavior b) attention control influences voice behavior and c) attention control influences voice
frequency of social anxious people. At the agreed date and time the participants made a visit to the M building of the University of Amsterdam, were the experiment took place. The day before their visit they received a reminder via e-mail. The experiment took place in a small room (5x4m2), which was divided in two parts by a cabinet. On one side of the cabinet the participant took place on an office chair 40 centimeters away from a computer. The experiment leader gave a brief explanation and mentioned that the experiment consisted of two parts: 1) two games on the iPad and 2) a test on the computer by which the participants could earn money. Hereafter the participant was asked to sign the informed consent. The
experiment leader explained the games on the iPad and returned to the other site of the cabinet. Silence was essential while the participant completed the iPad games.
In the second part of the experiment the participants used the computer with a 27-inch screen, a mouse and a keyboard. The screen of the participant was linked with a screen on the other side of the cabinet allowing the experiment leader to see the participants’ actions. The experimenter and participant could not psychically see each other, although the experimenter could follow every step of the participant and also judged the choices they made. Further instructions were given on the computer.
The first task was a practice of the WM load task. Participants were asked to countdown from 100 with steps of two every two seconds, which were monitored through beeps (BEEP 100, BEEP 98, BEEP 96,…). The countdown was performed vocally. In the meantime, the experimenter audited and corrected the participant if mistakes were made. If the last number was correct the participant received €2,50
The next task was the Brainstorm task (De Dreu, Baas & Nijstad, 2008). In this task, participants generated 10 ideas for improving the current education system at the faculty of Economics and Business. They were told, that the greatest ideas would be communicated to the board of directors of the faculty. These 10 ideas were immediately coded for originality by the experimenter based on a coding-system developed with brainstorm data from the Dreu et al. (2008). Using these ratings for originality, all 10 ideas were distributed over 2
experimental conditions, leaving each condition with an equal original/unoriginal idea ratio. The computer decided randomly if the participant started with the first or second group.
To manipulate the participants, two groups of equal numbers of ideas were created and linked to two different conditions; either condition 1 where participants started with the low-load and after the high-low-load task or condition 2, which was the other way around. This low-load task is to express the in-role behavior, in other words the real job for which the participant
gets the regular salary. All participants performed the low and high WM load. If the participant filled in the right number in the end, 2,50 euro was earned. In the meantime 5 ideas were shown on the screen for 30 seconds, which could be selected to ‘voice’ towards the experimenter. If the participant ‘voiced’ an idea that was unoriginal they would lose 0.50 cents per idea, however if they ‘voiced’ an idea that was original they would earn 0.50 cents per idea. This was created to monitor the ex-role behavior. Engaging in voice thus involved a risk; participants could either loose or gain money (0.50 euro) with every idea they voiced. This behavior is embedded in the current understanding of going above and beyond the requirements of jobs (Van Dyne et al., 1995; Van Dyne & LePine, 1998). The amount of money is only collected when the participant wrote down the right number at the end of the counting down task. The expectation was that participants in the high attention control condition were more able to focus and for that reason produced more voice than participants in the low attention control condition.
Analyses and Predictions
The participants were tested between and within subjects, resulting in the use of a Mixed Repeated Measures ANOVA in a 2x2x2 design. First, they were tested between subjects, meaning I compared social anxiety of one participant with another participant. Besides, I tested the participant within subjects. I manipulated attention control through engaging in two different load tasks to test whether attention control influences voice behavior.
First, I conducted a reliability test on social anxiety to test whether it is consistent over time. Secondly, the correlation between the variables was explored. The first model described the main effect of social anxiety on voice versus silence and the influence of attention control on voice versus silence. The second model described the main effect of social anxiety on voice frequency and attention control on voice frequency. The third model described the
interaction effect of attention control and social anxiety on voice frequency.
I predicted two main effects for the first model. First, I expected a negative main effect of independent variable social anxiety on dependent variable voice versus silence and no main effect of third variable attention control on dependent variable voice versus silence. In other words, I expected that participants who are socially anxious to remain silent because they fear the reaction of the experimenter and for that reason do not voice. Furthermore, I expected that there would be no difference in voice versus silence in the high or low load condition.
In the second model, I expected a negative main effect of independent variable social anxiety on dependent variable voice frequency and a main effect or no main effect of
moderating variable attention control on dependent variable voice frequency. In other words, I expected that participants, who are socially anxious to remain silent because they fear the reaction of the experimenter. Furthermore, I expected that in two different cases attention control could lead to a positive or negative effect on voice frequency or could also not creating a difference in the high or low load condition. If there is indeed a main effect of attention control on voice frequency, I expect in model 3 an interaction effect of social anxiety and attention control on voice frequency. Meaning, high attention control could mitigates the negative influence of social anxiety and encourage socially anxious participants to voice more often, because they are less distracted and more able to filter out negative distractions and the fear of negative feedback. On the other hand, low attention control could give socially anxious participants an extra negative impulse, because they are too afraid of the negative feedback, resulting in a lower voice frequency.
Figure 4. Measurement Model 1
Figure 5. Measurement Model 2
Working Memory Load Task
High vs. Low load
Attention Control (Carleton, 2006)
0-4 likert, 11 items
Voicing Task (Wolsink, forthcoming) Ideas communicated towards the
leader or not Working Memory Load
Task High vs. Low load
Attention Control
Voice versus Silence Social Anxiety
4 Results
Participants
In total 71 participants completed the experiment. 22 participants were removed due to not completing the counting down task, load task, correctly. A sample of 49 remained for
analyses. The final sample of participants were economics and business students (60% male, 38% female) in all years of enrollment with an average age of 22.10 years (SD = 1.432, range = 20-25)
Social Anxiety
To test whether the measurements mentioned earlier can be used for repeated measures analyses a reliability analysis is conducted. Table 1 shows the Crohnbach’s alpha of social anxiety on the diagonal. Social anxiety was highly reliable (Crohnbach’s alpha = .93). The mean of social anxiety is 3.78 (SD = 0.58). To create a between subjects factor we created a mean split, meaning that participants with an average social anxiety lower than 3.78 are low socially anxious and participants with an average above 3.78 are high socially anxious.
Control Variable WMC
The average WMC of the participants is 7.358 (SD = 0.829, range = 5-8.94). Table 1 shows unexpectedly that WMC is not correlated to voice or silence in the low WM load, r (49) =
(Carleton, 2006) 0-4 likert, 11 items
Voicing Task (Wolsink, forthcoming) Quantity of ideas communicated
towards the leader Voice Frequency Social Anxiety
correlated to voice frequency in low WM load, r (49) = 0.371**.
Control Variable Perceptions of Leader
Results of perceptions of the leader are shown in table 1. The scores are extremely high for the experiment leader creating a safe environment (M = 6.540, SD = 0.734), being an honest person (M = 6.640, SD = 0.663) and being trustful (M = 6.540, SD = 0.788). Scores were also high for appreciates initiative (M = 5780, SD = 1.250). Furthermore, no extreme between leader differences were found. Table 1 shows that only appreciates initiative is correlated to voice or silence in high WM load, r (49) = -.364**. The other 3 elements were not correlated to any variable. Table 1 shows that appreciates initiative is also correlated to voice frequency in the high WM load, r (49) = -0.238*.
Repeated Measures ANOVA Main Effects Model 1 Social Anxiety and Voice or Silence
Table 1 shows the correlation between the main variables. Unexpectedly it turns out that social anxiety and sum score of voice or silence over both condition are not correlated (low r (49) = 0.177, ns ; high r (49) = 0.087, ns). Unexpectedly, there is also no main effect found between the two variables, F (47) = 1.431, ns, η = 0.030. Table 2 shows the averages, another notable fact is that voice is higher for high social anxious people, than for low social anxious people. This could be the result of social anxious participants being more aware that the leader wants initiative and for that reason voice, to avoid a negative reaction for not voicing.
Attention Control and Voice versus Silence
Table 2 shows that the frequency of voice in the low load is higher than in the high load (low
= 45 ; high =43). However as expected, the difference is not substantial enough for a main
effect shown in table 3, F (47) = 0.485, ns, η = 0.100. This means that, the difference in attention control will not influence whether the participant engages in voice or remain silent.
This confirms the hypothesis.
Figure 6. Conceptual Model 1 of the Variables and their Partial Etas
Repeated Measures ANOVA Main Effects Model 2 Social Anxiety and Voice Frequency
Table 1 shows the correlation between the main variables. Unexpectedly it turns out that social anxiety and sum score of voice frequency over both loads are not correlated (low r (49) = -.012, ns ; high r (49) = -.074, ns). Also, unexpectedly, there is no main effect found
between the two variables, F (47) = 0.219, ns, η = 0.005. Table 5 shows the averages; expectedly high social anxious participants voice a lower frequency than low social anxious participants.
Attention Control and Voice Frequency
Table 4 shows that average voice frequency with high attention control is higher than low attention control (high = 2.610 ; low = 2.310). However, the difference is not substantial enough for a main effect, F (47) = 1.245, ns, η = 0.026. This means that, the difference in attention control did not influence the frequency of voicing of the participant. This confirms the alternative hypothesis proposed.
Repeated Measures ANOVA Interaction Effect Model 2
Table 5 shows the interaction effect of attention control and social anxiety on voice η = .10
WM load
Voice versus Silence Social Anxiety
η = 0.00
η = 0.03
frequency. Because no main effect is found between attention control and voice frequency, no interaction effect is found F (47) = 0.154, ns, η = 0.003. However as you can see in table 4 it seems that people with high social anxiety and low attention control voice less than people with low social anxiety and high attention control. Reasons why we think this difference is not substantial and not significant will be discussed later on. The control variables are not included in the repeated measures analysis, because the correlations only accounted for the high or low attention control.
Figure 4. Conceptual Model 2 of the Variables and their Partial Etas
η = .01
Attention Control
Voice Frequency Social Anxiety
Table 1. Descriptives and Correlations Between Variables Model 1 and 2 (Crohnbach’s alpha on diagonal)
M SD 1 2 3 4 5 6 7 8 9 10
1 Voice/Silence Low Load (DV1) .92 .27 1 .12 .50** .05 .18 .07 -.09 -.16 -.17 .12
2 Voice/Silence High Load (DV1) .88 .33 1 -.02 .55** .09 -.36** -.15 -.20 -.22 -.05
3 Voice Frequency Low Load (DV2) 2.58 1.55 1 .25 -.01 .00 -.01 -.11 -.06 -.37**
4 Voice Frequency High Load (DV2) 2.28 1.55 1 -.07 -.29* -.04 -.16 -.13 -.08
5 Social Anxiety (IV) 3.73 1.29 (.93) .22 -.08 -.03 .01 -.05
6 Appreciates Initiative (CV) 5.79 1.25 1 .47** .64** .64** .02
7 Creates Safe Environment (CV) 6.54 .73 1 .66** .61** -.10
8 Seems Honest (CV) 6.64 .66 1 .89** -.04
9 Trust (CV) 6.54 .79 1 -.06
10 Working Memory Capacity (CV) 7.36 .83 1
Note. N = 49. ** p < .01. * p <.05. 0 = Silence, 1 = Voice
Table 2. Repeated-Measures ANOVA Descriptives Model 1
Social Anxiety
Voice Silence
High Attention Control
Low 22 3
High 23 1
Total 45 4
Low Attention Control
Low 21 4
High 22 2
Total 43 6
Note. N = 49. ** p < .01. * p <.05
Table 3. Repeated-Measures ANOVA Results Model 1
Voice or Silence
F η p
Attention Control .49 .10 .49
Social Anxiety 1.43 .03 .24
Table 4. Repeated-Measures ANOVA Descriptives Model 2 Social
Anxiety
Voice Frequency
M SD
High Attention Control
Low 2.64 1.66
High 2.58 1.47
Total 2.61 1.55
Low Attention Control
Low 2.44 1.73
High 2.17 1.37
Total 2.31 1.56
Attention Control, Social Anxiety and Voice Behavior Voice Frequency
F η p
Attention Control 1.26 .03 .27
Social Anxiety .22 .01 .22
Attention Control * Social Anxiety .15 .00 .70
5 Discussion
Summary of Study Results
The main goal of this study is to show a fundamental difference in voice behavior between low and high socially anxious employees and between low and high attention control and specifically that high attention control might mitigates the negative influence of social anxiety on voice frequency and low attention control strengthen this negative influence.. This results in the research question; what is the relationship between social anxiety, attention control and voice versus silence and voice frequency?
First expectations address the influence of attention control on voice versus silence. This study confirms recent research stating that engaging in voice versus silence is an automatic process (Morrison, 2014). For that reason, attention control is not an influential factor, because voice is driven by non-deliberative, schema driven and automatic processes that are triggered by basic emotions such as fear (Detert & Edmonson, 2011; Wolsink, 2014; Morrison, 2014, Kish-Gephart et al., 2009). However, unexpectedly no effect of social anxiety on voice versus silence is found.
Second expectations address the influence of attention control on voice frequency. Two possible outcomes are considered. Most recent research states that proactive action and voice frequency is part of your conscious, slow, deliberative and goal direct process (Parker, Bindl & Strauss, 2010) and it is voluntarily chosen and initiated from within (Grant & Ashford, 2008). However, this study confirms Wolsink’s (2014) and Detert and Edmonson’s (2011) view, because the manipulation of the ability to engage in controlled processing did not affect voice frequency. This indicates that controlled processing does not seem to play (a strong) role in voice decision-making processes, which could be the result of voice frequency
also being an automatic process. Further, no effect between social anxiety and voice frequency is found.
Third expectations address the influence of attention control on the relationship between social anxiety and voice frequency. By distracting attention and lower attention control, it was expected that low attention control strengthen the negative influence of social anxiety on voice frequency, because socially anxious employees cannot focus and are too distracted to the possible negative judgments of their leaders. Thereby, it was expected that high attention control could alleviate the negative influence of social anxiety on voice frequency. However, because no main effects are found, there is also no interaction effect found.
Alternative explanations
The fact that no effect is found of attention control on voice versus silence confirms the theory that voice versus silence is driven by non-deliberative, schema driven and automatic processes (Detert & Edmonson, 2011; Wolsink, 2014; Morrison, 2014, Kish-Gephart et al., 2009). Also no effect is found of attention control on voice frequency, which confirms one way of explanation in voice decision-making processes of the amount of suggestions. It could be the case that voice frequency also flows more through the automatic path and controlled processing does not play a strong role. Besides this explanation also the perceptions of the leader can play a role.
According to the implicit voice theory of Detert and Edmonson (2011) it seems likely that employees, who are socially anxious or have implicit theories about speaking up in public, will voice less than people who do not own this. This theory is based on the assumption that speaking up could result in undesired outcomes, such as damaging of reputation, reduced self-esteem or negative work evaluation, which would result in the
remaining of silence of social anxious employees. Considering these predictions, it is remarkable that no interactions are found between social anxiety and voice behavior.
However, table 8 provides an explanation for this outcome. Most students participating in the experiment are friends or co-students of the experiment leaders. The perceptions of the leader shown in table 8 confirm this matter. Scores regarding the evaluation of safety, honesty, appreciation and trust of the student towards the experiment leader are extremely high and the deviations are extremely low. Wang and Hsieh (2013) already suggested that the creation of caring climates stimulated pro-social behavior. This means that employees would engage in voice behavior if they are motivated through processes such as the desire to bring a important change to the company and feel the obligation to do so. Besides, the study of Morrison, Wheeler-Smith and Kamdar (2011) suggested that work group climates positively influence voice behavior above the individual satisfaction. Most participants of this study were friends or co-students of the leader, meaning that engaging in voice was perhaps more stimulated because of the comfortable environment. This results in an explanation why only 4 students remain silent and why even social anxious students speak up, where perhaps in a less comfortable environment they would remain silent. Interesting of this outcome is, that apparently through creating a comfortable and good environment more social anxious people speak up, this contributes to the research of stimulating proactive behavior through creating a pro-social climate (Wang & Hsieh 2013; Morrison, Wheeler-Smith & Kamdar, 2011). The results also highlight the importance of broadening our conceptual models of voice, to engage the influence of work group climates above the effects of individuals satisfaction and
identification, to include shared cognitions and of conducting additional cross-level research on voice in the future.
Additionally, it is noteworthy that high social anxious people in this experiment voice more often than low social anxious people regarding voice versus silence. This could address
the reversed effect, meaning high social anxious employees could be more aware that the leader expects initiative and to avoid the negative reaction, engage in voice.
When regarding the interaction effect of attention control and social anxiety on voice frequency there is still a disparity. Moriya and Sugiura (2013) suggested that social anxiety influences attention control. If you are highly capable of controlling your attention and for that manner keep goals active, you will have a higher focus and filter out distractions. Moriya and Sugiura (2013) found that socially anxious people get easily distracted, unless they have a high capacity to control their attention. Also, according to the study of Moriya and Sugiura (2013), if the employee possesses a lower WMC, it is less capable to cope with the negative consequences of social anxiety. If you distract attention, which creates low attention control, social anxiety should be of maximum influence. This could result in the remaining of silence or lower frequency of voicing. If you take a look at graphic 2 it is noticeable that the
difference in voice frequency in the low or high load of high social anxious employees is slightly larger than the difference for low social anxious employees. Also, this difference is larger than the difference of social anxiety in the high and low load of voice versus silence. This graphic, slightly confirms Moriya and Sugiura’s (2013) presumptions, however the difference was not large enough for an effect. This could be the result of the creation of a perfect good and comfortable environment or the socially anxious people were not distracted enough to influence their voice behavior. Future research can enlarge the difference of distraction to optimize the contrast of both loads or enlarge the social hierarchy to get a clearer view of the influence of social anxiety.
Positive points
The positive points of this study are embedded in several elements. First of all, this research concerns a mixed empirical and experimental design, which explains more causal relations
than a survey design. It is important to fill the gap between research and practice and one way to handle this, is simulating the business environment carefully.
Secondly, this study confirms previous research that attention control seems to not influence voice versus silence. However, there is still a disparity in the influence of attention control on voice frequency. This study confirms, that voice frequency might also be
automatically processed. Here is where future research can built further and examine if this is really the case.
Thirdly, as explained above, it is remarkable that social anxiety did not seem to influence voice behavior of employees. With creating a less pro-social climate and increasing the social hierarchy it can be explored if social anxiety still does not influence voice behavior. Thereby, exploring if attention control still does not influence voice behavior if the distraction is higher or when including social anxiety, if the person is more afraid and distracted of negative feedback. Future research can also built further on these presumptions.
Methodological point of critique
Every experiment contains limitations. An obvious limitation of this study is our sample of highly educated students. To check the outcomes of this study, the findings will have to be replicated in a business environment and be performed in different layers of the population to generalize the results.
More points of improvement are the design and location. Due to noise and movements in the building, some of the students were obliged to take the WMC test in a room across the hall. The last 10 participants took the whole experiment in another part of the building. Also, the first 20 participants gave some commentary about the function of the mouse and the selecting of ideas in the program. After the first week, we corrected the program and no
problems were remarked anymore. All these problems were taken into account and considered in the analysis. This did not result in big differences.
Also, although we tried our best to simulate the environment carefully, this experiment remains just a brief snapshot of what might happen with voice in a specific work environment situation.
Future research
The biggest limitation of this experiment is the social hierarchy. Students felt really
comfortable and trusted the experiment leader enough to voice. Creating an experiment with a clear hierarchy can explore if the results of this experiment are consistent. Also, the
interaction of leader’s behavior and individual factors can be more specifically explored. This results in a higher focus on emotions and nonconscious processes instead of cognitive and deliberative processes (Morrison, 2014). To examine this, I propose the following experiment;
After every course, students are asked to fill in an evaluation form about the course and the professor. This evaluation form consists of question about the content, but also perceptions of the professor, such as; did the professor stimulate you to work hard? What if we also include questions about the professor on personal level. For example, the professor seems honest and created a comfortable and safe environment. This results in an average variable of the students’ perception about a certain professor. Consequently, when
participating in the experiment, students fill in the pre-test including some questions about courses they followed and who was their main professor. When participants engage in the brainstorm task, a selection of professors, who are/were main professors of the students are shown. We tell the student that this or these professor(s) will evaluate the ideas that are communicated. The name of the student will be linked to the ideas voiced. Besides, we can increase the difference in difficulty of both counting down task, to increase the difference in
distraction. Furthermore, we can expand this idea and include the research points students have to earn in their bachelor. The professor will evaluate ideas that are communicated and based on this evaluation the amount of research points earned. When making it more personal and include professors who posses a high power, the limitations of low hierarchy will be diminished. Also, including the evaluation forms give the opportunity to check what the influence of the professor on voice would be and use the average perceptions as a control variable. In case it is impossible or too expensive to actually include the evaluations of the professor, this is not a complication, because the student still feels evaluated by a professor. This possibly increases the feeling of social anxiety, which will influence the decision to voice ideas or not.
Besides changing the experiment it is also important to perform the experiment at different levels in education to be able to generalize the results. Thereby, although we simulated the business environment carefully, it is also necessary to perform a similar
experiment in a business environment. Another opportunity is to include a longitudinal study monitoring the evolution of proactive goals, planning and behavior over time. Wolsink (2014) suggest that WMC may play a role in goal-maintenance and persistence in the face of
setbacks. WMC allows planning, thinking back and forward, and focus of attention. Since scholars assume that proactive behavior involves conscious planning, anticipation, and persistence in the face of setbacks (Grant & Ashford, 2008; Frese et al. 1996) WMC might benefit the route to proactivity as well. Consider these opinions, research could also develop more in the quality of voice, based on originality and utility, which enhances more the creativity and could result in more positive outcomes for the organization. (Wolsink, 2014)
A question resulting from the findings is: do we want everyone to voice? In our sample, only 4 people remained silent. If this would happen in a real life organization, this would take up a lot of time. Especially when the ideas are of low quality this could lead to unfortunate
outcomes. This means that merely activating everyone’s goal to voice, might give the people who want to say something a chance to overrule the people who have something to say (Wolsink, 2014). Perhaps, you can say we need a diversity of people in organizations. From people who voice often, maybe of low quality but thereby stimulating others to voice, to people who voice less often, but when they do it is of high quality. The design of the present study allows for attention control and voice versus silence and voice frequency to link
different aspects between affect, unconscious processing and the influence of social factors such as social anxiety and the interaction between the leader and student. This was done through a mixed design of empirical en experimental research, which contributes to the theoretical research of the influence of attention control on voice behavior. This contribution especially lies in the creation of a good and caring climate in which people felt comfortable enough to express themselves in upward communication towards a leader. This can help organizations to improve human capital effectiveness.
Concluding Thoughts
The results from our study showed that the difference of attention control does not necessarily leads to voice or high voice frequency. We conclude that voice frequency might also results more from automatic and unconscious processing. Also, we might have eliminated the
negative influence of social anxiety through creating a pro-social environment in which social anxious people felt comfortable enough to voice. So apparently, it does not matter how social anxious or distracted you are, as long as you are in a pro-social climate. Thereby, no
Resulting in the question if we should actually invest such an amount of time and money to create an environment where employees perform higher frequency. In doing so, we hope to inspire future studies to pay more attention to the quality aspect of voice, because quality can improve organizational practice and an overload of frequency would only complicate practice. In this way it can be examined if it is actually useful to create a comfortable and good
environment for everyone to voice. The results also highlight the importance of broadening our conceptual models of voice, to engage the influence of work group climates above the effects of individuals satisfaction and identification, to include shared cognitions and of conducting additional cross-level research on voice in the future. Perhaps if my old company, MG de Jong, would have implemented a caring and pro-social environment and emphasized its importance on group dynamics and not hierarchal pressure from above, it could have prevented them from going bankrupt.
References
Baas, M., De Dreu, C. K., & Nijstad, B. A. (2008). A meta-analysis of 25 years of mood-creativity research: Hedonic tone, activation, or regulatory focus? Psychological
bulletin, 134(6), 779.
Baddeley, A. (1992). Working memory. Science, 255(5044), 556-559.
Baddeley, A. (2003). Working memory: looking back and looking forward. Nature Reviews
Neuroscience, 4(10), 829-839.
Bolino, M., Valcea, S. & Harvey, J. (2010). yee, manage thyself: The potentially negative implications of expecting employees to behave proactively. Journal of Occupational
and Organizational Psychology, 83(2), 325-345.
Academy of Management Executive (1993-2005), 14(3), 52-66.
Carleton, R. N., Collimore, K. C., & Asmundson, G. J. (2007). Social anxiety and fear of negative evaluation: Construct validity of the BFNE-II. Journal of Anxiety Disorders,
21(1), 131-141.
Centraal Bureau voor de Statistiek. (2013, June 25). CBS Statline. Arbeidsrekeningen; arbeidsvolume naar bedrijfstak en geslacht. Retrieved April 30, 2014, from CBS Statline:
http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=81108NED&D1=a&D2= 1-2&D3=1&D4=0&D5=1,11,21,31,36-43&HDR=T,G1,G2&STB=G3,G4&VW=T
Crant, J. M. (2000). Proactive behaviour in organizations. Journal of Management, 26(3), 435-462.
De Dreu, C. K., Nijstad, B. A., Baas, M., Wolsink, I., & Roskes, M. (2012). Working memory benefits creative insight, musical improvisation, and original ideation through
maintained task-focused attention. Personality and Social Psychology Bulletin, 38(5), 656-669.
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(4), 624-668.
Dyne, L. V., Ang, S., & Botero, I. C. (2003). Conceptualizing Employee Silence and Employee Voice as Multidimensional Constructs*. Journal of Management Studies,
40(6), 1359-1392.
Evans, J. S. B. (2008). Dual-processing accounts of reasoning, judgment, and social cognition. Annu. Rev. Psychol., 59, 255-278.
Grant, A. M., & Ashford, S. J. (2008). The dynamics of proactivity at work. Research in
organizational behavior, 28, 3-34.
Kane, M. J., & Engle, R. W. (2003). Working-memory capacity and the control of attention: the contributions of goal neglect, response competition, and task set to Stroop
interference. Journal of experimental psychology: General, 132(1), 47.
Kish-Gephart, J. J., Detert, J. R., Treviño, L. K., & Edmondson, A. C. (2009). Silenced by fear: The nature, sources, and consequences of fear at work. Research in organizational
behavior, 29, 163-193.
LePine, J. A., & Van Dyne, L. (1998). Predicting voice behavior in work groups. Journal of
Applied Psychology, 83(6), 853.
Luck, S. J., & Vogel, E. K. (2013). Visual working memory capacity: from psychophysics and neurobiology to individual differences. Trends in cognitive sciences, 17(8), 391-400.
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(6), 1453-1476.
Moriya, J., & Sugiura, Y. (2013). Socially anxious individuals with low working memory capacity could not inhibit the goal-irrelevant information. Frontiers in human
neuroscience, 7.
Morrison, E. W. (2011). Employee voice behavior: Integration and directions for future research. The Academy of Management Annals, 5(1), 373-412.
Morrison, E. W. (2013). Employee Voice and Silence. Annual Review of Organizational
Psychology and Organizational Behavior, (0).
Morrison, E. W., & Milliken, F. J. (2000). Organizational silence: A barrier to change and development in a pluralistic world. Academy of Management Review, 25(4), 706-725.
Morrison, E. W., & Milliken, F. J. (2003). Speaking up, remaining silent: The dynamics of voice and silence in organizations. Journal of Management Studies, 40(6), 1353-1358. Morrison, E. W., Wheeler-Smith, S. L., & Kamdar, D. (2011). Speaking up in groups: a
cross-level study of group voice climate and voice. Journal of Applied Psychology, 96(1), 183.
Tangirala, S., Kamdar, D., Venkataramani, V., & Parke, M. R. (2013). Doing right versus getting ahead: The effects of duty and achievement orientations on employees’ voice.
Journal of Applied Psychology, 98(6), 1040.
Tugade, M. M., & Engle, R. W. (2004). Individual differences in working memory capacity and dual-process theories of the mind. Psychological bulletin, 130(4), 553.
Wang, Y. D., & Hsieh, H. H. (2012). Toward a better understanding of the link between e thical climate and job satisfaction: a multilevel analysis. Journal of business ethics,
105(4), 535-545.
Wolsink, I. (2014). Attention control & Voice quality
Appendix A
Social Anxiety, Carleton et al (2006). Each item is rated on a 5-point Likert scale,
(Extremely characteristic of me).
(Translated)
1. Ik pieker over wat andere mensen van mij denken, ondanks dat ik weet dat het geen zin heeft
2. Ik vind het vervelend als mensen een ongunstige indruk van mij hebben
3. Ik vrees regelmatig dat mijn tekortkomingen worden opgemerkt door anderen
4. Ik maak me druk over hoe ik op andere mensen overkom
5. Ik ben bang dat anderen mij niet zullen goedkeuren
6. Ik maak me druk over wat andere mensen van mij vinden
7. Wanneer ik met iemand praat, pieker ik over wat zij van mij denken
8. Ik maak mij doorgaans zorgen over de indruk die ik maak
9. Het hindert mij, als ik weet dat iemand mij beoordeelt
10. Soms ben ik me te bewust wat mensen van mij denken
(Original English version)
1. I worry about what other people will think of me even when I know it doesn’t make any difference.
2. (original) I am unconcerned even if I know people are forming an unfavorable impression of me.
2. (revised) It bothers me when people form an unfavorable impression of me.
3. I am frequently afraid of other people noticing my shortcomings.
4. (original) I rarely worry about what kind of impression I am making on someone. 4. (revised) I worry about what kind of impression I make on people.
5. I am afraid that others will not approve of me.
7. (original) Other people’s opinions of me do not bother me. 7. (revised) I am concerned about other people’s opinions of me.
8. When I am talking to someone, I worry about what they may be thinking about me.
9. I am usually worried about what kind of impression I make.
10. (original) If I know someone is judging me, it has little effect on me. 10. (revised) If I know someone is judging me, it tends to bother me.
