Creativity and voice –
From idea generation to idea communication
Bas van Hoorn
University of Amsterdam
June, 24, 2014
Master thesis under supervision of Inge Wolsink
Word count: 10.394 (excl. references and abstract)
Abstract
The aim of this paper is to further knowledge of voice behavior, by investigating the process from idea generation (creativity) to the communication of ideas, while differentiating between the frequency of voice behavior and the quality of voice expressions. To investigate whether creativity influences voice quantity, if voice quality mediates this relationship and if impulsivity plays a role, a field study (N= 147 participants) is conducted. Self-rated measures, peer-rated measures and a
creativity test were used. Creativity has no direct effect on voice quantity, but this relationship is fully mediated by voice quality. It was hypothesized that creativity has a positive effect on voice quality, because originality and usefulness are components of both creativity and voice quality. In turn it was expected that voice quality has a positive effect on voice quantity, because positive reactions of others encourage employees to voice more often. Both hypotheses are confirmed by the results. No effect for impulsivity is found in the results.
Introduction
Communication of potential useful ideas (voice) from employees can be essential for an organization. Imagine yourself working as a department manager, who feels the pressure of the president to achieve a bunch of goals. It seems extremely hard to achieve all goals simultaneously, since your employees feel that they have to sacrifice one goal in order to achieve the other, for example. When you, as the manager, are not aware of their ideas, because they do not voice them, it is unlikely that your
organization will succeed its plans and that the president will be satisfied in the end. When employees tell you about the problems that arise, you can discuss it with the president or adjust the goals. This is just one practical example that illustrates the importance of voice. Voice could be upward
communication to managers and work-related communication to other people higher in the hierarchy, but also communication of ideas to team members. Team members could use information from their colleagues in order to perform better or to make better decisions. Managers could use information from their employees to guide decisions as well, and also in order to solve problems early (Morrison, 2011).
The goal of this paper is twofold. The first goal is to examine the process of voice in the business context, in more detail. The communication of an idea starts with the generation of an idea. So, this paper describes the process from idea generation (creativity) to the communication of ideas. Second, a distinction between how often one communicates and the quality of these expressions is investigated. From the literature review of Morrison (2011) it turns out that most of the current research focuses on voice quantity, so basically why and when employees speak up and the
consequences of this expression. However, a distinction should be made between voice quality and voice quantity. Voice quantity measures how often one speaks up, while voice quality judges the content of the expressed voice (Wolsink, forthcoming). This distinction is important, because low quality ideas or ideas where others already came up with, do not help to improve things. Wolsink (forthcoming) proposes that in order to create value for businesses, the quality of voice should be examined. For example, ideas that are not useful, do not contribute to anything and are therefore a waste of time in most cases. In other words, the role of the usefulness of voice should be studied. Wolsink (forthcoming) shows that voice quality and voice quantity are separate constructs, and that at
least one antecedent of these two constructs differ. She found that attention control influences voice quality, but she found no effect of attention control on voice quantity. Her findings on this difference are based on an experiment and she suggests that future research should address the distinction between voice quality and voice quantity again in a real life setting (Wolsink, forthcoming).
Furthermore, voice and creativity are discussed separately in the literature, but they are not studied together. It seems interesting to include creativity in voice research, because creativity and voice quality are closely related concepts. They are related, because voice quality is the usefulness and originality of ideas that are expressed (Wolsink, forthcoming), while creativity is the ability to generate useful and novel ideas (Amabile, 1983). Wolsink combines these two constructs in her theoretical part, but in her analysis she uses attention control instead of creativity. Attention control is related to creativity, since attention control leads to a systematic and focused analysis and
combination, of different possibilities and is shown to benefit creativity when tasks last longer (De Dreu, Nijstad, Baas, Wolsink & Roskes, 2012).
It is likely that the process which employees go through when choosing to voice or not to voice, is different when voice is expressed to managers as opposed to voice expressed to team members, because in most cases team members are more easily accessible. Besides that, individuals consider the risks that come with voice behavior (Morisson, 2011). Since managers have more power than colleagues, it seems to be more risky to voice towards managers. This is in line with the findings of Liu, Zhu and Yang (2010), who found that voice to managers and voice to colleagues is not the same and that employees express voice mostly towards persons with whom they identify themselves. In this article, the focus is on what influences upward communication, or in other words the voice expressed by employees towards managers, because they have the most power to change things. This means that voice quantity towards managers is studied. However, reports from managers and co-workers about voice quality are taken into consideration, because both their evaluations and reactions can influence ones decision to express voice. Previous studies have primarily focused on voice behavior as a whole (Morisson, 2011).
In this article the relation between creativity and both voice quality and voice quantity is examined. In addition, it is investigated if voice quality is related to voice quantity. In other words,
does a higher quality of expressed voice leads to more expression of voice? Furthermore, it is tested if voice quality is a mediator in the relationship between creativity and voice quantity. Stated another way, it is tested if the relation between creativity and voice quantity is strengthened by voice quality. Also, it is expected that impulsivity influences the voice process. Impulsive individuals often do things started by an impulse, instead of handling as a result of careful reasoning (Evenden, 1999). Therefore, the moderating role of impulsivity in the relation between voice quality and voice quantity is tested. Put differently, it is studied if voice quality is less important for the decision whether to voice for a highly impulsive employee, compared to a low impulsive employee.
This article contributes to the current theory, by investigating more on the distinction between voice quality and voice quantity and by examining both constructs in a business context. Colleagues and managers are asked to rate the voice quality and quantity of a selected colleague. This way, voice quality and voice quantity are measured objectively in a real-life setting. Building on this, the question if voice quality has an effect on voice quantity, and if the relation between creativity and voice
quantity is mediated by voice quality, could generate a better understanding of the effects of voice. More specific, a better understanding on how voice quality and voice quantity interact will give first insights into the process of voice, from idea generation to communication and evaluation, and will thereby establish the basics for more detailed future research.
Besides the theoretical importance, this article is important for practice aspect as well. The distinction between voice quantity and voice quality enables organizations to maximize the outcomes of voice. More specific, a better understanding of the role of voice quality in the voice process is beneficial for organizations and managers, since managers only need useful ideas and ideas that are not yet known by them, like new ideas or ideas that they could not come up with themselves. Moreover, employees with creative abilities can be selected for positions and situations where (high-quality) voice is needed or vice-versa. Also, organizations can manage and develop the creativity of their members while taking voice behavior and quality into consideration.
In summary, this paper tries to elucidate how the process from idea generation to the communication of ideas in a business context works. The focus is on two questions: ‘How does the ability to generate ideas (creativity) influence the frequency of idea communication (voice quantity)? ’
and ‘how is this relationship mediated by voice quality, in the sense that more creative individuals generate better ideas, get more positive feedback and in turn communicate ideas more often?’
Theory Voice
Voice is one specific form of proactive behavior (Morisson, 2011). Morisson (2011) contrasted several slightly different definitions of voice. These definitions share a few similarities. First, voice can be viewed as an act of verbal expression. Moreover, people choose voluntary to voice or not to do so. In other words, it is a discretionary behavior. Furthermore, voice is intended to be constructive, so it has to be expressed with the aim to change or improve situations (Morisson, 2011). While taking these similarities into account the following definition of voice can be drawn: ‘discretionary
communication of ideas, suggestions, concerns, or opinions about work-related issues with the intent to improve organizational or unit functioning’ (Morisson, 2011 p. 375).
Three different types of voice are established by Morisson (2011). She made a distinction between suggestion-focused voice, opinion-focused voice and problem-focused voice. focused and suggestion-focused voice both have the intention to benefit the organization. Problem-focused voice is the expression of voice by employees regarding practices that could (potentially) harm the organization, while suggestion-focused voice is the communication of ideas to improve the organization. On the other hand, opinion-focused voice is the expression of ideas that are different from the ideas that are held by others within the organization. It is important to distinguish between these different types of voice, because both the antecedents and outcomes of voice can be different for suggestion-, problem- and opinion-focused voice (Morisson, 2011).
There is a lot of research done about voice, especially about the quantity of voice (Wolsink, forthcoming; Morisson, 2011). Voice quantity measures how often someone engages in voice behavior, so how often one speaks up. There are a lot of antecedents and outcomes of voice quantity studied already, which are summarized in a framework by Morisson (2011). Motivation to help the organization is an important antecedent of voice quantity, and contextual and personal differences moderate this relation (Morisson, 2011). In turn, an increase in voice leads to for example, better
decision making and better error correction (Morisson & Milliken, 2000). Like already mentioned in the introduction, employees could voice to their team members or colleagues or to their manager. This article focuses on voice to managers.
There are some concepts that are closely related to voice (Morisson, 2011). An early established concept is upward communication. This is any flow of information from an employee to ones supervisor (Athanassiades, 1973), so it is broader in its scope than voice. Another related concept is issue selling, which entails calling attention to trends and developments that have implications for an organization (Ashford, Rothbard, Piderit & Dutton, 1998). This is strategic oriented communication on an organizational level (Morisson, 2011). A concept that is just the opposite of voice, is employee silence, which refers to the withholding of information that could be important for the organization, by employees (Pinder & Harlos, 2001). The findings in this related literature focus mainly on voice quantity as well.
Voice: One concept or different steps in a process?
Notwithstanding, the conceptualization of voice as just one concept in some of the earlier studies, it should be seen as a process consisting out of multiple phases. Parker, Bindl and Strauss (2010) argue that the process of proactive behavior consists of four phases: the generation of proactive goals, the planning of proactive behavior, the actual proactive behavior and the reflection on the behavior. Since voice is a form of proactive behavior, these four phases can probably be used for voice too.
With regard to voice, first, ideas are generated in the mind of the employee (creativity). After that, the employee thinks about how to express the voice best. Then, the actual message is expressed. After that, a reflection on the expressed voice follows. In other words, voice quality is judged in the last phase. Creativity plays an important role in the first phase of the voice process.
Less is known about voice quality, which is the evaluation of expressed voice, compared to voice quantity. Voice quality refers to the originality and usefulness of the ideas that are expressed (Wolsink, forthcoming). So, voice quantity could be high, but if these communications lack quality, the outcomes of voice behavior will be less positive and could even be negative.
Creativity
Most of the definitions of creativity focus on the creative output, or in other words the product that results from the use of creativity. A product has to be novel, useful and appropriate in order to be creative (Amabile, 1983). In this article creativity means creativity at the individual cognitive level, especially focused on the creation of ideas, solutions and suggestions. So, the definition above is slightly adjusted to the following one: Creativity is the ability of one to create novel, useful and appropriate ideas and suggestions. The level of creativity someone possesses is determined by the cognitive ability and style, personality factors, motivation, knowledge and social and contextual influences (Woodman, Sawyer & Griffin, 1993).
There are two types of cognitive processes to creativity: cognitive flexibility and persistence De Dreu, Baas and Nijstad (2008). Cognitive flexibility means that one thinks flexibly and breaks set. Ideas are generated from many different categories. The other way to creativity is persistence. This means that one is creative by means of perseverance. Many ideas are generated then within a few categories. Often, persistence takes more effort and time. Besides being a cognitive process, flexibility and persistence can be measures of creative performance as well (De Dreu e.a., 2008).
Next to that, there are three other measures of creative performance. First, fluency is the number of unique ideas that are generated. Furthermore, creativity is measured by both the usefulness and originality of ideas that are generated (De Dreu e.a., 2008).
Creativity and voice quality
So, the definitions of creativity and voice quality are quite similar and share important characteristics. Coming back to these definitions, creativity is the ability to create novel and useful (Amabile, 1983) ideas and suggestions, while voice quality measures the originality and usefulness of ideas that are expressed (Wolsink, forthcoming). Some would say that these concepts are the same, but they are different in the sense that creativity is the ability to create ideas, while voice quality is the
communication of ideas. In addition, creativity is an ability that one possesses, while voice quality is a rating of expressed ideas by the manager in this case. So, one can have the abilities to imagine
creative solutions, but does not communicate this well.
Wolsink (forthcoming) proposed that creativity is related to the quality of voice. She tested the effect of attention control, on voice quality. De Dreu, Nijstad, Baas, Wolsink and Roskes (2012) found already that attention control is related to creativity. Attention control leads to a systematic and focused analysis and combination, of different possibilities, because it enhances and holds
concentration (De Dreu e.a., 2012). So, people with high attention control could stay focused longer and are better in analyzing different options and insights, which makes them more creative when tasks last longer. For short tasks, the effect of attention control on creativity is less strong. Wolsink’s (forthcoming) results confirm her expectation that higher attention control leads to more voice quality. Since attention control positively influences creativity when time on task is longer, this makes it likely that more creativity results in higher voice quality as well.
So, attention control influences creativity for longer tasks and is shown to influence voice quality positively. Also, creativity and voice quality are related concepts, but in order to judge voice quality, actual communication of ideas is needed first. This leads to the first hypothesis:
H1: Creativity has a positive influence on voice quality of employees, because creativity measures the ability to generate useful and original ideas, while voice quality is the evaluation of communicated ideas. All hypotheses are shown in figure 2, in the analysis and predictions section.
Creativity and voice quantity
Wolsink (forthcoming) also tested the effect of attention control on voice quantity. Beforehand, she proposed that attention control does not have an impact on voice quantity. This hypothesis was confirmed in her experiment. So, there was no positive impact, nor a negative influence of attention control on voice quantity. Since a higher level of attention control results in more creativity (De Dreu e.a., 2012), it could be that creativity has no influence on voice quantity either. However, attention control and creativity are different concepts, with different characteristics, and there are also arguments that creativity could result in more voice quantity.
First, there is evidence that some employees voice more than others, without taking the context in consideration (Morisson, 2011). In other words, individual differences determine a part of voice quantity. Creativity might be such an individual difference, since creativity is the ability to
create novel, useful and appropriate ideas. Creativity people are more open than non-creative people (Goldberg, 1990) and it is argued that creativity and divergent thinking are related to openness (Judge, Heller & Mount., 2002). Le Pine and van Dyne (2001) expected that openness is positively related to voice quantity, because open people look for change and improvement and they are willing to view things from different perspectives. Voice behavior is often change-oriented, and therefore they expected that openness is related to voice quantity. No positive, nor a negative relation between openness and voice quantity could be established from their study (LePine & Van Dyne, 2001)..
However, there are more arguments that show that it is possible that creativity leads to more frequent voice behavior. Attention control benefits the route of persistence to creativity, while there is a way of flexibility too (De Dreu e.a., 2008). Creative individuals possess the ability to think
divergent (Woodman e.a., 1993; LePine & Van Dyne, 2001). This means that creative people are able to think of a wide range of solutions. Especially, unconscious thoughts, so moments when people think about a subject without a clear focus, are more divergent than conscious thoughts (Dijksterhuis & Meurs, 2006). Moreover, one of the components of creativity is fluency, which measures the number of unique ideas that one generates (De Dreu e.a., 2008). One who scores well on this specific component of creativity is able to generate a lot of ideas. So, creative people are able to explore a lot of ideas, and therefore it is likely that creative employees generate more different ideas and see more often a solution to a problem. It is plausible that this, in turn, will increase the quantity of voice behavior.
Though, it does not have to be that creative individuals communicate their ideas. Creativity is just an ability to generate ideas and is not shown to directly influence the communication of it. When deciding whether to voice, individuals consider the chance that their idea will get the attention of the top management (Ashford e.a., 1998) and the perceived efficacy of their idea (Morisson, 2011). The chance that one believes that his idea reaches top management successfully, increases when one thinks that he is able to produce good ideas. High creative people often have high creative self-efficacy (Tierney & Farmer, 2011), which means that they believe that they have the ability to produce creative outputs (Tierney & Farmer, 2002). So, people with high creativity have more
efficacy, which leads to better perceived chances that ideas will reach top management successful and therefore more communication (voice quantity).
Based on these arguments, which point in the direction that creativity influences voice quantity positively, it is proposed that creativity has a positive influence on voice quantity. This effect is probably a weak one, and it is likely that the effect of creativity is weaker for voice quantity than for voice quantity. This leads to the second hypothesis:
H2: Creativity has a positive influence on manager rated voice quantity of employee, because creative individuals are able to explore a lot of ideas (fluency).
Mediation effect of Voice quality
According to the Expectancy Theory of Vroom (1964), individuals choose to engage in a specific behavior, based on what they expect to be the outcome of the behavior. Individuals consider both the chance that their effort will be effective and the reward to which the behavior will lead (Vroom, 1964). With regard to voice, this means that individuals who think of voicing, consider the extent to which their voice behavior will be effective and what their reward will be. Voice quality consists of both manager rated and colleague rated quality, since both their reactions to ones voice could influence expectations of the employee who voice. Ashford et. al. (1998), who studied issue selling, found that employees consider the chance that their idea will get the attention of the management before engaging in such behavior. Also, the silence literature says that employees who remain silent, so those that do not voice, belief that their voice will be not effective (Pinder & Harlos, 2001). This shows that employees consider the effectiveness of speaking up. When the quality of the expressed voice is high, at least in the eyes of colleagues and managers, it is more likely that an idea or
suggestion is put into practice and thus, is effective. In addition, an idea that is high in quality and that could be put into practice, will mostly get more and better rewards, albeit small rewards like better feedback or a better appraisal.
Another argument that high quality voice leads to more voice quantity, is the finding that employees are concerned about the reaction of their colleagues and their boss, when they consider to speak up to change things (Morisson, 2011). In this study, a questionnaire is used to measure voice
behavior and voice quality of the past, and these are rated by managers and colleagues. Both the manager and the colleague ratings of voice quality are taken into account, because both their reactions and opinions influence the confidence of an employee with regard to voicing. When co-workers and the manager view ones voice as highly qualitative, the perceived chance of getting appreciation increases. Therefore, the employee who considers to voice, is more likely to do so. In this article, the focus is on the quantity of voice that is expressed to the manager, since the theory shows that voice quantity towards employees is quite different from voice quantity towards managers. Therefore, the choice is made to focus on voice quantity towards managers, since managers have most power to use the voice. When voice quality is high, it is thinkable that the relationship between the employee and the manager is stronger, because they agree with each other in the end. This, in turn, increases the chance that an employee expresses more voice. So, the quality of the relationship between an employee and a manager, in the eyes of the employee, has a positive effect on how often one voices (Ashford e.a., 1998).
A counterargument can be made as well, since it takes more time on average to develop ideas that are original and useful. This is in line with the dual-pathway theory to creative performance, described by De Dreu e.a. (2008). This theory assumes that one could create more and original ideas by working hard, persevering and systematic processing (De Dreu e.a., 2008; Nijstad, De Dreu, Rietzschel & Baas, 2010). This kind of exploration is more in-depth and from this perspective, relatively a lot of time is invested in one idea, which results in less time for more ideas. However, this could be true from an idea generation perspective, but this paper is about the communication of ideas, namely voice. It is expected that good ideas will be communicated more often that mediocre ones. Besides that, the dual-pathway theory tells us that it is also possible to create ideas in less time, as a result of cognitive flexibility. This means that people with high cognitive flexibility are able to create a lot of ideas in different categories that are original as well (De Dreu e.a., 2008), which is a
component of voice quality. The findings and considerations in this section lead to the next hypothesis:
H3: Voice quality has a positive influence on voice quantity towards the manager, because
the voice quality of one is rated by managers and colleagues, and therefore positive evaluations of voice influence the confidence and expectations of one and thereby the voice quantity of one.
Since it is expected that the effect of creativity on voice quality is stronger than the effect on voice quantity, and that a higher voice quality results in more voice quantity, we propose that voice quality mediates the relationship between creativity and voice quality. In other words, creativity will result in more voice quality, and that in turn will lead to more voice. This combined effect then is expected to be stronger than the direct effect of creativity on voice quantity. To test this, the fourth hypothesis is:
H4: Voice quality mediates the relationship between creativity and voice quantity towards the manager.
The role of impulsivity
Since Wolsink (forthcoming) was the first who differentiated between voice quantity and voice quality, no other factors are studied in combination with voice quality. It is likely that the influence of voice quality on voice quantity is moderated by factors that still have to be explored.
Impulsivity, which is ‘the tendency to act premature without foresight’ (Dalley e.a., 2011 p. 680), is one of the factors that could moderate this relation, since the level of impulsivity of one determines if actions are a handling started by an impulse or handling as a result of careful reasoning (Evenden, 1999). This means, with regard to voice, that individuals who are highly impulsive could express voice as a result of an impulse, while low impulsive persons will think carefully about whether to voice.
Related to this, is the dual process theory of mind, which explains that the human mind processes information either automatic or systematic, to guide ideas and behavior (Barrett, Tugade & Engle, 2004). Automatic processing is passively processing of stimuli, which in turn leads to actions or ideas (Barrett, Tugade & Engle, 2004). Automatic processing is in line with high impulsivity, since both are about handling based on stimuli, without careful reasoning. On the other hand, systematic processing takes more effort and is more controlled (Barrett, Tugade & Engle, 2004), which is more applicable to less impulsive people. When someone scores high on impulsivity, this person will not
think of all the consequences of an action, because information is processed automatic, and therefore it is likely that an impulsive person expresses voice more often. With regard to the relation between voice quality and voice quantity, it is expected that for employees who are impulsive, a lower level of voice quality is needed in order to express this voice. Also, it is expected that the effect of voice quality on voice quantity will be less strong for high impulsive individuals, because they do not think carefully of all the consequences of their action. The direct effect of impulsivity on voice quantity is expected to be positive, because high impulsive individuals will express voice more often. This leads to the last two hypothesis:
H5: The relation between voice quality and voice quantity is less strong for employees with a high level of impulsivity, because high impulsive individuals do not think carefully about their actions and express voice as a result of automatic processing. H6: Impulsivity has a positive influence on voice quantity towards managers.
Method Procedure
A total of seven researchers of the University of Amsterdam worked together to collect the data from triads of participants. A triad is a group of three people, who are answer questions about each other. In this case, a triad means a manager and two employees from the same company, who see each other on a regular base. The purpose of these triads is that the employees answer questions about each other and about their manager. This way, data from a colleague and the manager is collected for two employees. The manager could give insights in his view on his employees. Participants had to fill in a
code before starting each test or survey, so that their different assignments could be matched and that they could be linked in the analysis to their colleagues.
The seven researchers contacted managers and employees in their network, personally, by telephone or by email, with the request if they were willing to participate in this study. Participants had to speak Dutch, since the surveys were in Dutch. Their contact asked two more people within their organization to participate as well. While these people considered participation, they got an email with more information. Employees could win 250 euro’s by participating and managers could win 60 euro’s.
At the agreed time, the researchers showed up at the workplace of the participants. The data in this article is part of a larger research project, so the employees had to make two intelligence tests first. After these test they were asked to make a creativity test and fill in an online survey. All the tests and surveys were done in a quiet and closed room, in order to make sure that the conditions for the tests were equal for all participants. The creativity test and the survey took around twenty-five
minutes together for employees. The managers were asked to fill in an online survey only, which took around fifteen minutes. The researchers waited till everything was finished, in order to ensure that no problems could occur. An email to thank the participants was send after the appointment.
Sample
Participants were recruited within personal networks. Both fulltime and part-time employees
participated. A total of 76 triads were collected. So, the total sample consists of 76 managers and 152 employees. However, due to unanswered questions about key variables, eleven employees are excluded. This means that the final sample consists of 141 employees. 71 of the 141 employees was female (51.1%), against 69 males (48.9%). The average age of employees was 33.3 years (SD=12.2, range=16-61)
Measures
Creativity is measured by a task that participants had to perform. They were asked to type in as many ideas as possible, on how to use one single brick. The minimum time they had for this task was four
minutes, but they could voluntary choose to extend this period with a maximum of four more minutes. The ideas that they wrote down were judged by two independent people, who had to agree about the usefulness and originality of these ideas. A special scale was available for the two judges. A similar task to measure creativity is used by Friedman, Fishbach, Förster and Werth (2003) and De Dreu e.a. (2012) judged creative output by two independent people and a special scale as well.
Voice quantity is measured by the answers that the manager gave about the employees voice behavior. Examples of questions used here are: ‘How often comes your employee with suggestions for improvement?’ and ‘How often comes your employee with plans to change things?’. A total of fifteen statements had to be answered to measure voice quantity. For each form of voice behavior (opinion-, suggestion- and problem-focused) there were five statements. Before the questions about voice quantity a few control
questions were asked, in order to make sure that the participant understood the questions.
Furthermore, it had to be clear where the questions were about, so voice quantity in this case. Participants were asked to indicate to which extent they agreed with fifteen statements on a 5-point Likert scale. These statements to measure voice quantity are adopted from Wolsink (forthcoming).
Voice quality is also measured by the answers of the
colleague and the manager of one employee. The questions about voice quality measure two components, the usefulness and originality of voice. Respondents had to answer, among others, to what extent they agreed with the following statements: ‘My employee comes up with original
advices’ and ‘the ideas of my colleagues are well applicable’. Control questions to make sure that respondents understood the questions were asked again, to make sure that they knew that the questions were about voice quality and not still about voice quantity. There were twenty-six 7-point Likert scale items to measure voice quality adopted from Wolsink (forthcoming).
Impulsivity is measured by seventeen statements that respondents had to answer about themselves. Impulsivity consists of three components: concentration (six items), impulsivity (five items) and planfulness (six items).Examples of such statements are: ‘I plan things carefully’ and ‘I take rapid decisions’. For the analysis, the impulsivity items and one item from planfulness (recoded) will be used. Participants had to indicate to which extend they agreed with 7-point Likert scale items that were adapted from Barratt (1994).
Analysis and predictions
Before testing these hypotheses, the inter-rater reliability for creativity will be determined to check if the two raters agreed with each other. After that, a factor analysis and a reliability analysis will be conducted. We expect that the factor analysis for voice quality will show two different components, usefulness and originality. For voice quantity, we expect three different components: suggestion-focused voice, opinion-suggestion-focused voice and problem-suggestion-focused voice. For impulsivity, also three components are expected: concentration, impulsivity and planfulness. After these steps, the hypotheses can be tested.
In figure 2, the conceptual model with all the hypotheses, is depicted. The first hypothesis will test the effect of creativity on voice quality and a positive effect of creativity on voice quality is expected. For this hypothesis, an average score of usefulness and originality is calculated for both voice quality and creativity. The second hypothesis will test the effect of creativity on voice quantity and it is expected that creativity has a positive effect on voice quantity. The influence of the average of usefulness and originality will be determined. Also, since it is expected that especially creative fluency influences voice quantity, the influence of creative fluency on voice quantity will be tested under the heading ‘additional analysis’. The third hypothesis tests the effect of voice quality on voice quantity, and I expect that voice quality has a positive effect on voice quantity. All these hypotheses
will be tested by a single regression analysis with process and since all the expected effects are positive, we expect all the regression coefficients to be positive and significant (p<0,05).
Hypothesis four tests whether voice quality mediates the relationship between creativity and voice quantity. We expect that voice quality explains at least a part of the relationship between creativity and voice quantity. So, we expect that the direct effect of creativity on voice quantity is less strong when voice quality is included in the analysis. The fourth hypothesis will be tested with the software plug-in Process in SPSS. A 95 % confidence interval will show if the mediation effect is significant, when zero is not included in the confidence interval.
The expectation for the fifth hypothesis is that impulsivity is a negative moderator for the relationship between voice quality and voice quantity. So, it is expected that the effect of voice quality on voice quantity is stronger for high impulsive persons. This moderation effect will be tested with the software plug-in Process in SPSS as well. Again, when zero is not in the 95 % confidence interval, the moderation effect is significant.
Results
Only complete questionnaires were included in the analyses. This resulted in a loss of 11 participants, leaving us with a sample of 141 participants.
Factor- and reliability analyses
A principal components factor analysis with oblimin rotation is conducted for voice quality, voice quantity and impulsivity and not for creativity, since that was a test.
Table 1 shows the factor loadings for impulsivity. The first item is recoded from the
planfulness scale. The last two items do not have sufficient loading on factor 1 (impulsivity) and will probably measure something else. Therefore, they are excluded from the factor impulsivity. The Cronbach’s alpha for this variable is 0.83 and the scale does not become more reliable when items are deleted.
Table 2 shows the factor loadings for voice quantity items. As expected, three factors can be identified, problem- , suggestion- and opinion-focused voice. For the final analysis, all the three factors will be grouped together for the factor voice quantity. The Cronbach’s alpha for voice quantity is 0.90 and the scale does not become more reliable when items are deleted.
Table 3 shows the factor loadings for voice quality, rated by managers. Factor one is originality and factor two is usefulness. Consequently, items in bold (mostly recoded) were deleted, because of low factor loadings on the two final factors. The recoded items formed a separate factor, what indicates that these items were not interpreted as the opposite as originality or usefulness. The alpha for originality is 0.96 and for usefulness is 0.86. For this scale, again no items needed to be deleted because of the possibility to increase reliability. The alpha for the total scale (originality and usefulness combined) is 0.96
Table 4 shows factor loadings for voice quality, rated by co-workers. Again, items in bold are deleted, because they were not interpreted as the opposite of originality or usefulness. The reliability for originality here is 0.93, for usefulness it is 0.87 and the reliability for these two combined is 0.94.
Inter-rater reliability
For creativity, ideas from participants are rated by two independent judges on the usefulness and originality of these ideas. To determine if these judges agreed with each other, the inter-rater reliability (IRR) was calculated with correlation coefficients. The IRR for originality is 0.70 and for usefulness it is 0.52. The IRR for usefulness is quite low and this is a first sign to be careful with the use of usefulness for further analysis.
Furthermore, for voice quality, which is rated by managers and co-workers, a IRR should be calculated. Since they did not judge exactly the same idea (as with originality and usefulness in the creativity test), we expected a lower correlation. The IRR for voice quality originality is 0.37, for usefulness it is 0.25 and for the total scale the IRR is 0.36. Since correlations between raters in
source research is often not higher than 0.35, we decided to combine the quality ratings of managers and co-workers. Furthermore, a mean score is calculated for all the different factors identified in the factor analysis. These are: impulsivity, suggestion-focused voice, opinion-focused voice, problem-focused voice and voice quantity (total).
Correlations and hypotheses Table 5
Means, standard deviations and correlations.
_____________________________________________________________________________________________________________________________________________________________ Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 _____________________________________________________________________________________________________________________________________________________________ 1. Employee Age 33.66 12.25 -2. Employee Gender 1.51 0.50 -0.04 -3. Impulsivity 3.32 1.19 -0.15 0.04 (0,83)
4. Problem-focused voice quantity 3.07 0.77 0,18* 0.03 -0.07 (0,74)
5. Suggestion-focused voice quantity 3.14 0.81 0.06 0.08 0.05 0,67** (0,87)
_____________________________________________________________________________________________________________________________________________________________ 6. Opinion-focused voice quantity 3.21 0.82 0,26** 0.05 0.11 0,63** 0,57** (0,81)
7. Voice quantity (total) 3.14 0.69 0,19* 0.06 0.04 0,87** 0,89** 0,84** (0,90)
8. Voice quality - usefulness 4.53 0.83 0.15 0.01 0.07 0,55** 0,62** 0,44** 0,63** (0,87)
9. Voice quality - originality 4.01 0.96 0.06 0.02 0.15 0,59** 0,75** 0,52** 0,73** 0,75** (0,94)
10. Voice quality (total) 4.20 0.86 0.10 0.02 0.13 0,61** 0,75** 0,53** 0,74** 0,88** 0,97** (0,95)
_____________________________________________________________________________________________________________________________________________________________ 11. Creativity - usefulness 3.30 0.45 0.03 0.07 -0.13 -0.12 -0.09 -0.07 -0.10 -0.13 -0.15 -0.15 -12. Creativity - originality 2.60 0.41 -0.07 0.06 0.05 0.15 0.11 0.11 0.14 0,20* 0,21* 0,22* -0,68** -13. Creativity (total) 2.95 0.17 -0.04 0.16 -0.12 0.03 0.02 0.04 0.03 0.07 0.05 0.06 0,49** 0,31** -14. Creative fluency 13.18 9.81 -0.06 -0.01 0.11 -0.09 -0.01 -0.01 -0.03 0.00 0.03 0.02 -0,49** 0,45** -0,10 -_____________________________________________________________________________________________________________________________________________________________
Table 5 shows the correlations between the most important (sub)variables. Some variables, like voice quantity, are divided into sub-variables (factors) as well. The usefulness component of creativity is negatively correlated with both voice quantity and voice quality, while the originality component correlates positive with voice quality and voice quantity. Furthermore, the correlation between the two components (creative originality/usefulness) is negative and significant (-0,68). Apparently,
usefulness in the creativity test and usefulness in business ideas are not quite the same. Since both originality-usefulness correlations, the IRR and usefulness- voice quality correlations are low or negative, we excluded the measure of creative usefulness from analyses. Creativity will therefore be represented by the originality measure.
All hypotheses are tested with the process macro (model 14). Bootstrapping is used 1000 times, with a confidence interval of 95 %. Significant results are indicated by a confidence interval that excludes zero or a p-value below 0.05. First, we tested the direct effect of creativity on voice quality. Creativity has a (β =0.461, t= 2.63, p=0.001) positive effect on voice quality. Bootstrapping confirms that there is a significant effect (0.11, 0.81). We conclude that in our sample, the
independent variable creativity influences the mediator voice quality, as expected in hypothesis 1. Second, we tested the effect of voice quality on voice quantity. The effect of voice quality on voice quantity is strong and positive (β= 0.530, t= 3.57, p= 0.001). The bootstrapping confidence interval confirms this effect, since it did not contain zero (0.24, 0.82). So, voice quality has a significant positive effect on voice quantity. So, in this sample employees who voice high quality managers and colleagues, also voice more frequently to managers, which means that H3 is supported.
Next, we tested the direct effect of creativity on voice quantity. Unexpectedly, we did not find an effect (β = -0.035, t= -0.35, p= 0.729). Bootstrapping confirms that in our sample, creativity does not directly affect voice quantity (-0.23, 0.16), indicating that a high level of creativity does not directly influence ones voice frequency and therefore H2 is not supported. However, it is still possible that we find (full) mediation in our indirect effects.
Now, we tested the mediating effect of voice quality for the relation between creativity and voice quantity. The total effect has a coefficient of 0.239 and is not significant based on the 95 % confidence interval (-0.05, 0.52), because zero is in the interval. The direct effect, with a coefficient of
-0.035, is again not significant based on the CI (-0.24, 0.16). The indirect is significant with a positive coefficient of 0.278, since the 95% CI does not contain zero (0.05, 0.51). So, the relation between creativity and voice quantity is fully mediated by voice quality (H4 is supported). This indicates that the level of creativity someone possesses influences the quality of voice, which in turn influences how often one voices (indirect effect).
Lastly, we tested the proposed moderation an direct effect of impulsivity. We did not find the predicted interaction (β = 0.023, t= -0.35, p=0.579), indicating that the relationship between how often people voice and the quality of voice is not moderated by impulsivity. We therefore reject H5.
Furthermore, we did not find a direct effect of impulsivity on voice quantity (β = 0.127, t= -0.73, p=0.465). This indicates that the level of impulsivity of one, does not have an effect on how frequently one expresses voice and means that H6 is not supported.
Additional analysis
The test results for H2 indicate that in our sample, creativity has no effect on the voice quantity of employees. However, as described in the theoretical part, it is expected that especially the number of ideas that people generate (creative fluency) influence voice quantity. So, the effect of creative fluency on voice quantity is tested as well, with a linear regression analysis. In our sample, no effect of creative fluency on voice quantity is found (β = -0.026, t= -0.31, p= 0.759). This means that the number of ideas that people are able to generate does not influence the frequency with which voice is expressed.
Discussion
The study was designed with two goals in mind. In the current literature, voice is conceptualized as one concept which focuses mainly on how often and why employees voice. However, Wolsink (forthcoming) proposes that voice should be split up into voice quantity and voice quality. She claims that a distinction should be made between how often one communicates ideas and the quality
(originality and usefulness) of these ideas, because there is a difference between ‘having to say something’ and ‘having something to say’ (Wolsink, forthcoming). In other words, not all expressions of voice contribute equally to improvements. The first goal of this paper was to investigate the
constructs of voice quantity and voice quality in a business context. Besides, creativity shares a lot of similarities with voice quality, but probably because of the fact that the focus was merely on voice quantity, creativity and voice are not frequently studied together yet (Morisson, 2011; Wolsink, forthcoming). Therefore, the second goal was to integrate creativity in voice research and investigate the process from idea generation to the communication of these ideas. These goals led to the
following research questions: ‘How does the ability to generate ideas (creativity) influence the frequency of idea communication (voice quantity)? ’ and ‘is this relationship mediated by voice quality, in the sense that more creative individuals generate better ideas, get more positive feedback and in turn communicate ideas more often?’
Table 6
Summary of regression, interaction and mediatin analysis with process.
___________________________________________________________________________
Effect β t 95% CI
___________________________________________________________________________
Creativity on voice quality 0.461** 2.63 0.11, 0.81
Creativity on voice quantity -0.035 -0.35 -0.23, 0.16
Voice quality on voice quantity 0.530** 3.57 0.24, 0.82
Indirect effect (creativity -> voice quality) on voice quantity 0.278* - 0.05, 0.51
Impulsivity on voice quantity -0.127 0.73
-Interaction (impulsivity*voice quality) on voice quantity 0.023 -0.35
-Creative fluency on voice quantity -0,026 -0,31
-______________________________________________________________________________________ Note. ** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
Summary
The results indicate that voice quality and voice quantity are indeed different constructs, in the sense that the influence of creativity on them differs. Despite the fact that creativity is not frequently included in voice research (Wolsink, forthcoming), this study shows that creativity plays an important role in the process from idea generation to idea communication. No direct effect of creativity on voice quantity is found, but the effect of creativity on voice quantity in this study is fully mediated by voice quantity. This means that when voice quality is included in the model, voice quantity is influenced by creativity indirectly, because creativity results in a higher quality of voice, which in turn results in more communication of ideas.
To answer the research question, six hypotheses were tested. The results show that, as hypothesized based on the similarities between creativity and voice quality (Wolsink, forthcoming; Amabile, 1983), creativity influences voice quality positively. This means that employees who are more creative, communicate ideas of higher quality in the eyes of both colleagues and the manager.
Furthermore, voice quality has a positive effect on voice quantity, which means that
employees who communicate high quality ideas, communicate ideas more often than employees who score low on voice quality in the eyes of both colleagues and managers. This finding was expected, because we assume that the positive feedback of colleagues and managers encourages employees to express voice more often. Also, with regard to the expectancy theory (Vroom, 1964), employees who communicate high quality ideas, expect that their ideas are evaluated positively again and will therefore communicate ideas more often. However, it could be the other way around as well, in the sense that employees who voice more often, will communicate ideas of higher quality. It could be that employees learn from the feedback that they receive on their ideas and consequently voice higher quality in the future. In the model used in this study (from creativity to idea communication), it is likely that first ideas are generated in the mind of the employee (creativity). After that, the employee thinks about how to express the voice best. Then, the actual message is expressed. After that, a reflection on the expressed voice follows. This process of multiple steps is in line with the article of Parker, Bindl and Strauss (2010), who argue that proactive behavior is stepwise process. So, we think that voice quality has a positive influence on voice quantity, because the reflection on the expressed
voice influences the frequency of one’s voice behavior. However, it is possible that more voice quantity in turn results in even higher voice quality (learning). In other words, it could be that employees with a high quality of voice, express their ideas more often because of positive
feedback and as a result of that, even increase the quality of the ideas that they communicate (circular effect). A visual model of this possible effects is depicted in figure 5.
Moreover, no direct effect of creativity on voice quantity is found. Based on the theory that creative fluency is a component of creativity (De Dreu e.a., 2008) and that creative individuals have high creative self-efficacy (Tierney & Farmer, 2011), it was expected that creativity had a direct positive effect on voice quantity. However, for none of the components of voice quantity a direct effect was found.
The fact that there is no direct effect of creativity on voice quantity is not a completely unexpected finding. It was expected that voice quality mediates this relationship, because creativity enables employees to produce better ideas and because of more positive evaluations of communicated ideas, individuals will voice more often. In this study, even full mediation is found, which means that there is no direct effect of creativity on voice quantity, because voice quantity mediates this
relationship completely. So, creativity does not directly results in a higher voice quantity, but it results in higher voice quantity indirectly, because creativity enhances the quality of communicated ideas and therefore creative employees communicate ideas more frequently.
For impulsivity, a negative moderating effect for the relationship between voice quality and voice quantity and a positive direct effect on voice quantity were expected beforehand, based on the theory that impulsive individuals often do things based on an impulse instead of acting as a result of careful reasoning (Evenden, 1999). However, the results indicate that neither the moderating effect, nor the direct effect are present in our sample. Also, impulsivity does not have significant correlations
with any of the variables. A possible explanation for the absent effects of impulsivity is that the questions about impulsivity in this study were self-report items. So, participants had to indicate how impulsive they are themselves, while voice quantity and voice quality were rated by others. When we look back at the dual process theory of the mind, we see that there are two different ways the human mind works (Barrett, Tugade & Engle, 2004). Impulsivity is related to a route of automatic
processing, which means that impulsive individuals do not think carefully about what they do. This could be a reason that it is hard for impulsive individuals to judge their own level of impulsivity, since when answering a questionnaire the answers are often a result of careful and conscious (as opposed to automatic) thinking. It could be that impulsive individuals do not even know that they are impulsive, because they do not consciously experience it when they act impulsive. A solution for this, that could be applied in future research, is to use a test to measure impulsivity. A risk taking test can be used for this, because impulsivity is related to risk taking and is shown to correlate with the outcome of the balloon analogue risk task (BART) of Lejuez et. al. (2002). The BART measures how willing individuals are to take risks. In this task, participants get a balloon on a computer screen and they have two options to choose. They can choose to inflate the balloon by clicking on it, which means that they earn an additional amount of money which is the same for each pump. Or they can choose to collect their earnings so far, by clicking on the collect earnings button. This means that they get another balloon out of a total of ninety to repeat the same task. By each pump, the chance that the balloon explodes will grow, but the participants do not know what these chances are. When the balloon explodes, the collected amount so far by the current balloon is lost. Different balloons with different chances of exploding could be used (Lejuez, 2002). To measure impulsivity, the reaction times for this task could be studied as well. To compare the reaction times, equal circumstances have to be created for all participants. So, this task can be used as a test for impulsivity. Also, risk taking itself could be studied in relation to voice, since it is related to impulsivity.
Theoretical implications
The findings connect to the current theory in two ways. First, it connects to the dual pathway to creativity theory of De Dreu et. al. (2008). This theory assumes that there are two ways to creativity:
the route of flexibility which is more automatic processing and the route of persistence, which is more effortful. Creative individuals that use a route of flexibility think divergent, which means that these people are able to think of a wide range of solutions (Woodman e.a., 1993; LePine & Van Dyne, 2001). People with high cognitive flexibility are able to create a lot of ideas in different categories that can be original (De Dreu e.a., 2008). So, the route of flexibility can result in original ideas. However, to score high on voice quality, it is likely that both originality and usefulness are needed. Think of a very useful idea, that already will be implemented next month by your organization. This idea is not original anymore and will not help the organization, so the quality of this idea will be rated low. Contrary, an idea that is very original, but of which the costs are higher than the benefits, is not useful and therefore voice quality for this idea will also be low. To create useful ideas, often more effort and thinking is needed, which means that a route of persistence benefits the usefulness. It could be that employees with high voice quality are able to concentrate better or that they are willing to put in more effort in the idea generation process, because they have different motives for voicing than people with low voice quality. So, in order to reach a high voice quality, employees can use the route of flexibility to explore ideas that are original, but probably a route of persistence will benefit the usefulness of an idea.
Besides, the findings of this study connect with the theory of Wolsink (forthcoming), in the sense that voice quantity and voice quality are different constructs, also in the business context. The route of persistence will benefit from a high level of attention control, because creativity will be higher when tasks last longer (De Dreu e.a., 2008). It is shown now that both attention control and creativity have a positive direct influence on voice quality, but not on voice quantity. This means that the distinction between voice quantity and voice quality holds and that they have different
antecedents.
The distinction between voice quantity and voice quality makes it possible to see voice as a process consisting of multiple steps, instead of one concept. Voice quality is the evaluation of ideas that are communicated (voice quantity). Since voice quality has similarities to creativity (Wolsink, forthcoming; Amabile, 1983), the step of idea generation (creativity) is added to the voice process as well. This is a more complete overview than before, of the process from the generation of
ideas to the communication of ideas. However, still a lot is unknown concerning this process and especially about voice quality.
Practical implications
There are implications for practice as well. Because of the fact that creativity is not frequently studied in relation with voice, this article gives some new insights, like the positive effect of creativity on voice quality. Managers who believe that voice from employees is essential for the success of the company, could select creative employees for specific positions. Especially for positions where decisions do not have to be taken quickly and where discussions and new ideas could improve things, creativity and qualitative voice could contribute to these improvements. When decisions have to be taken quickly, a lot of voice will only slow down the processes unnecessarily. Also, when a manager makes decisions mainly on information and feedback from employees, creativity can benefit the quality of these information (voice). Furthermore, it is plausible that because of the positive feedback on high quality voice, employees will voice more often. When managers want more voice quantity from their employees, it could be wise at least not to criticize ideas from employees too hard, but rather give constructive comments. It could be that there is a learning effect as well, so that when employees express voice more often, they will become better in it over time.
Limitations and positive points
There are some limitations to this study. First, some questions were answered by participants about themselves (self-reports), while other questions were answered by colleagues or the manager. Often, self-reported questions are more subjective, because participants do not exactly know how they behave. Also, the participants could be a bit biased, because the questions were about work related subjects and the manager was participating as well. This could have resulted in socially desirable answers. In the case of impulsivity, besides the possibility that individuals do not know whether they are impulsive, it could be that they state that they are less impulsive than they actually think they are. Another limitation is the fact that we cannot be sure of the causality in this study. The results are based on correlations (regression) and therefore it is not possible to be certain about the direction of
the relationships. This makes is hard to tell whether voice quality really results in more voice quantity or that it is the other way around. However, with the theoretical background, we assume that voice quality results in more voice quantity or that it is a circular effect. Furthermore, the sampling
technique that is used for this study was a mix of convenience sampling and snowball sampling, since acquaintances were approached to participate in this study. Then this person asked two others within the company. A disadvantage of this type of sampling is that the sample is often less representative. The companies of the participants were concentrated in a small geographical area (mostly
Amsterdam) and this reduces the generalizability of the results. Although relatively most Dutch employees work in this area, also a lot of employees in the Netherlands work somewhere else. A replication of this study in other areas (preferably other countries) with more random sampling, will validate the results. Another limitation is that the questionnaire for employees took a lot of time (25 minutes), because a lot of variables were in the questionnaire. The benefit of this, is that more variables can be studied. However, participants could get less concentrated and think less careful about their answers after a while.
Alongside the limitations, an obvious positive point is the use of triads in this study. A triad means that three persons from one company are involved and answer questions about each other. This way, a complete picture is created and bias is reduced. For voice quantity and voice quality, the judgment of others is necessary, because they have to deal with the ideas and give feedback. Another positive point is the fact that the researchers visited all the companies and waited on location till the questionnaires were finished. This ensured that the participants completed the surveys and created the feeling at the participants that it was a serious study. The presence of the researchers was also a positive point for limitation that the questionnaire took a lot of time, since the researchers invested a lot of time as well (noticed by the participants). Furthermore, the creativity test was a positive point. By this test, creativity is measured objectively, which is better than asking questions about how creative one is. In most cases, participants cannot answer questions like this about themselves, since many employees do not exactly know what creativity is, or how well they score on the different components. The test in this study solved these problems. The fact that the ideas from the creativity test are rated by two independent judges makes the creativity scores even more reliable.
Future research
Like already mentioned, it is one of the first times that creativity is studied in relation with voice and that voice quantity and quality are separated for research in a business context. To make the results more generalizable, this study could be replicated, preferably in other countries. Furthermore, there are a lot of factors known that influence voice quantity, but there are no other factors known that influence voice quality. Other factors that can influence the process from idea generation to
communication and their effect on voice quality could be explored in other studies. We believe that it is interesting to study the effect of transformational leadership on voice quality. There is already a lot known about transformational leadership and voice quantity (Morisson, 2011), but it could be that transformational leaders inspire employees to come up with very useful and original ideas as well. Future research can study this influence for example, but there are more factors (like risk taking) that are interesting to study. This could make the focus on the quality of ideas, more easily applicable into practice too.
Also, since it is plausible that voice quality is partly achieved by the route of persistence to creativity (see theoretical implications), it seems interesting to study the effects of some other facets of creativity in this model. For example, creative persistence and flexibility can be measured in a creativity test, by focusing on how many categories participants use (based on a scale that indicates categories). Flexibility could be measured by the number of different categories that are used, while persistence can be measured by the number of unique ideas divided by the number of categories that are used (De Dreu e.a., 2012). Then, it could be studied if the route of persistence is related to the originality, the usefulness or both components of voice quality. The same can be done for the route of flexibility.
Furthermore, the direction of the relation between voice quality and voice quantity could be studied in more detail. As mentioned previously, it could be that employees with high voice quality, express their ideas more frequently because of positive feedback they receive and as a result of that, the quality of the ideas that they communicate increases as well. This circular effect would indicate that employees learn from communicating ideas by performing the behavior. However, we need experimental designs to study the direction of this relationship is. To study this, for example a social
setting should be created, where participants act as employees. The manager should be someone who is hired by the researcher and acts if he is a participant. This way, the feedback and the degree to which ideas are used by the manager could be manipulated. A control group could be used to compare the effects. Employees could choose whether to advice their managers about decisions, based on the information they have. The quality of these advices can be judged by the managers. Managers could choose to make use of the advice or to ignore it and to give feedback on the ideas. Also, a possibility is to let other employees give feedback if they want to. Employees could rate how they perceive the feedback. Then, employees could choose to advice the manager again if they want, about other decisions. The effect of voice quality on the feedback and decisions of managers and the effect of the feedback on voice quantity of employees could be studied then. When this is repeated a few times, it could also be studied if the quality of voice increases when employees voiced more often.
Conclusion
We conclude that creativity is the first step in the voice process and is related to voice quality. Creativity also seems to have an indirect positive effect (but no direct effect) on voice quantity, because creativity influences voice quality and this in turn influences voice quantity. This means that the feedback from colleagues and the manager is important for creative individuals who think whether to express voice to their manager. It can be concluded that voice quality and voice quantity are, in any case different concepts with regard to creativity. Future research should study more factors that influence voice quality.
