Personality and Voice Behavior:
The Effect of Behavioral Systems on the Relation between
Working Memory Capacity and Voice Behavior Quality
H.A.A. Spelt
BACHELOR’S THESIS
30-06-2014
Cijfer: 8.5
University of Amsterdam
Student number: 10002396
Supervisor: I. Wolsink
Word count: 6.533
ABSTRACT
Previous studies have proved that there is a positive relation between working memory capacity (WMC) and voice quality (Wolsink, 2013). This study investigates whether the different behavioral systems, BIS and BAS, developed by Gray (Carver & White, 1994) result in different calculations of costs and benefits of voice behavior, and thereby have a respectively negative (BIS) and positive (BAS) mediating effect on the relationship between WMC and voice quality. We measured BIS/BAS, WMC and voice quality with a within-subject design among students (N = 62). Unexpectedly we found a negative relationship between BAS and voice quality and a positive relation between WM distraction and BAS, meaning that BAS-employees are easily distracted resulting in worse WM performance causing a lower quality of voice. Findings suggest that WM of BIS-employees is better trained than BAS-employees and that BAS-employees only provide high quality ideas if that is their primary task.
Keywords: voice behavior, voice quality, working memory capacity, behavioral
Bob and Barry have been working for the same company for ages. Santa’s Distributions is specialized in wrapping and distributing gifts from the North Pole to all the children in the world during Christmastime. Barry and Bob are a fabulous team, but they behave very differently towards their boss Santa Claus. Barry has lots of improvement ideas on the working method of the company which he forwards to Santa Claus. For example, Barry thinks that the packaging paper should be blue instead of red and that the gifts should not be bigger than one cubic meter. Santa Claus does not agree with Barry’s suggestions and leaves things the way they are. In contrast to Barry, Bob never forwards his ideas towards Santa Claus. But secretly he thinks that the company should stop with providing different sorts of gifts and just give all the children an iPad. The latter is actually a really good idea, because of the low diversity it saves a lot of administrative work, the iPad fits through every chimney and since Mr. Jobs was a true believer in Santa Claus the company gets a lot of discount. But why does Bob remain silent towards Santa Claus considering that his idea could benefit the whole organization? And why does Barry keep forwarding his ideas to Santa Claus, although he knows the boss does not like his ideas?
Bob and Barry illustrate an activity that in behavioral science is called voice behavior (Morrison, 2011). Voice behavior concerns the way people act, speak up (like Barry) or remain silent (like Bob), in certain situations considering their ideas, opinions, concerns and suggestions. A main difference between Bob and Barry lies within the frequency and the content of their ideas. Barry has a lot of ideas, but the content is not very good, whereas Bob has only one really good idea. Since the goal of voice behavior is to change the status quo towards a better situation, it makes sense that for an effective improvement of an organization the content of ideas is more important than the amount of ideas. The content of an idea depends on the creativity of the person, which in turn depends on persistent processing of information (De Dreu, Nijstad, Baas, Wolsink, & Roskes, 2012). Facilitated by working memory, this processing makes it possible to stay focused on one thing. Working memory has multiple functions. The most relevant for this study is the amount of information one can process and the ability to ignore distractors.
Besides the capacity of one’s working memory more factors are involved in the explanation why Bob kept his mouth shut, while Barry just kept on forwarding his ideas. They both reacted
differently, even though the situation was the same. Maybe it has something to do with the fact that Bob and Barry are two totally different persons.The effect of personality on the intent of voice was investigated with the use of different theories; for example Bindl and Parker (2010) used proactive personality and lePine and van Dyne (2001) used the Big Five. However, these studies lack in neurological basis. Therefore we decide to focus on a behavioral system developed by Gray which finds his neurological basis in the same neural structure that is also important in attention control (Gray & Braver, 2002). This brain region (anterior cingulated cortex) that is specialized in demand for control and reward is related to behavior activation (reaction to reward) and behavior inhibition (avoidance of punishment) systems (Gazzaniga, Irvy, & Mangun, 2009). Considering this shared neural structure, we think that investigating with the behavioral system developed by Gray should provide more insight in the relationship between cognition, personality and the content of voice behavior.
Even though the story about Bob, Barry and Santa Clause is fiction, similar situations are very common in organizations. Every employee has different ideas and behaves differently towards them, but they are still an important source for improvement for the manager. We want to know how we can stimulate employees to voice their ideas with good content. The goal of the present study is to investigate whether the amount of memory capacity is related to different behavioral systems and thereby influence the quality of voice in the business environment. Therefore we state the following research question;
To what extend do different behavioral systems, in explicit BIS and BAS, guide the
relationship between working memory capacity and voice quality?
Voice Behavior
The reason why employees do or do not speak up in certain types of situations when they have opinions, suggestions, ideas or concerns is a called voice behavior (Morrison, 2014). Within the literature there are three important commonalities for the definition of voice behavior; 1) voice is an act of verbal expression, where a message is conveyed from a sender to a recipient, 2) voice is a
discretionary behavior and 3) voice is constructive in its intent (Morrison, 2011). The most recent definition of voice is given by Morrison (2014):
I am defining employee voice as informal and discretionary communication of ideas, suggestions, concerns, information about problems, or opinions about work-related issues by an employee to persons who might be able to take appropriate action, with the intent to bring about improvement or change. (p. 174)
Morrison adds to this definition that voice is a form of extra role upward communication behavior towards the management that, although constructive in intent, challenges and seeks to alter the status quo. Extra-role behavior is the type of voicing that is not required in association with the current employment. Therefore it is not recognized by the formal reward systems (Van Dyne & LePine, 1998).
Another common phenomenon in the voicing literature is silence. In this action an employee decides to remain silent when he or she has a suggestion, idea, information, concern or a divergent point of view that could be useful or relevant to share (Morrison, 2014). Looking back to Bob, we see that silence does not mean the absence of a good idea. Barry however shows us that the absence of a good idea does not directly imply silence. Whether to voice or remain silent is an employee’s decision based on different motivators and inhibitors described by Morrison (Van Dyne, Ang & Botero, 2003).
If employees decide to voice there are differences in the frequency and the content, as we saw with Bob and Barry. These differences are better known as voice quantity and voice quality, where the first is about the frequency of the ideas and the latter about the content (Wolsink, 2013). We highlight this difference, because they may be facilitated by different cognitive processes. This study will focus on voice quality, which we will explain further by means of the dual process theory.
Quality of Voice Behavior by means of the Dual Process Theory
The dual process theory is the foundation for the idea that voice quality and quantity are facilitated by different cognitive processes. The theory states that there are two different modes of cognitive processing; one is fast, automatic, and unconscious and the other is slow, deliberate, and conscious (Evans, 2008). For clarification I will use the terms System I and System II in consensus with the article by Evans. In evolutionary perspective System I evolved earlier. The unconscious,
automatic, rapid and implicit cognition, responsible for much of our behavior and associated with System I, is the type of cognition that we share with other animals (Evans, 2008). Whereas the conscious, slow, explicit and controlled cognition in System II is responsible for higher cognitive tasks like language, higher-order control and activities as thinking about the future and counterfactual possibilities. The later characteristic of the human mind is one that we do not share with most other animals, and therefore must be evolved over time (Evans, 2008). The precise definition of System I and II is a debatable subject, but all theories share the common idea that thoughts, behaviors, and feelings result from the interplay between endogenous and exogenous forms of attention. Both types of attention can be applied to either increase or decrease the level of activation. Therefore, behavior is determined by the interplay of System I and II, respectively automatic and controlled processing (Feldman Barrett, Tugade & Engle, 2004).
This interplay also happens while voicing. The purpose of voicing is to make a constructive change in the status quo towards a better situation (Morrison, 2014). To do this efficiently there needs to be a high quality of voice, which is based on useful and original ideas and suggestions. Therefore, creativity plays an important role. Creativity is known as the generation of insights, ideas or problem solutions that are new and suitable (De Dreu et al., 2012). Creative performance can be influenced either by cognitive flexibility; fast and relatively effortless skipping through a large number of broad categories, or by persistence and perseverance; systematically scanning cognitive categories, using prolonged effort and time to creatively explore a smaller number of categories (De Dreu et al., 2008). What De Dreu et al. describe is the dual process theory in creative processes. Creativity benefits from the ability to focus on a problem for a longer period of time allowing persistent processing of information (De Dreu et al., 2012). This persistent processing is facilitated by the central executive of the working memory (Gazzaniga, Irvy, & Mangun, 2009).
Central Executive Functioning in Creativity
According to Feldman Barrett et al. (2004) working memory capacity (WMC) reflects individual differences in the ability to process information. Working memory is a type of short term memory with a limited-capacity store for retaining information over the short term and for performing mental operations on the contents of this storage (Gazzaniga, Ivry & Mangun, 2009). The information
in the working memory system can be acted on and processed, not merely maintained by rehearsal. Working memory acts like the central executive mechanism of the brain. It is the command-and-control center that is responsible for holding information available for complex activities (McVay & Kane, 2009).
The most prominent theory of working memory is Baddeley’s multi-component model (figure 1). This study focuses on the central executive (CE), which prioritizes and selects what information from the other components is
most relevant for the current behavioral goals. However, this selection is not always perfect. Failure to inhibit unnecessary information causes distraction (Kim et al., 2005). Therefore, the CE does most of the work of working memory and is responsible for the process that we consider as “thinking”, a reasoning process that allows the cognitive system to go beyond information readily available for the environment (Radvansky, 2011; Markman & Gentner, 2001). Differences in the persistent information processing, needed in creativity, arise from the functioning of the CE (De Dreu et al., 2012).
It is important to know that components 1, 2 and 3 are unconscious subsystems under control of a generalized conscious executive controller, component 4 (Radvansky, 2011). Only the information that is selected by the CE will become conscious, this shows again that the CE is responsible for ignoring distractors. This conscious versus automatic debate may ring a bell. Evans (2008) suggests that system II processes are those that require access to the CE, while system I processes do not require such access.
Summarizing, if employees use system II while voicing, they would need an active CE to make these. Wolsink (2013) already proved that, given the CE is located in working memory, there should be a difference in voice quality depending on WMC. We expect a replication of these results, therefore we hypothesize that;
Figure 1.
H1: Working memory capacity predicts the quality of voice; employees with high working
memory capacity will reach higher voice quality than employees with low working memory capacity.
Gray’s Behavioral Systems as Framework of Different Personalities
In our introduction we proposed the idea that besides the relationship between WMC and voice quality (H1), individual differences at the personality level may also play a role in the way
employees act in their business environment. One theory that describes how different people react differently to similar situations was developed by Gray. He believed that behavior does not only depend on the situation, but is also influenced by individual differences, also known as traits (Gray & Braver, 2002). He developed a difference between a behavioral inhibition system (BIS) and a behavioral activation system (BAS). BIS relates to
behavior that avoids negative outcomes, thus inhibits behavior that will result in punishment. BAS approaches positive outcomes, thus activates behavior that will result in reward (Berkman, Lieberman & Gable, 2008). Given these characteristics, BIS and BAS do not act on one scale, but rather as two dimensions in the same model (figure 2)(Vermeersch et al. 2011).
The difference between these two systems is important because they relate to the working memory capacity. Employees with more active BAS have greater influence in social situations due to better WMC (Lieberman & Rosenthal, 2001). Lieberman and Rosenthal state that more efficient WMC would translate into more effective multitasking, which results in a more active BAS. The contrary is true for employees with more BIS. Mostly they have the same skills but are unable to apply these in certain circumstances, especially when it involves multitasking. This is caused by the fact that their working memory gets occupied by external stimuli. BIS-employees react better cognitively, behaviorally and physiologically in the absence of external stimuli, whereas BAS-employees will become more active in cases of increased arousal (Lieberman & Rosenthal, 2001).
The Anterior Cingulate Cortex:
The link between Behavioral Systems and Working Memory Capacity.
The assumption made by Lieberman and Rosenthal has a neurophysiologic basis namely the Anterior Cingulate Cortex (ACC), an important neural structure in the working memory located in the medial prefrontal cortex (Gray & Braver, 2002). The ACC plays a role in autonomic bodily functions, as the regulation of blood pressure and heart rate, but also in higher cognitive functions (Gazzaniga, Irvy, & Mangun, 2009). These higher cognitive functions result in an activation of the bottom region of the ACC. This region is sensitive to the demand for control and is part of a larger network sub serving central executive functioning (Gray & Braver, 2002). We are interested in this cognitive control part because it tells us how an employee handles incomming information in deciding what is relevant and where to focus on.When WM load is high employees need a higher cognitive control to complete their tasks succesfully (Gray & Braver, 2002).
The ACC also plays an important role in personality (figure 3). Gray and Braver (2002) propose that the two behavioral dimensions of Gray reflect individual differences in two systems for control, either approach or
withdrawal. This perspective, suggesting that personality is related to cognitive aspects for control, is in agreement with Lieberman and Rosenthal’s view on the relationship between personality and WMC. Several neuroimaging studies also suggest associations between personality and activation of the bottom part of ACC during resting state (Johnson, et al., 1999). Gray and Braver (2002) conclude their article with the finding that, in brain activity, personality predicted task related activation in the bottom part of the ACC during a WM task. More specific, individuals with high BAS and low BIS have better WM performance and lower ACC activity. These results support the theories which state that personality is associated with differences in CE functioning (Carver et al., 2000). It also shows that personality is not only associated with resting state ACC activity, but also with differences in cognitive reactivity. Behavioral Systems WMC Anterior Cingulate Cortex
Looking back at hypothesis 1, we stated that WMC possibly predicts voice quality. Furthermore, we argued that one important part of WMC, the anterior cingulated cortex, is also important in the distinction between the two behavioral systems, because the differences in ACC activation during reward approach or punishment withdrawal. Therefore, it could be that behavioral systems guide the relationship between WMC and voice quality. This brings us to the following hypotheses;
H2a: Employees with lower WMC are more likely to have more active BIS and less active BAS.
Thus, WMC is negatively related to BIS.
H2b: Employees with higher WMC are more likely to have more active BAS and less active
BIS. Thus, WMC is positively related to BIS.
The Relationship between Behavioral Systems and the Quality of Voice
Assuming that indeed, CE functioning predicts the quality of voice (H1) and is positively
related to BAS (H2b) and negatively to BIS (H2a), we argue that perhaps, CE functioning influences
voice quality through these behavioral systems. As proven by Gray and Braver (2002) BAS-employees should have better WM performance and lower ACC activity. Also, Lieberman and Rosenthal (2001) argued that BAS-employees have better WMC and therefore are better in multitasking. On top of this, Wolsink’s (2013) findings suggest that better WMC leads to better quality of voice. Therefore, we argue that BAS-employees guide the relation between WMC and voice quality in a positive manner. This should be logical, considering that BAS-employees are focused on achieving better outcomes, which are more easily obtained with a high quality of voice.
On the contrary BIS-employees are characterized with being more worried. This results in an occupied working memory, leaving less working memory for immediate use and a lower ability to multitask. Considering Wolsink’s findings, we assume that this impaired WMC results in a lower voice quality. In other words, the different behavioral systems describe the psychological process that occurs to create the relationship between WMC and voice quality in the business environment. In other words, BIS and BAS have a mediating effect on this relationship. Therefore, we state the following hypotheses;
H3a: There is an indirect negative effect of WMC on voice quality mediated by BIS; employees
with more active BIS will have lower voice quality
H3b: There is an indirect positive effect of WMC on voice quality mediated by BAS; employees with more active BAS will have higher voice quality.
Design and Expectations
In this study we expect that, a) working memory capacity predicts the quality of voice, b) there is an indirect negative effect of WMC on voice quality mediated by BIS, c) there is an indirect positive effect of WMC on voice quality mediated by BAS and d) there is a positive indirect effect of WMC on voice quality mediated through BIS and BAS. A clear overview of the hypotheses is given in figure 4. To test these hypotheses we simulated an organizational environment, in which the participants got the opportunity to voice their ideas to the experiment leader, after which the ideas got rated on voice quality. Both WMC and BIS/BAS were also measured.
Figure 4. Hypotheses of this study
Method
Design and Sample
The study has a between subject design to test whether BIS and BAS mediating the relationship between WMC and voice quality. The dependent variable was voice quality, the independent WMC and the mediating variables BIS and BAS. Data was collected through a pre-test and an experiment completed by Dutch speaking students of the Faculty Economics and Business of
WMC
BIS
Voice
Qualit
y
H1+
H2
a-H2b
+
H3a-BAS
H3b+
the University of Amsterdam. We used these students because we wanted to create an organizational environment. We believe that their behavior towards their faculty is the same as their behavior towards their real working environment. Seventy-one students (45.1% women, average age 21.87, SD=1.392 all years of enrollment) successfully completed both the pre-test and the experiment.
Measurements, Tasks and Manipulations Measuring the Dependent Variable: Voice Quality
We measured voice quality with a voicing task developed by Wolsink (2013). Participants had to generate 10 ideas with improvements for the Faculty Economics and Business. They had to focus on suggestions which included solutions that were useful and original. Then the participants had the opportunity to voluntarily communicate ideas towards the experiment leader. Two raters rated the communicated ideas on a 1-5 Likert scale based on usefulness and originality (1 = not useful and original, 5 = very useful and original). High scores indicated high voice quality. An example of a suggestion is: ‘To improve student performance exams shouldn’t start earlier than 10.00 in the morning.’
Measuring the Independent Variable: Central Executive Functioning
To measure CE functioning we used WMC as defining factor. WMC was measured with an iPad app which focuses on visual and spatial representations developed by Ilja Sligte. In the first part of the experiment participants were shown a pattern with red and yellow balls for a very short moment. After the balls disappeared the participants had to repeat the pattern of the red balls. Alternating with the balls, a similar task was performed with tennis rackets, but instead of repeating the pattern the participants had to appoint which tennis rackets were rotated. With every move in the game participants earned or lost points. The game stopped after a certain amount of mistakes. The final amount of points earned is associated with the total WMC. Besides WMC, the app also measured WM Distraction, which indicates to what extend the CE selected the relevant information to focus on and ignored other information.
Measuring the Mediating Variables: BIS/BAS
To test our mediating variables we used a survey with the BIS/BAS Likert-scales developed by Carver and White (1994) (αBIS = .74, αBAS Reward Responsivenes = .73, αBAS Drive = .76 and αBAS Fun Seeking =
.66). The survey contained 20 questions about different behaviors of people divided in four subscales; BIS, BAS Drive, BAS Fun Seeking and Bas Reward Responsiveness. Participants were asked to rate to what extent these statements were applicable to them on a 7 point Likert-scale, 1 standing for ‘totally NOT applicable’ and 7 for ‘totally applicable’ (Appendix B). BIS questions (7x) focused on inhibitive behavior (‘I worry about making mistakes’). BAS Drive (4x) questions focused on the effort one is willing to take to achieve their goals (‘When I want something, I usually go all-out to get it’). BAS Fun seeking (4x) focused on how innovatory our participants were (‘I crave excitement and new sensations’). BAS Reward Responsiveness (5x) focused on how reward focused our participants were
(‘It would excite me to win a contest’) (Carver & White, 1994). Measuring the Control Variables
We created a demographics section in the survey for the sole purpose of giving background information of our participants. It contained seven questions concerning sex, age, attitude towards the Faculty Economics and Business, commitment and future plans which might involve the faculty (Appendix C). Besides the demographics we also controlled multiple times for affect state (Appendix A). We also controlled for the influence of the experiment leader during the voicing task. Perceptions were measured with a set of questions concerning the experiment leader (‘The experiment leader created a safe environment to take initiative’). Participants were asked to rate the statements with a
Figure 5. Measurement methods of the variables
Procedure and Materials
Pre-testParticipants were personally approached by e-mail, Facebook or during lectures, after which we made an appointment for the experiment. One week before the appointment the participants received the pre-test with the demographic and the BIS/BAS questions via e-mail. Thus, the survey was conducted by computer/telephone/iPad using internet on any possible time during that week. On behalf of professionalism and the ability to randomize questions, Qualtrics, an online survey site, was used to conduct the pre-test. Under the participants a price of € 50,- was raffled, as motivation to continue as participant.
Experiment
One day before the appointment participants received a reminder with an affect test (Appendix A). 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 experiment was conducted 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
WMC
BIS
Voice Quality
BAS
BIS Questionnaire (Carver and White,
1994) 1-7 likert,
7 items
BAS Drive and
Reward Responsiveness Questionnaire (Carver and White, 1994) Sports Q app by Ilja Sligte (2014) Useful and originality-ratings by experiment leaders
brief explanation of the experiment, in which was mentioned that it consisted two parts: two games on the iPad and 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. At first the participants were introduced to the ‘counting back task’ and were told to practice this (developed by Wolsink, 2014). The design of the task was as follows: the computers appointed the beginning of the task, thus the participant began with the number 100. Every two seconds the computer would make a beeping sound, after which the participants will say the next number with steps of two out loud (100, beep, 98, beep, 96, ...). If the task was completed successfully the participant received €2,50. After the rehearsal of the counting-back task the participants were told to generate 10 improving suggestions on the faculty of Economics and Business. The participants were told that they had the opportunity to communicate their ideas to the experiment leader, who would send the useful and original ideas to the faculty-management. All the ideas, communicated and not communicated, got rated by the experiment leader with a 1, being not useful and original, or a 2, being useful and original.
In organizational environments employees have in-role and extra-role behavior. To simulate the in-role behavior the participants had to fulfill a counting back task from 100 with steps of 1 (100, beep, 99, beep, 98 ...). If they completed this successfully they received € 2,50. During the countdown, 5 of the 10 generated ideas were shown with the opportunity to communicate to the experiment leader to simulate extra-role behavior. For every communicated idea ranked with a 2, useful and original, the participant earned € 0,50. For every communicated idea ranked with a 1 the participants loses € 0,50. After the computer-part the participant filled in a questionnaire about the experiment leader (appendix D) and received the money he/she earned.
Analysis and Predictions
The data was tested in a model based on a regression-based path analysis using PROCESS, an SPSS plug-in for testing various types of mediation and moderation (Hayes, 2008). The analysis was based on model 6; a serial multiple mediation model (Appendix E). With this model we test the indirect and direct effects of WMC using two mediators; BIS and BAS. In specific there are three indirect effects; 1) through BIS only, 2) through BAS only and 3) through BIS and BAS. The effect from WMC on voice quality is the direct effect. We tested these models by using the bootstrapping method with 5000 interactions, which means that the program tests 50% of the data sample, repeating the process for 5000 times, each time testing another half.
We predict a direct positive effect of WMC on voice quality. This means that we expect to find participants with higher WMC will communicate more useful and original ideas than participants with lower WMC. Furthermore, we predict a strong indirect positive effect of WMC on voice quality mediated by BAS. In other words, we expect that participants with more active BAS are more likely to voice useful and original idea. Furthermore we predict an indirect negative effect of WMC on voice quality mediated by BIS; employees with more active BIS are less likely to voice useful and original ideas. Last indirect effect is a mediation of both BIS and BAS on the relationship of WMC on voice quality. We expect a negative indirect effect.
Results
Reliabilities and correlations
Although we had 71 participants, we only used data from the participants who voiced, because of our dependent variable voice quality (N = 62). To test whether the measurements mentioned earlier can be used for a regression analyses a reliability analysis is conducted. Voice quality was rated by two independent judges, who scored the ideas on originality and usefulness. An inter-rater reliability test was conducted to see whether the raters gave more or less the same scores to ideas. For both aspects there was a strong positive relation between the experiment leaders; originality is r(620) = 0.690, p<0.01 and for usefulness r(620) = 0.539, p<0.01.
Table 1 shows the descriptive statistics, Cronbach’s alpha (on the diagonal) and correlations for all the variables. Some main variables are made from different variables; these are also shown in
the table. The alphas are considered reasonable to good. The alpha for BIS was originally α=.735, but by deleting two items (question 5 and 7) the alpha improved considerably to α=.815. We removed BAS Fun Seeking because of the low reliability (α=.500). BAS Drive question 2 was removed which increased α=.806 (Appendix B). BAS Drive and Reward Responsiveness together had a reliability of α=.788.
Table 1. Descriptives and correlations between the variables (Cronbach’s Alphas on diagonal)
Variables Pearson’s Correlation
M SD 1 1a 1b 2 2a 3 4 4a 4b 1 Voice Quality (DV) 2.690 .586 a. Usefulness 2.887 .636 **.646 b. Originality 2.492 .903 **.843 .134 2. WMC (IV) 7.259 .844 .004 -.186 .136 a. WM distraction (CV) .829 .957 -.207 -.077 -.214 .123 3. BIS (M1) 4.203 1.233 -.064 -.173 .038 .001 -.124 (.815) 4. BAS Total (M2) 5.333 .725 *-.288 -.070 *-.324 .027 *.272 .027 (.788) a. BAS Drive 4.914 1.088 *-.252 -.083 *-.301 .099 .114 -.114 **.859 (.806) b. BAS Reward Responsiveness 5.584 .686 -.247 -.083 *-.262 -.048 .154 .154 **.874 **.501 (.663) Note. N= 62. *p<.05. **p<.01,
Surprisingly, an inspection of the correlations suggest only one correlation between the main variables that is strong enough to be interesting; BAS has is negatively correlated with voice quality r(62)=-.288, p<.05. However, this correlation conflicts with our prediction that BAS facilitates voice quality. Besides the main effects of BAS and voice quality we also computed the correlations between the different components of BAS and voice quality. Interestingly, there is a difference between originality and usefulness; where usefulness is uncorrelated to BAS, originality correlates negatively with BAS in the framework with voice quality (r(62)=-.324, p<.05 with BAS total, r(62)=-.301, p<.05 with BAS Drive and r(62)=-.262, p<.05 with BAS Reward Responsiveness). The other variables are not significantly correlated with each other.
Main Effects: Testing the Mediation Model
As discussed in the analysis we used PROCESS model 6 with Bootstrap 5000 to test our mediation. Table 2 outlines the regression results for our hypotheses. Our model (figure 4) revealed an indirect effects model between WMC and voice quality mediated by BIS and BAS. To prove our hypotheses, there should be an overall positive indirect effect with BAS as moderator and a negative indirect effect with BIS as moderator. However, as depicted in Table 2, the only significant effect was in contradiction with H3b (β = -.232, p = .026). Given that all the results are either negative or not
Table 2. Regression Results for the Mediations in the Original model.
Model 1 Model 2 Model 3 Model 4
Outcome Voice Quality BIS BAS Voice Quality
Predictor β p β p β p β p
Intercept 4.192 *.003 5.097 **.000 3.980 **.000
WMC .008 .926 H2a .002 .993 H2b .023 .836 H1 .008 .926
BIS .016 .838 H3a -.027 .653
BAS H3b -.232 *.026
Bootstrap result for indirect effect on voice quality
Effect Boot SE LLCI ULCI
BIS .000 .015 -.031 .030
BAS -.008 .038 -.100 .058
BIS and BAS .000 .005 -.011 .010
Total -.008 .041 -.106 .063
Effect SE T p LLCI ULCI
Total effect of WMC on VQL (H2a* H3a+ H3a* H3b+ H1)
.003 .090 .030 .976 -.177 .182
Direct effect of WMC on VQL (H1) .008 .087 .094 .926 -.166 .183 Note. N = 62. *p<.05. **p<.01, Standardized regression coefficients are reported.
Bootstrap sample size=5000. CI=95%,
Post-Hoc analyses
Due to the found correlations between BAS, Originality and WM Distraction (table 1), we exploratory tested a more specific version of our earlier model. Since WM Distraction and BAS have a positive relationship and BAS and originality a negative relation, we tested a simple mediating model with originality as dependent variable, WM distraction as independent variable and BAS as mediating variable. In other words, do easily distracted employees higher BAS which leads to lower originality? For this analysis we used PROCESS model 4 (Appendix F), bootstrap 5000; simple mediation model. Table 3
shows that there is a significant positive relation between WMDist and BAS (β = .212, p = .030) and a significant negative relationship between BAS and originality (β = -.360, p = .028). However the bootstrap for the indirect effect concludes negative values, which indicates that there is no mediation effect (CI = [-.199, -.0113]).
Table 3. Regression Results for the Mediation in the Simple Mediation Model
Model 1 Model 2
Outcome BAS Originality
Predictor β β
Intercept **5.162 **4.507
WMDist. .*206 -.128
BAS *-.358
Bootstrap result for indirect effect Effect Boot SE LLCI ULCI
BAS -.074 .044 -.199 -.013
Effect SE T P LLCI ULCI
Direct effect of WMDist on Originality -.128 .120 -1.069 .289 -.3673 .112
Note. N = 62. *p<.05. **p<.01, Standardized regression coefficients are reported = β. Bootstrap sample size=5000. CI=95%,
Because little significant results were found some other tests with the control variables were done. Females score higher on BIS than male (female; M = 4.859, vs. male; M = 3.679). BIS correlates with sexes r(62) = .474, p < 0.05 (male = 0, female = 1). Other results indicated that full-time FEB students were often too late for an appointment compared and felt less committed to their faculty compared to normal students.
Discussion
Summary and Contributions
The main goal of this study was to prove that different behavioral systems, explicitly BIS and BAS, have an influence on the quality of voice behavior. We stated different hypotheses concerning a direct positive effect of WMC on voice quality, an indirect positive effect of WMC on voice quality mediated by BAS and an indirect negative effect of WMC on voice quality mediated by BIS. We did not find any mediation effects. The most important contribution of this study lies within the finding that BAS is negatively correlated to voice quality. This finding is in line with the current theories that personality has an effect on the quality of voice behavior (Bindle & Parker, 2010; Le Pine & van Dyne, 2001), but it conflicts with our prediction that BAS would increase the quality of voice. Although this was unexpected, it shows that the behavioral systems developed by Gray indeed measure something different than proactive personality or the Big Five and that further research needs to be done.
A second contribution of this study is that BAS correlates positively with the distractibility of employees, which means that employees with more active BAS are more susceptible to distraction. This emphasizes a link between the behavioral systems and WM. However, the correlation is in conflict with the statement of Lieberman and Rosenthal (2001) that a more efficient WMC would result in more active BAS. This confliction indicates that full understanding of this subject is not yet achieved. Presumable, different functions of WM relate differently to BAS and simple WM activation is not equal to individual differences in WMC. Another finding is the difference between originality and usefulness with respect to BAS. Only originality showed significant results with BAS. This shows that originality and usefulness are too distinct from each other to create one variable.
Extending Findings and Alternative Explanations
The findings of a negative influence of BAS on voice quality and the positive influence of WM distraction on BAS are, presumably, related to each other. First we look at the common factor in both relationships. The behavioral systems developed by Gray (Gray & Braver, 2002) have different underlying neural systems. BAS includes brain regions which are involved in the regulation of arousal with a special focus on reward, such as the cerebral cortex, thalamus and striatum (Berkman et al., 2008). BIS includes regions as the brain stem and neocortical of projections to the frontal lobe, which are involved in arousal with focus on non-rewarding stimuli (Vermeersch et al., 2011). These two systems are associated with individual differences in neural activity during a working memory task (Gray & Braver, 2002). Since participants with more BAS are more focused on reward, it is assumable that completing the experiment successfully was their primary task. The quality of their ideas was their secondary tasks. Therefore, this focus on reward got at expense of the quality. This is in line with findings of a field study by Wolsink (2013), showing that employees who plan their work extensively and in detail have lower voice quality ideas that employees who plan. Apparently, being very focused on getting your primary job done creates a tunnel vision that distracts attention from doing secondary things. A possible explanation for the fact that BIS did not correlate with WM distraction is that employees with more BIS, being more worried and social anxious than BAS- employees, are used to have an occupied working memory. Therefore the working memory of an employee with more active BIS is better trained and less impaired with distraction in contrast to BAS-employees, who have more difficulty with staying focused. Lieberman and Rosenthal found similar results (2001); the reaction time of BAS-participants during a memory task with distractions was much lower than the reaction time of BIS-participants. They conclude that prefrontal cortex distraction and impaired CE functioning result in different levels of brain activity for BIS and BAS people.
Alongside this explanation, we look back at the two systems of the dual process theory. Evans (2008) stated that thoughts, behaviors, and feelings result from the interplay between endogenous and exogenous forms of attentions. We assume that our participants had the intrinsic motivation to earn
money by completing the experiment successfully. However, we also found that BAS-employees are more subject to external stimuli and distractions (Carver & White, 1994). In other words, the CE also selects irrelevant external stimuli, making the endogenous and exogenous forms of attention conflict which results in less attention to the quality of voicing. This is in line with the article of Kane and Engle, which states that active maintenance of information is useful in many contexts, but it is critical under interference (exogenous attention). Interferences will result in slow responding and errors (2003).
Within the main finding of BAS correlating negatively with voice quality, there is a difference between the two components of voice quality; originality and usefulness. Whereas originality correlates negatively with BAS, usefulness does not correlate. We did not make a clear distinction between originality and usefulness in our theoretical framework, but this finding did not come as a surprise. Voice quality depends mostly on creativity, which depends on central executive functioning facilitated by the working memory, allowing persistent processing of information (De Dreu et al., 2012). Creativity is seen as a way to solve problems (Wolsink, 2013). Originality is an ability to think or express oneself in an independent and individual manner using creative ability. But usefulness, however, is the quality of having utility and especially practical worth or applicability (English Dictionary). Thus, originality and usefulness are quite distinct in their characteristics and controlled by different processes. Thereby, usefulness is much broader defined than originality. Maybe it would be convenient to create different categories of usefulness, investigate which characteristics an idea should have to qualify in certain usefulness-categories and which usefulness-categories are relevant for voice quality.
Limitations and Future Research
An obvious limitation of this research is the group of participants. With only 62 useful participants the power of the statistical tests was just sufficient or too low to really prove or reject our hypotheses. We tried to simulate an organizational environment where our participants, originally students, were proposed as employees and the Faculty Economics and Business had to function as the organization. Unfortunately, the fulltime students who participated felt little commitment to the faculty
and often came late for the appointment, indicating that our simulation of an organizational environment with dedicated employees, who really want to change the status quo towards a better situation, failed. In addition, students of the Faculty Economics and Business are not representative for average employees; therefore the results are not generalizable.
Furthermore, this study suffers from a validity problem. It is questionable if the experiment measured what we wanted it to measure. First of all, the experiment was not realistic enough. In the experiment participants were forced to contrive 10 ideas in a short time. In reality the construction of ideas happens with a less strict time limit and generally not on command. In addition, from the beginning of the experiment it is clear that one can receive a reward if a good performance is delivered. In reality the choice to voice or stay silent is a tricky one, because the implications are unpredictable (Morrison, 2014).
This is also a factor in the second limitation that concerns the elaboration of the behavioral systems. The focus of the experiment was to obtain money. Given that BAS-employees are characterized by their attitude to approach positive outcomes, it is likely that the experiment had the desired effect on this group. However, it is given that BIS has a focus on avoiding negative outcomes and responds mainly to anxiety relevant cues (Carver & White, 1994). But since the maximal negative outcome was the loss of only a couple of euros, it could be that this did not have a strong enough effect on our BIS-participants. In addition, many participants scored high on both the BIS- and the BAS-questionnaire, which made the distinction between the BIS and BAS group smaller.
Furthermore, we still do not know what really happened within the ACC during the experiment. Where there really differences in ACC-activity between the BIS- and BAS employees during the experiment? We took this the ACC as theoretical basis for our hypotheses, but we measured behavior instead of brain activity. Behavior represents the output of the whole system (Gray and Braver, 2002). Therefore, behavior is a relative insensitive measure. ACC activity, in contrast, is specific measure reflecting the demand for control.
To bypass these limitations in future research, we suggest to use a larger participant group with actual employees of an actual company to make sure that the participants are dedicated enough to want to change the status quo. Perhaps a company like Albert Heijn or another supermarket chain is eligible. Before constructing the study it would be useful to state a clear definition of voice quality or to use solely originality. In addition, a longitudinal study would make the construction of the ideas in an organizational setting more realistic. Also it should be possible to achieve a bigger negative outcome for the purpose of the desired effect on the BIS-employees. At last, a more multidisciplinary design could be useful to tell us more about what really happens concerning brain activity during these experiments. Therefore, collaboration between neuroscience and psychology would be convenient.
Practical Implications
With the results, the extended findings and the limitations taken in consideration we can safely say that our theory was not verified. We did not find a mediating effect of any behavioral system on the relationship between WMC and voice quality. Given the power of the statistical test, it is assumable that our small group of participants was one of the reasons for this outcome. Aside from the fact that we did not find as many results as we hoped, we did found something surprising. Employees with more active BAS are more subject to distractions, causing lower originality in voicing. Does this mean that organizations should hire only BAS-employees and make sure that they get no distraction? No, many other factors also play an important role. Therefore, further research is necessary.
But it does answer the question whether the act of voicing happens conscious or unconscious. In our introduction and theoretical framework we stated that some employees will voice towards better outcomes and that some employees stay silent to avoid negative outcomes. The fact that BAS-employees are subject to WM distractions shows WM activation during voice behavior. Evans suggested that system II processes require access to a single capacity limited working memory source (2008). In other words, voice behavior needs conscious calculation. This is in line with Wolsink’s theory (2013) that people with high voice quality use more systematic and persistent processes to reach their voice goal.
We wanted to say something about how to stimulate employees to voice their high quality ideas in practice. Conscious calculation takes place, our result suggest that different behavioral systems result in different calculations of cost and benefits of voice behavior. Therefore, it is necessary to make sure that no negative outcomes are possible to stimulate employees to voice their high quality ideas.
Concluding Thoughts
The goal of our study was to investigate whether different behavioral systems have different calculations of cost and benefits of voice behavior and thereby influence the quality of voice behavior in the business environment. We proved that BAS and BIS are associated with different levels of voice quality; employees who are strongly reward driven voice less quality than employees who are not focused on rewards. This is caused by the fact that BAS-employees are focused on their primary task and have difficulty focusing on a secondary task. Further research needs to shed light on differences in brain activity between BIS- and BAS-employees during different tasks. But first, we need clear definitions of voice quality, originality and usefulness.
References
Berkman, E., Lieberman, M., & Gable, S. (2008). BIS, BAS, and response conflict: Testing predictions of the revised Reinforcement Sensitivity Theory. Elsevier: Personality and Individual differences, 46, 586–591.
Bindl, U. K., & Parker, S. K. (2010). Proactive Work Behavior: Forward-Thinking and Change-Oriented Action in Organizations. Washington, DC: American Psychological Association.
Carver, C., & White, T. (1994). Behavioral Inhibition, Behavioral Activation, and Affective Responses to impeding Reward and Punishment: the BIS/BAS Scales. Journal of Personality and Social Psychology, 67, 319-333.
Carver, C., Sutton, S., & Scheier, M. (2000). Action, Emotion and Personality; Emerging conceptual integration. Personality and Social Psychology Bulletin, 26, 741-751.
De Dreu, C., Baas, M., & Nijstad, B. (2008). Hedonic tone and activation level in the mood-creativity link: Toward a dual pathway to creativity model. Journal of Personality and Social Psychology, 94, 739–756.
De Dreu, C., Nijstad, B., Baas, M., Wolsink, I., & Roskes, M. (2012). Working Memory Benefits Creative Insights Musical Improvisation, and Original Ideation Through Maintained Task-Focused Attention. Personality and Social Psychological Bulletin,38, 656-669.
Evans, J. (2008). Dual-Processing Accounts for Reasoning, Judgement, and Social Cognition. Annual Review of Psychology, 59, 255-278.
Feldman Barrett, L., Tugade, M., & Engle, R. (2004). Individual Differences in Working Memory Capacity and Dual-Process Theories of the Mind. Psychological Bulletin, 130,, 553–573. Gazzaniga, M., Irvy, R., & Mangun, G. (2009). Cognitive Neuroscience: The Biology of The Mind.
London: Norton.
Gray, J., & Braver, T. (2002). Personality predicts working-memory- related activation in the caudal anterior cingulated cortex. Cognitive, Affective & Behavioral Neuroscience, 2,, 64-75.
Hayes, A. (2008). Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York: The Guilford Press.
Johnson, D., Wiebe, J., Gold, S., Andreasen, N., Hichwa, R., Watkins, G., & Boles Ponto, L. (1999). Johnson,D.L.,Wiebe,J.S.,Gold,S.M.,Andreasen,N.C.,Hichwa, R.D.,Watkins,G.L.,&BoCerebral bloodflow and personality: A positron emission tomography study. American Journal of Psychiatry, 156, 252 - 257.
Kane, M., & Engle, R. (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), 47–70.
Kim, S., Kim, M., & Chun, M. (2005). Concurrent Working Memory Load can reduce Distraction. Proceedings of the National Academy of Sciences of the United States of America, 102, 16524 -
16529.
lePine, J. A., & van Dyne, L. V. (2001). Voice and cooperative behavior as contrasting forms of contextual performance:evidence of differential relationships with Big Five personality characteristics and cognitive ability. Journal of Applied Psychology, 86: 326–36. Lieberman, M., & Rosenthal, R. (2001). Why Introverts Can't Always Tell Who Likes Them:
Multitasking and Nonverbal Decoding. Journal of Personality and Social Psychology, 80, 294– 310.
Markman, A., & Gentner, D. (2001). Thinking. Annual Review of Psychology, 52,, 223-248.
Mc Vay, J., & Kane, M. (2009). Conducting the Train of Thought: Working Memory Capacity, Goal Neglect and Mind Wandering in an Executive-Control Task. Journal of Experimental Psychology, 35, 196-204.
Morrison, E. (2011). Employee Voice Behavior: Integration and Directions for Future Research. The Academy of Management Annals, 5,, 373-412.
Morrison, E. (2014). Employee Voice and Silence. Annual Review Organizational Pscyhology and Organizational Behavior, 1, 173-1197.
Radvansky, G. (2011). Human Memory. Boston: Pearson. Schmajuk, N. (2008). Classical Conditioning. Scholarpedia, 3, 3.
Van Dyne, L., & LePine, J. (1998). Helping and Voice Extra-role Behaviors: Evidence of Constructs and Predictive Validity. Academy of Management Journal, 41, 108-119.
Van Dyne, L., Ang, S., & Botero, I. (2003). Conceptualizing Employee Silence and Employee Voice as Multidimensional Constructs. Journal of Management Studies, 40, 1359–1392.
Vandervert, L., Schimpf, P., & Liu, H. (2007). How Working Memory and the Cerebellum Collaborate to Produce Creativity and Innovation. Creativity Research Journal, 19, 1-18.
Vermeersch, H., T'Sjoen, G., Kaufman, J., & van Houte, M. (2011). Social Science Theories on Adolescent Risk-Taking: The Relevance of Behavioral Inhibition and Activation. Youth & Science, 45, 27-53.
Wolsink, I. (2013). On the Differences Between Voice Actions and Outcomes: The Ability to Control Attention Facilitates. Master's thesis, University of Amsterdam, Netherlands.
Appendices
Appendix A: Affect-state questionnaire
In welke mate voel je je OP DIT MOMENT:
1. In zeer lichte mate 2. in lichte mate 3. Gemiddeld 4. In sterke mate 5. In zeer sterke mate Kalm Sereen Ontspannen Blij Opgetogen Uitgelaten Droevig Ontmoedigd Teleurgesteld Ongemakkelijk Gespannen Angstig Boos Gefrustreerd
Appendix B: BIS/BAS questionnaire
BIS questions:
1. If I think something unpleasant is going to happen I usually get pretty “worked up” 2. I worry about making mistakes
3. Criticism or scolding hurts me quite a bit
4. I feel pretty worried or upset when I think or know somebody is angry at me
5. Even if something bad is about to happen to me, I rarely experience fear or nervousness 6. I feel worried when I think I have done poorly at something
7. I have very few fears compared to my friends BAS Reward Responsiveness questions:
1. When I get something I want, I feel excited and energized 2. When I'm doing well at something, I love to keep at it 3. When good things happen to me, it affects me strongly 4. It would excite me to win a contest
5. When I see an opportunity for something I like, I get excited right away BAS Drive
1. When I want something, I usually go all-out to get it 2. I go out of my way to get things I want
3. If I see a chance to get something I want, I move on it right away 4. When I go after something I use a “no holds barred” approach BAS Fun Seeking:
1. I will often do things for no other reason than that they might be fun 2. I crave excitement and new sensations
3. I'm always willing to try something new if I think it will be fun 4. I often act on the spur of the moment
Appendix C: Demographic questions
Ik ben een: - Man - Vrouw Wat is je leeftijd? ….
Ben je voltijd student aan de Faculteit Economie en Bedrijfskunde? - Ja
- Nee
Hoeveel punten heb je gevolgd bij de Faculteit Economie en Bedrijfskunde? - Minder dan 30 punten
- Tussen de 30 en de 70 punten - Meer dan 70 punten
Ben je van plan om in de toekomst meer vakken bij de Faculteit Economie en Bedrijfskunde te gaan volgen?
- Ja - Nee
Hoe lang studeer je al aan de Faculteit Economie en Bedrijfskunde? - 1 jaar - 2 jaar - 3 jaar - 4 < jaar In zeer lichte mate 1. 2. 3. 4. 5. 6. In zeer sterke mate 7. In welke mate voel jij je
betrokken bij de Faculteit Economie en
Bedrijfskunde?
Appendix D: Control variable experiment leader
1. The experiment leader created a ‘safe enviroment’ to take initiative. 1-7 Likert: 1 = strongly disagree, 7 = totally agree.
2. The experiment leader seems like an honest person. 1-7 Likert: 1 = strongly disagree, 7 = totally agree.
3. The experiment leader appreciates initiative. 1-7 Likert: 1 = strongly disagree, 7 = totally agree.
4. I trusted the experiment leader.
