Creativity under output constraints of expected specifications: The mediating effect of intrinsic motivation and the moderating role of certain personality traits
Master Thesis
Submitted by: Jingyue Wen (13329391) Supervisor: Dr. Philip Eskenazi
MSc in Business Administration – Strategy Track University of Amsterdam, Amsterdam Business School June 23rd, 2021, Final version
EBEC approval number: 20210317060335
Statement of Originality
This document is written by Student Jingyue Wen who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Contents
ABSTRACT ... 5
1. INTRODUCTION ... 6
2. LITERATURE REVIEW ... 10
2.1. CREATIVITY UNDER CONSTRAINTS ... 10
2.2. THE RELATIONSHIP BETWEEN EXPECTED SPECIFICATIONS AND CREATIVITY ... 11
2.3. INTRINSIC MOTIVATION AS MEDIATOR BETWEEN EXPECTED SPECIFICATION AND CREATIVITY ... 13
2.4. CERTAIN PERSONALITY TRAITS AS MODERATORS BETWEEN EXPECTED SPECIFICATION AND INTRINSIC MOTIVATION ... 15
2.5. THE RELATIONSHIP BETWEEN CERTAIN PERSONALITY TRAITS AND CREATIVE PERFORMANCE... 18
3. METHOD ... 20
3.1. GENERAL RESEARCH APPROACH ... 20
3.2. SAMPLE... 21
3.3. MEASUREMENT OF VARIABLES ... 21
3.3.1. Dependent variable, creative performance ... 21
3.3.2. Independent variable, output constraints of expected specifications ... 23
3.3.3. Mediating variable, intrinsic motivation ... 24
3.3.4. Moderating variable, conscientiousness and openness to experience ... 24
3.3.5. Control variables ... 25
4. RESULTS ... 26
4.1. DESCRIPTIVE STATISTICS AND CORRELATIONS ... 26
4.1.1. Missing value and recoding ... 26
4.1.2. The computation of the dependent variable ... 26
4.1.3. Reliability analysis ... 27
4.1.4. Correlation analysis ... 28
4.2. HYPOTHESES TESTING ... 31
4.2.1. Main effect of expected specifications for creative performance ... 31
4.2.2. Mediating effect of intrinsic motivation on the relationship between output constraints of expected specifications and creative performance ... 31
4.2.3. Moderating effects of conscientiousness and openness to experience on the relationship between expected specifications and creative intrinsic motivation ... 32
4.2.4. Direct effects of conscientiousness and openness to experience for creative performance ... 35
4.3. SUMMARY OF HYPOTHESES ... 36
5. DISCUSSION... 37
5.1. DISCUSSION OF THE RESULTS ... 37
5.2. THEORETICAL AND PRACTICAL IMPLICATIONS ... 39
Theoretical implications ... 39
5.2.2. Practical implications ... 40
5.3. STUDY LIMITATIONS AND FUTURE SUGGESTIONS ... 42
6. CONCLUSION ... 44
7. REFERENCES ... 45
8. APPENDIX... 53
APPENDIX A:INSTRUCTIONS FOR THE EXPERIMENT... 53
APPENDIX B:MEASURES AND ITEMS ... 62
Abstract
Creativity is a topic of ever-increasing interest in business, given the fact that it’s crucial for an organization’s survival and development (Anderson, Potočnik & Zhou, 2014). This study focused on one of the output constraints that could kill creativity. To be specific, it explored how output constraints of expected specifications affect creative performance through the mediating mechanism of intrinsic motivation, and how certain personality traits moderate the relationship between expected specifications and intrinsic motivation. This study not only partially filled the gap in existing literature but was also of value for managers. The Data were collected from an online experiment where participants were randomly allocated to two groups: one performing creative tasks without any expected specifications, and the other with strict expected specifications. The results indicated significant negative main effect of expected specifications on creative performance, and significant positive direct effects of the two personality traits, conscientiousness and openness to experience on creativity, although the former one was unexpected. The results also indicated no significant mediating effect of intrinsic motivation between expected specifications and creative performance, and no significant moderating effects of the two personality traits.
Key words: creativity, output constraints, expected specifications, intrinsic motivation, conscientiousness, openness to experience.
1. Introduction
Creativity is defined as ‘‘novel and useful products, ideas, or procedures that provide an organization with important raw material for subsequent development and possible implementation’’ (Oldham & Cummings, 1996, p. 607). Both academics and
practitioners believe that a company's survival and development depend on its ability to innovate and create, whether that creativity is with high originality, introduces groundbreaking innovation or not (Anderson, Potočnik & Zhou, 2014). It has long been thought that generating creative ideas and turning them into innovation is a way for a company or an organization to obtain competitive advantage, which is with high importance (Anderson, Potočnik & Zhou, 2014; Hye, Sun, Jin, Kyungmook & Seongsu, 2015; Chen, Farh, Campbell-Buth, Wu & Wu, 2013).
However, there are always constraints such as time pressure, scarce resources, lack of task autonomy, and regulations that affect creativity through different routes. The effects of a variety of constraints that trigger diverse mediation mechanisms such as intrinsic motivation, opportunity identification, and interaction anxiousness on creative performance have attracted interests across the fields of strategic management, industrial organization, entrepreneurship, and marketing (Acar, Tarakci &
Knippenberg, 2019). The prior research is fragmented, and sometimes yield conflicting results, leaving fertile ground where there is remained unexplored.
Acar et al. (2019) divides the constraints into three clusters, including input constraints, process constraints and output constraints. The role of input constraints in creativity has received widespread attention and has been the most common and well-
studied constraint type, spanning the strategic management spectrum (Acar et al., 2019).
There are also many research focusing on the process constraint type, especially the role of formalization and job autonomy (Liu, Jiang, Shalley, Keem, & Zhou, 2016;
Evanschitzky, Eisend, Calantone, & Jiang, 2012; Oldham & Cummings, 1996;
Damanpour, 1991). However, prior research on output constraints is not as much as the other two types, which has mainly focused on the role of standards and regulations at firm, industry, and country level (Acar et al., 2019). In response to the scholarly call for more research on the creativity under output constraints at an individual level, this study examines how output constraints of expected product or project specifications affect individual creativity through the mediating mechanism of intrinsic motivation.
In addition, this study proposes that certain personality traits can moderate the relationship between expected specifications and intrinsic motivation.
According to Amabile (1988) and Amabile (1996), external constraints have a decidedly negative connotation in general. First, this study proposed that output constraints of expected specifications, one of the external constraints, have a negative effect on creative performance. Prior research also found that creators can lose intrinsic motivation when confronted with external constraints engaging in creative activities, which results in lower level of creativity (Rosso, 2014; Amabile, Hadley, & Kramer, 2002; Andrews & Smith, 1996). Given the characteristics that intrinsic motivation is both closely related to expected specifications and creativity, this study then proposed that intrinsic motivation mediates the relationship between the output constraints of expected specifications and creative performance. Furthermore, considering the scarce
explore of the moderating role of contextual factors between constraints and mediating mechanisms, this study also brought two personality traits as moderators into the model.
Based on the characteristics that individuals scoring high in conscientiousness or openness to experience have (Barrick et al., 2002; Barrick, Mount, & Strauss, 1993;
Burke & Witt, 2002), this study proposed that the relationship between expected specification and intrinsic motivation is moderated by conscientiousness and openness to experience. In addition, this study proposed that conscientiousness has a negative effect on creative performance, while openness to experience has a positive one.
In order to test these hypotheses, an online experiment was conducted, in which participants were randomly allocated into two groups to perform creative tasks, namely generating creative titles for two short stories (Eisenberger and Rhoades, 2001).
Participants in one group were confronted with strict expected specifications, while the other with no expected specifications at all. By doing so, the independent variable was manipulated and consisted of two levels. The dependent variable was measured by the originality of what participants generated, which were scored by two independent judges. The mediating variable and moderating variables were measured by scales developed by Jimmy, Tamiko, & Alice (2010) and Donnellan et al. (2006) respectively.
The results indicated significant negative effect of expected specifications on creative performance, and significant positive direct effect of openness to experience on creativity, which was consistent with what was proposed. However, the results also indicated significant positive effect of conscientiousness on creativity, which was unexpected. In addition, the results indicated no significant mediating effect of intrinsic
motivation between expected specifications and creative performance, and no significant moderating effects of the two personality traits.
This study contributes to the literature and practice in at least three ways. Firstly, this is the first study that empirically examines how output constraints affect creativity through a motivational route. According to Acar et al. (2019)’s cross-disciplinary integrative framework, there’s no empirical evidence on how output constraints affect creativity through motivational route or social route. Rosso (2014) also indicates that there’s limited empirical research on constraints as inhibitors of creativity from the perspective of intrinsic motivation. Secondly, there’s little prior research that considered the moderators between output constraints of creativity and mediating mechanisms. Upon reviewing the relevant literature, it can only be found that Medeiros, Partlow & Mumford (2014) highlighted the employees’ need for cognition as a moderator at an individual level. Thus, this study is also the first one empirically examines how certain personality traits moderate the relationship between expected specifications and creative intrinsic motivation. Thirdly, considering that creativity is crucial for a firm’s survival and development (Anderson, Potočnik & Zhou, 2014), it’s important for managers to know how not to kill creativity at an individual level in practice. This study gives useful suggestions on it after unraveling the mystery of the relationships among output constraints of expected specifications, certain personality traits, intrinsic motivation and creative performance.
2. Literature review
Prior relevant research will be discussed in this section, supported by which, the hypotheses will be explained, and the conceptual model will be presented.
2.1. Creativity under constraints
Constraints can be defined as “externally imposed requirements on the requisite characteristics of creative problem-solutions” (Medeiros, Partlow, & Mumford, 2014, p. 198; Weisberg, 2011). Based on the widely used input-process-output model, Acar et al. (2019) presents a taxonomy of constraints and mediating mechanisms, as well as an integrative synthesis of how constraints affect creativity. According to Acar et al.
(2019)’s cross-disciplinary integrative framework, there are three clusters of constraints including input constraints, process constraints and output constraints, meaning restraints on the input to be used, steps and processes to be followed, and factors of the end result to be met for creativity, respectively. In addition, the framework also presents three clusters of mediating mechanism/ route, including motivational route, cognitive route, and social route, which refers to mechanisms related to the motivation in engaging in creative activities, cognitive process of creativity, and interactions among individuals and teams in creativity-related activities, respectively (Acar et al., 2019).
Output constraints were often seen negatively in early research (e.g.Ambec, Cohen, Elgie & Lanoie, 2013). Standards were also thought to promote uniformity while reducing variety, thus would inhibit creativity (Gilson, Mathieu, Shalley & Ruddy, 2005). Some recent research argues for a positive depiction of output constraints. For example, de Vries & Verhagen (2016) and Xie et al. (2016) argue that standards can
help to foster innovation by channeling attention towards new knowledge and providing reliable interfaces to build on (Acar et al., 2019). In addition to some conflicting conclusions as discussed above, the empirical evidence examining the mechanism through which output constraints affect creativity is scarce, especially at individual level (Acar et al., 2019). To be specific, there’s only Moreau & Dahl (2005)’s research that have tested deviation from path of least resistance as the cognitive mediating mechanism between output constraints and creativity. According to Acar et al. (2019)’s integrative framework, there is a lack of empirical evidence on how output constraints affect motivational and social mechanisms. Thus, this study examining the effects of output constraints of expected specification on creativity through the intrinsic motivational mechanism at an individual level can partially fill the gap.
2.2. The relationship between expected specifications and creativity
Expected specification, a type of output constraints imposed externally, gives the expected elements or requirements that should be met by the end result, for example, a specific design element that should be presented in the product, which narrows or defines the expected outcomes of product or project (Acar et al., 2019; Rosso, 2014).
According to Acar et al. (2019)’s integrative framework, prior research on regulations and standards has primarily concentrated on firm-, industry-, and country-level, whereas research on expected product or project specifications has a tendency to focus on individual- and project-level. This study that examines the effect of output constraints of expected specifications on creativity also focus on individual level. In
addition, Rosso (2014) points out that limitations connected to the requirements of the expected project or product is one of the most commonly identified type of constraints.
As shown in the definition, expected specifications are often externally imposed by managers or leaders, in order to meet specific requirements, such as customer needs (Rosso, 2014). Defined as “factors introduced by the social environment that are intended to control an individual’s engagement in a task” (Rosso, 2014, p.553), external
constraints have a decidedly negative connotation in general (Rosso, 2014; Amabile, 1996; Amabile, 1988). To be specific, Amabile (1988) conducted an interview study to examine factors influencing creativity and innovation in organizations, and he found that external constraints in deciding what to do or how to accomplish the task inhibit creativity. In addition, Amabile (1996) did quantitative analysis to assess the work environment for creativity, and also found that external constraints of freedom or workload pressure could undermine creativity. The negative connotation discussed above that external constraint could bring, focuses on the lack of creators’ sense of
control over the creative activity and output, which also works in expected specifications.
What’s more, similar as constraints related to market needs and intellectual property do, expected specifications have an impact on creativity by limiting creative solution’s options (Rosso, 2014). For example, individuals and teams might have to discard some potentially interesting and creative ideas in order to meet the expected product or project specifications (Rosso, 2014). Furthermore, output constraints of expected specifications also limit the participants’ imagination by shaping the
boundaries during the creative activities, which can be also negatively related to creative performance. Thus, based on the literature reviewed above, there’s a prediction:
H1: The output constraints of expected specifications have a negative effect on creative performance.
2.3. Intrinsic motivation as mediator between expected specification and creativity Given that output constraints of expected specifications potentially have a negative impact on creative performance, it is important to identify the mediating mechanism through which they affect creativity. Intrinsic motivation is defined as the desire to engage in an activity just for the sake of the activity itself, to enjoy doing it without any expectations, or with the aim of obtaining external rewards (Malik, Choi,
& Butt, 2019; Lepper, Greene, & Nisbett, 1973). In addition, Malik, Choi, & Butt (2019) operationalize intrinsic motivation as a task-specific situational motivation, representing a task setting where the participants are motivated to indulge in creativity by the task itself.
The componential model of creativity proposed by Amabile (1988) implies that intrinsic motivation, the desire to perform an activity due to its inherently interesting or involving features, is critical to creative output (Amabile, 1996; Amabile, 1988). The model has influenced the relevant literature heavily. Prior research often posits that there’s a positive correlation between creative intrinsic motivation and creative performance (Malik, Choi, & Butt, 2019; Yoon, Sung & Choi, 2015; Grant & Berry, 2011). Shin, Yuan, & Zhou (2017) point out that the behavior driven by the intrinsic motivation that comes from the task itself, induces a deep and sustained involvement
in the creative activity. Intrinsic motivation is typically linked to interest, curiosity, and a drive to learn, which makes an individual expand effort on a creative task initiatively, and it is also thought to boost an individual’s creativity by enhancing the positive effect, cognitive flexibility, risk taking, and perseverance (Grant & Berry, 2011).
However, output constraints externally imposing requirements for the end result can reduce such intrinsic motivation, which leads to the potential to impede creativity by reducing the individual’s desire to explore without boundaries, the interest, the curiosity, the cognitive flexibility and risk taking. Rosso (2014) concluded from Amabile’s (1988) theory and other studies on this topic that the intrinsic motivation,
positively correlated to creativity, is fostered by freedom, and inhibited by constraints (Rosso, 2014). In order to examine the role of different categories of constraints in creativity, Rosso (2014) did a field research with real R&D teams who were engaged in ongoing creative projects in an organization. Through the interviews, he found that both process and product constraints tended to inhibit creativity when reducing intrinsic motivation or experimentation (Rosso, 2014). Those external constraints such as output constraints/ product constraints often come with lack of freedom in deciding what to do or how to accomplish the task, which can reduce one’s intrinsic motivation and thus inhibit the creative performance. Other proponents of this perspective argue that creators can lose intrinsic motivation, fall back on routines and surface-level thinking when confronted with external constraints engaging in creative activities (Amabile, Hadley, & Kramer, 2002; Andrews & Smith, 1996). For example, Amabile et al.
(2002)’s study indicated that the more external constraints of time pressure that an
individual feels on a given day, the more likely he/she feels frustrated, and the clearly lower level of creative thinking he/she shows in evidence. Likewise, expected specifications as an external constraint can also negatively affect creativity by influencing one’s “feeling” and intrinsic motivation.
Based on the characteristics of a mediating variable (Barron & Kenny, 1986), the above discussion suggests that (a) intrinsic motivation is closely related to output constraints of expected specifications, and (b) it is also strongly associated with creative performance. Specifically, output constraints of expected specifications can negatively affect an individuals’ creative performance through reducing their intrinsic motivation.
Thus, based on the literature reviewed above, there’s a prediction:
H2: The intrinsic motivation mediates the relationship between the output constraints of expected specifications and creative performance.
2.4. Certain personality traits as moderators between expected specification and intrinsic motivation
Some researchers (e.g.Ana et al., 2021; Johns, 2018) agree that the importance of context in organizational research has been undervalued. Similarly, most research on the link between output constraints and creativity has mainly focused on the direct effects rather than delving further into the moderating role of some contextual factors such as organizational culture, control mechanism, supportive leadership, and personality traits (Ana et al., 2021). This study focuses on an individual level and chooses to explore the moderating role of certain personality traits. All personality measures, according to Goldberg (1990), can be classified into 5 factors, labelled as the
“Big-five” factors, including extraversion, conscientiousness, agreeableness,
neuroticism, and openness to experience (Costa & McCrae, 1992; Goldberg, 1990).
Different individuals with different personality traits have different reactions towards the same constraints.
As mentioned, this study explores certain personality traits, to be specific, the conscientiousness and the openness to experience’s moderating effects between output constraints of expected specifications and intrinsic motivation. Choosing to further explore on these two personality traits rather than the other three not only because the scarce explore on them as the moderator between expected specifications and intrinsic motivation, but also because they are more relevant to output constraints, intrinsic motivation, and creativity. For example, conscientiousness has been related to creative performance many times before (e.g. Sung & Choi, 2009). When it comes to openness to experience, it is often associated with the flexibility that situations where there’re output constraints lack, which makes it also highly relevant. These will be explored in more detail below.
Salgado (1997) indicates that individual’s conscientiousness is measured by how organized, hardworking and dependable they are. Costa & McCrae (1992) also describe people with high conscientiousness as reliable, determined and achievement oriented.
As mentioned before, output constraints of expected specifications are often externally imposed by managers or leaders, with the aim of meeting specific requirements, which can reduce the intrinsic motivation related to creativity. However, according to Barrick, Mount, & Strauss (1993)’s research, the sales representatives with high
conscientiousness are more likely to be committed to accomplishing the set goals in order to achieve better performance that they expected. Barrick et al. (2002) also points out that individuals scoring high in conscientiousness have desire to exercise self- control and seek to fulfil the obligations. This would suggest that conscientiousness can moderate the relationship between output constraints of expected specifications and intrinsic motivation. To be specific, individuals scoring high in conscientiousness are more likely to be positive towards facing and accepting the externally imposed requirements, and they are also more likely to continue being hardworking and organized without losing too much intrinsic motivation, in order to fulfill the goals.
As for openness to experience, it is stated to be the most relevant trait with the most empirical evidence linking it to creativity (Feist, 1998; McCrae & Costa, 1997).
Openness to experience is described as measuring the creativity, curiosity, and flexibility of an individual (Salgado, 1997; Judge & Ilies, 2002). It also reflects characteristics such as imaginativeness, intellectual curiosity and broadmindedness (Costa & McCrae, 1992). Burke & Witt (2002) indicates that individuals with high openness to experience prefer to generate insights independently and control their own work, which indicates that they prefer independence and autonomy, and are more likely to be negative when facing externally imposed requirements, for example, facing the output constraints of expected specifications. Thus, based on the literature reviewed above, there are two predictions:
H3a: The relationship between expected specification and intrinsic motivation is moderated by conscientiousness. The relationship is weaker among individuals high in conscientiousness.
H3b: The relationship between expected specification and intrinsic motivation is moderated by openness to experience. The relationship is stronger among individuals high in openness to experience.
2.5. The relationship between certain personality traits and creative performance As mentioned, individual’s conscientiousness is measured by how organized, hardworking and dependable they are (Salgado, 1997). Individuals with high conscientiousness are often more purposeful and strive for achievement, which results in more focus on how to carry out a given work in an efficient and structured manner rather than introducing interruptions such as seeking problems or new opportunities in the existing work or generating new ideas that can interrupt the process of accomplishing the given task (George & Zhou, 2001; Goldberg, 1990). What’s more, conscientious people have been found to be less willing to take risks, to explore and do experiments that creative activity needs, due to that they are more inclined to avoid uncertainties that can lead to unexpected delays or unforeseen setbacks (James &
Mazerolle, 2002). Thus, conscientiousness is often negative to creative performance.
Individuals scoring high in openness to experience are more broad-minded, curious and flexible in embracing novel ideas and exploring new things (Mount &
Barrick, 1995; Sung & Choi, 2009). On the contrary, people scoring low in openness to experience are more cautious and conservative thus less interested in exploring new
things or generating new ideas, because they are more comfortable with the status quo and prefer familiar ideas and activities in order to control uncertainty (Sung & Choi, 2009; George & Zhou, 2001). In addition, by conducting empirical testing, Sung &
Choi (2009) examined that openness to experience significantly positive related to creative performance (r = .26, p < .01). Prabhu, Sutton, & Sauser (2008) also found a significant positive correlation between openness to experience and creativity (r = .35, p < .01). Thus, based on the above, there are two predictions:
H4a: Conscientiousness has a negative effect on creative performance.
H4b: Openness to experience has a positive effect on creative performance.
3. Method
The general research approach will be explained in this section. Furthermore, the sample will be discussed, and the measurement of the variables will be presented.
3.1. General research approach
This study was conducted through an online experiment, in which the output constraints of expected specification is the independent variable to be manipulated. In addition, this research had a one-way between-subjects factorial design with two levels.
To be specific, there were two groups of participants to engage in the experiment.
Participants in one group were confronted with no expected specifications at all, while people in the other group had to meet strict expected specifications when performing creative tasks.
Conducting an online experiment has many advantages. First, it is cost efficient.
Second, it also helps to get access to a diverse sample of participants in terms of age, gender, nationality, education, and work experience. Last but not least, it is the safest way when facing the Covid-19 epidemic. The experiment of this study was designed in Qualtrics, which is an online survey and experiment software. In addition, the experiment was conducted via the platform Amazon Mechanical Turk, members of which are paid by researchers to participate in studies such as experiments. This research has been approved by The Economics & Business Ethics Committee (University of Amsterdam) with the number 20210317060335.
3.2. Sample
The group of participants (N = 164) consisted of 118 males, 44 females, and 2 who prefer not to say their gender. The most common age group was between 26 and 30 years old, where there were 73 participants (44.51%). In addition, 39 participants were 31-35 years old (23.78%), and 31 participants were 18-25 years old (18.90%). The majority held a bachelor’s degree (64.02%) or a master’s degree (22.56%). As for the
work experience, 95 participants fell in the 0-5-year range (57.93%), 39 and 15 participants fell in the 6-10 and 11-15-year range (23.78% and 9.15%) respectively.
3.3. Measurement of variables
3.3.1. Dependent variable, creative performance
As Liu et al. (2016) concludes, there are many existing measures of creativity, including both perceptual measures and objective measures. Perceptual measures mainly consist of that rated by employees themselves, coworkers, supervisors, and expert judges (e.g., Shalley, Gilson, & Blum, 2009; Binnewies, Ohly, & Sonnentag, 2007; Amabile et al., 2005; Tierney & Farmer, 2002; George & Zhou, 2001). As for objective measures, Liao, Liu, & Loi (2010) used that, for example. Considering the need of manipulation of the independent variable expected specifications and also the aim to reduce biases, the dependent variable of this study, creative performance, was measured by scores of the creativity task judged by “experts”.
There are some previous studies that have developed different methods to measure creative performance through creativity tasks. Eisenberger and Rhoads (2001) developed some experimental tasks. For example, one is to let participants come up
with names for a film about students’ summer break, and another is to let participants list five possible titles for two short stories. Sellier and Dahl (2011) developed a task where participants create scarf and Christmas tree ornament through a field experiment.
Hester et al. (2012) conducted an experiment where the participants need to develop a written marketing plan for a new high-energy root beer and describe their advertising campaign for it. Medeiros, Partlow, & Mumford (2014) also took inspiration from it.
What’s more, based on Moreau & Dahl (2005) and Finke et al., (1992)’s previous creativity studies, Scopelliti, Cillo, Busacca, & Mazursky (2014) developed a creative task, which is to let participants design a new toy for children aged 5 to 11.
Considering the situation of conducting an online experiment under the influence of the Covid-19 epidemic, to shorten the time of the experiment, and to have more people participating in it, the creativity task as applied by Eisenberger and Rhoades (2001) was used, which means to let participants list possible titles for two short stories.
In addition, Eisenberger and Rhoades (2001) also found that whether participants were allowed to name as many titles as they wanted to or were asked to generate a set number of responses, the results were identical. Based on the above, participants were aksed to generate a maximum of three titles for each story in the experiment of this study in order to keep their concentration.
The scores of the creativity task were determined by the originality of what participants generated, judged by two master students majored in Business Administration who are familiar with the literature both in creativity and business. The
judges were completely unaware of the participants’ circumstances, and they rated each participants’ work from 0 (no creativity) to 5 (highly creative) independently. Derived from Medeiros, Partlow, & Mumford (2014)’s Benchmark rating scales for originality of marketing plans, the scales for rating the originality of titles are displayed below.
Table 1
Benchmark rating scales for originality of titles generated
Ratings 1 3 5
Definition The title is predictable and fails to provide any new or unique ideas.
The title lacks richness and descriptiveness.
The title has a few original and unique elements; however, it still contains many predictable concepts.
The title is clearly unique and has core elements that appear is exceptionally rich and descriptive.
3.3.2. Independent variable, output constraints of expected specifications
The independent variable was not measured but was manipulated and consisted of two levels. One was no output constraints of expected specifications at all, the other strict expected specifications. To be specific, for the no-expected-specification group, the instruction was:
“You are about to read two short stories on the next pages. Please read the story carefully. When you finish reading, please write down three possible creative titles that fit the short story. There’re no constraints on the titles at all, and it is important that you try to come up with the most creative relevant titles possible.”
For the strict-expected-specification group, the instruction was:
“You are about to read two short stories on the next pages. Please read the story carefully. When you finish reading, please write down three possible creative titles that fit the short story. Note that each title that you created should meet two requirements:
1) within 5 words; 2) contains at least one color. It is important that you try to come up with the most creative relevant titles possible that meet the requirements.”
3.3.3. Mediating variable, intrinsic motivation
To measure the level of participants’ task-specific intrinsic motivation, there were seven questions about the interest and enjoyment of the activity derived from IMI (Intrinsic Motivation Inventory) after the creativity task, based on the Self-determinism Theory (SDT) of motivation. This measurement was also used by Ryan (1982), Ryan, Koestner, & Deci (1991), and Jimmy, Tamiko, & Alice (2010). Example item is “This task was fun to do” (Jimmy et al., 2010). All items of the scale that were used to assess participants’ intrinsic creative motivation towards the task can be found in Appendix B.
3.3.4. Moderating variable, conscientiousness and openness to experience
The moderating variable was measured by 8-item short survey derived from Mini- IPIP scales developed by Donnellan, Oswald, Baird, & Lucas (2006). Participants often find it boring or frustrated in completing a long questionnaire, which can lead to a negative mood or careless attitude towards the experiment (Donnellan et al., 2006).
Using such a short survey can reduce the transient measurement errors (e.g., Schmidt, Le, & Ilies, 2003) and also help in keeping the focus of the participants, compared to the longer existing 50-item International Personality Item Pool -- Five-Factor Model
measure (Donnellan et al., 2006; Goldberg, 1999). Moreover, Donnellan et al. (2006) also verified the validity of the Mini-IPIP scales, with four items per trait, which concluded that it is psychometrically acceptable and practically useful. Example item of conscientiousness is “Like order”, and example item of openness to experience is
“Have a vivid imagination” (Donnellan et al., 2006). All items of consciousness and openness to experience (also mentioned as Intellect/ Imagination in Donnellan et al., 2006’s paper) according to the Mini-IPIP scales can be found in Appendix B.
3.3.5. Control variables
At the end of the experiment, there were four questions about factors that might affect the propensity of participants to generate creative titles for the two short stories.
The four demographic questions to measure the control variables were age, gender, the highest level of education, and working experience. Instructions for the experiment in detail can be found in Appendix A.
4. Results
4.1. Descriptive statistics and correlations 4.1.1. Missing value and recoding
First of all, the participants who did not finish the experiment were excluded from the sample. The questions in the online experiment were set on forced answering mode, thus participants who completed the experiment succeeded in answering all questions.
Of the 283 participants who initially started the experiment, 164 completed validly and were used for the analysis. In addition, the answers to questions 5 and 9 in measuring intrinsic motivation, and questions 12, 13, 15, 16, and 17 in measuring conscientiousness and openness to experience that contained reversed statements were reversed coded before analyzing.
4.1.2. The computation of the dependent variable
As mentioned in the method chapter, to measure the independent variable creative performance, the two judges were asked to assign scores from 0 (no creativity) to 5 (highly creative) for each title based on the originality of what participants generated.
The judges were completely unaware of any participant’s circumstances and assign scores separately in this study. Then a bivariate correlation analysis was executed for the scores of the creative performance assigned by the two judges per idea. Table 2 presents a strong significant correlation between the two, r = .747, p < .01, thus a new variable which consists of the mean of the scores assigned by the two judges is computed as the dependent variable creative performance for further analysis.
Table 2. Results of bivariate correlation analysis of creativity score of the two judges
Variables M SD 1
Scores1 1.41 1.15
Scores2 1.23 1.24 .747**
Note: **p<.01
4.1.3. Reliability analysis
The reliability analysis was conducted to test the internal consistency of the constructs with subjective scales. At first, the Cronbach alpha of the mediator and moderators were relatively low due to the difficulty in controlling the quality of participants’ answers through conducting such an online experiment under the influence of the epidemic compared to a field experiment in a lab. For example, some participants just scanned through the questionnaire very quickly on the phone without serious thinking, which resulted in lower quality of the answers. In addition, the particular items possibly had confusing content, for example some are reversed statements while others are normal ones, which could also have a negative effect on the quality of the data. Table 3 shows the Cronbach alpha of the 4 constructs. Deleting two items of conscientiousness and one item of openness to experience resulted in change of Cronbach’s alpha from .184 and .670 to .781 and .845 respectively. In addition, deleting two items from the seven-item intrinsic motivation resulted in change of Cronbach’s alpha from .637 to .867. Thus, in order to enhance the reliability and
internal consistency of the constructs, items 5, 9, 10, 11, and 14 were deleted.
Subsequently, those items left were used for computing new variables for further analysis.
4.1.4. Correlation analysis
After checking the normality of the variables, the skewness of creative performance per participant was .052, and that per idea was .312, both of which were acceptable. Figure 2 and 3 shows the plot of the distribution of the dependent variable.
Furthermore, a bivariate correlation analysis was executed for all the variables in the model. Table 3 reports the correlation matrix in which the means, standard deviations, correlations, and Cronbach’s alpha of the variables are presented.
As table 3 displays, there’re significant negative correlations between age and creative performance (r = -.22, p < .01), as well as age and openness to experience (r = -.18, p < .05). This suggests that as people grow older, they become less open to experience and less creative. Table 3 also presents significant positive correlations between creative performance and conscientiousness (r = .35, p < .01), as well as creative performance and openness to experience (r = .37, p < .01), which is consistent with hypothesis 4b but opposite of hypothesis 4a. Furthermore, it can also be found that there’s a significant negative correlation between conscientiousness and intrinsic
motivation (r = -.28, p < .01). This suggests that individuals scoring higher in conscientiousness have less creative intrinsic motivation, which might be due to that they care more about accomplishing the given task rather than generating new ideas or exploring new methods that could increase the risk of interrupting the process (George
& Zhou, 2001; Goldberg, 1990).
Figure 2. Histogram of participant scores
Figure 3. Histogram of individual idea scores
Note: Each participant generated titles for two short stories, so N = 164*2 = 328.
Table 3. Means, standard deviations and correlations
Variables M SD 1 2 3 4 5 6 7 8 9
1. Gender - - -
2. Age 3.46 1.37 - -
3. Education - - - - -
4. Work experience 1.67 1.17 - .57** - -
5. Expected specification group - - - -
6. Intrinsic motivation 4.18 .71 - .02 - -.06 - (.87)
7. Conscientiousness 2.35 1.17 - -.00 - .13 - -.28** (.78)
8. Openness to experience 2.67 1.22 - -.18* - -.07 - -.09 .62** (.85)
9. Creative performance 2.65 2.04 - -.22** - -.01 - -.04 .35** .37** (.89)
Note: Nominal/ categorical/ ordinal variables such as gender, education and expected specification group are not included in this correlation matrix. Gender: 1=male, 2=female, 3=Non-binary/ third gender, 4=prefer not to say. Age: 1 = Below 18, 2 = 18 - 25, 3 = 26 - 30, 4 = 31 - 35, 5
= 36 - 40, 6 = 41 - 45, 7 = 46 - 50, 8 = 51 - 55, 9 = 56 - 60, 10 = 61 - 65, 11 = 66 and older. Education: 1 = Less than a high school diploma, 2
= High school degree or equivalent, 3 = Associate degree, 4 = Bachelor’s degree, 5 = Master’s degree, 6 = Doctoral degree. Work experience:
1 = 0 - 5, 2 = 6 - 10, 3 = 11 - 15, 4 = 16 - 20, 5 = 21 - 25, 6 = 26 - 30, 7 = 31 - 35, 8 = 36 - 40, 9 = 40 - 45, 10 = 45 and more. Output constraints of expected specifications group: 1 = no expected specifications at all, 2 = strict expected specifications. Cronbach’s alpha in brackets.
*. P < 0.05. **. P < 0.01.
4.2. Hypotheses testing
4.2.1. Main effect of expected specifications for creative performance
Hypothesis 1 suggested that the output constraints of expected specifications have a negative effect on creative performance. In order to examine the main effect between the expected specifications and creative performance, a one-way ANOVA test was executed, which can be used to examine whether there are significant differences between the means of the two independent groups with different levels of expected specifications. First, the homogeneity of variance was checked. Second, as table 4 displays, there was statistically significant effect of output constraints of expected specifications on creative performance, F (1,162) = 7.76, p < .05. Furthermore, the mean of the group with strict expected specifications was 2.17, which was lower than that 3.04 of the group with no expected specification. Thus, Hypothesis 1 was supported.
Table 4. Results of the one-way ANOVA test for creative performance
Sum of Squares DF Mean Square F Sig.
Group 30.87 1 30.87 7.76 .006
Error 644.48 162 3.98
Total 675.35 163
4.2.2. Mediating effect of intrinsic motivation on the relationship between output constraints of expected specifications and creative performance
Hypothesis 2 suggested that the intrinsic motivation mediates the relationship between the output constraints of expected specifications and creative performance. To
Group Mean SD N
No expected specifications 3.04 1.94 89 Strict expected specifications 2.17 2.05 75
Total 2.65 2.04 164
test this hypothesis, model 4 of Process was used (Hayes, 2012; Hayes & Preacher, 2014). As is displayed in Table 5, first of all, the direct effect of output constraints of expected specifications on creative performance was negative and significant, c’ = -.871, p < .05, which further supports Hypothesis 1. However, there were no significant relations between output constraints of expected specifications and the mediator creative intrinsic motivation, a1 = .002, p > .05. In addition, the results also showed that there were no significant relations between creative intrinsic motivation and creative performance, b1 = -.108, p > .05. Furthermore, the indirect effect, a1b1 = .0002 (see Table 6), also showed that it was not statistically different from zero. Therefore, no support was found for Hypothesis 2.
4.2.3. Moderating effects of conscientiousness and openness to experience on the relationship between expected specifications and creative intrinsic motivation In Hypotheses 3a and 3b, we proposed that the relationship between expected specifications and intrinsic motivation is moderated by conscientiousness and openness to experience. In addition, we also proposed that the relationship is weaker among individuals high in conscientiousness and stronger among individuals high in openness to experience. To test these hypotheses, model 8 of Process used, which is for conditional indirect effect: moderated mediation testing (Hayes, 2012). Before running moderation analysis in Process, the numerical variables conscientiousness and openness to experience were standardized.
Table 5. Mediation - Regression coefficients, Standard errors, proportion explained variance, and F statistic Consequent
Intrinsic motivation (M) Creative performance (Y)
Antecedent Coeff. SE p Coeff. SE p
Expected specification Group (X) a1 .002 .112 .989 c' -.871 .313 .006
Intrinsic motivation (M) - - - b1 -.108 .221 .625
Constant i1 4.180 .076 .000 i2 2.594 .946 .007
R2 = .000 R2 = .047
F (1,162) = .000 F (2,161) = 3.982
p = .989 p = .625
N = 164, Bootstrap resamples: 5000
Table 6. Mediation – Direct, total, and indirect effects
Effect SE p LLCI ULCI
Direct effect c1' -.8711 .3134 .0061 -1.4900 -.2523
Total effect c1' -.8709 .3126 .0060 -1.4883 -.2536
Boot SE Boot LLCI Boot
ULCI
Indirect effect a1b1 .0002 .0298 -.0657 .0648
N = 164, Bootstrap resamples: 5000
As is shown in Table 7 and 8, the moderated mediation effects both by conscientiousness and openness to experience between expected specifications and intrinsic motivation were not taking place, p > .05. Therefore, no support was found for Hypothesis 3a and 3b.
Furthermore, considering that the moderation effect between the independent variable and the mediator did not take place, which was to be expected since the mediation effect itself was not supported, the moderation effect between the independent variable and the dependent variable was further explored, using model 1of Process (Hayes, 2012). However, as Table 9 and 10 shows, the moderation effects of both conscientiousness and openness to experience between expected specifications and creative performance did not take place, p > .05.
Table 7. Moderated mediation - Conscientiousness
Variables Coeff. SE T p LLCI ULCI
Constant 4.175 .073 57.226 .000 4.031 4.319
Expected specification Group .012 .108 .110 .913 -.201 .225 Z_ Conscientiousness -.209 .075 -2.786 .006 -.358 -.061
Int_1 .022 .108 .205 .838 -.191 .235
Table 8. Moderated mediation – Openness to experience
Variables Coeff. SE T p LLCI ULCI
Constant 4.178 .076 55.230 .000 4.029 4.328
Expected specification Group .008 .112 .069 .945 -.213 .229
Z_ Openness -.041 .076 -.539 .591 -.190 .108
Int_1 -.055 .112 -.485 .628 -.276 .167
Table 9. Moderation - Conscientiousness
Variables Coeff. SE T p LLCI ULCI
Constant 3.970 .452 8.785 .000 3.078 4.863
Expected specification Group -.909 .293 -3.096 .002 -1.488 -.329
Z_ Conscientiousness .660 .460 1.436 .153 -.248 1.568
Int_1 .039 .293 .134 .894 -.540 .619
Table 10. Moderation – Openness to experience
Variables Coeff. SE T p LLCI ULCI
Constant 4.015 .447 8.985 .000 3.132 4.897
Expected specification Group -.938 .290 -3.233 .002 -1.511 -.365
Z_ Openness .856 .447 1.915 .057 -.027 1.739
Int_1 -.057 .291 -.197 .844 -.633 .518
4.2.4. Direct effects of conscientiousness and openness to experience for creative performance
In Hypotheses 4a and 4b, we proposed that conscientiousness has a negative effect on creative performance, and openness to experience has a positive effect on creative performance. In order to test them, a linear regression analysis was executed. As is shown in Table 11, openness has a significant positive effect on creative performance, β2 = .251, p < .05, which supported Hypothesis 4b. However, the results show that conscientiousness has a significant positive effect on creative performance as well, β1
= .192, p <.05, which did not support Hypothesis 4a that predicted a negative effect.
Thus, Hypotheses 4a was not supported, while Hypothesis 4b was supported.
Table 11. Results of the linear regression
Model Coeff. Std. Error Beta t Sig.
Constant .744 .375 1.984 .049
Conscientiousness .334 .160 .192 2.081 .039
Openness to experience .419 .154 .251 2.721 .007 Note. Dependent variable: creative performance
4.3. Summary of hypotheses
H1: The output constraints of expected specifications have a negative effect on creative performance.
The hypothesis was supported.
H2: The intrinsic motivation mediates the relationship between the output constraints of expected specifications and creative performance.
The hypothesis was not supported.
The results indicated no significant effect.
H3a: The relationship between expected specification and intrinsic motivation is moderated by conscientiousness. The relationship is weaker among individuals high in conscientiousness.
The hypothesis was not supported.
The results indicated no significant effect.
H3b: The relationship between expected specification and intrinsic motivation is moderated by openness to experience. The relationship is stronger among individuals high in openness to experience.
The hypothesis was not supported.
The results indicated no significant effect.
H4a: Consciousness has a negative effect on creative performance.
The hypothesis was not supported.
The results indicated a positive effect.
H4b: Openness to experience has a positive effect on creative performance.
The hypothesis was supported.
5. Discussion
5.1. Discussion of the results
This study aims to examine the role of output constraints of expected specifications in creativity, the mediating mechanism of intrinsic motivation between them, the effects of conscientiousness and openness to experience on the relationship between expected specifications and intrinsic motivation, and the direct effects of the two personality traits on creativity. The present analysis was conducted on the data collected from the online experiment.
Although scarce empirical evidence exactly on the role of expected specifications in creativity, there were prior research that explored some external constraints similar with expected specifications. Hypothesis 1, predicting that expected specifications have negative effects on creativity, was in line with the research of Amabile (1988), Amabile (1996), and Rosso (2014) that indicated negative effects of external constraints on creative performance. The results of this study proved that output constraints of expected specifications did have negative effects on creativity. This could be explained by the lack of participants’ sense of control over the task and output, and by limiting alternatives for the creative solution (Amabile, 1988; Rosso, 2014).
Hypothesis 2 expected a mediating effect of intrinsic motivation on the relationship between expected specifications and creativity. This hypothesis was supported by research of Amabile et al. (2002) and Rosso (2014). For example, Rosso (2014) conducted a field research and found that both process and product constraints tended to inhibit creativity when reducing intrinsic motivation or experimentation.
However, the results of this study indicated no significant mediation effect of intrinsic
motivation between expected specifications and creativity, which was not consistent with the prior theories. A possible reason for not finding a significant effect may be due to the limitations of this study, for example, the quality of the data, which will be discussed later. Another possible explanation could be that the tasks under both situations with no expected specifications and strict expected specifications were too simple to significantly affect participants’ feelings and motivation towards the activity.
The mean scores for the intrinsic motivation of the two groups, both really close to 4.18, can support the possible explanation mentioned above. Furthermore, there might be other motivational mediating mechanisms through which expected specifications affect creative performance. Based on Acar et al. (2019)’s framework summarized from prior research, it could be compliance, task enjoyment, and challenge appraisal, etc.
Hypotheses 3a and 3b expected moderation effects of conscientiousness and openness to experience on the relationship between expected specifications and intrinsic motivation. For hypothesis 3a, it could be supported by Barrick, Mount, &
Strauss (1993) and Barrick et al. (2002) who indicated that individuals scoring high in conscientiousness are more likely to be committed to accomplish the set goals, exercise self-control and to seek to fulfill obligations, thus less likely to be influenced by output constraints. For hypothesis 3b, it could be supported by Burke & Witt (2002)’s finding that individuals scoring high in openness to experience prefer working with independence and autonomy, and thus more likely to be negative facing output constraints. However, the results of this study indicated no significant effect. In addition, we tested the moderation effects of these two personality traits on the relationship
between expected specifications and creativity, and also found no significant effects.
There could be two possible reasons to explain these results. First, the quality of the data might be not high enough. Second, the scales in measuring the personality traits included so many reverse statements that might make the participants confused.
Hypotheses 4a and 4b expected a significant negative effect of conscientiousness and a significant positive effect of openness to experience on creativity, which could be supported by Sung & Choi (2009) and George & Zhou (2001). The results of this study proved hypothesis 4b that openness to experience did have a significant positive effect on creative performance, while also indicating a significant positive effect of conscientiousness on creativity, which was opposite of hypothesis 4a. In addition to the data quality issue, it could be explained that the opposite theory may also make sense.
To be specific, considering that individuals’ conscientiousness is measured by how organized, hardworking and dependable they are, those scoring high in conscientiousness are often more purposeful and strive for achievement (Salgado, 1997;
George & Zhou, 2001; Goldberg, 1990). Therefore, if the participants with high conscientiousness regarded generating the most creative titles for the two short stories as their given task, they would also strive for it. This viable theory could explain why there was a significant positive effect of conscientiousness on creative performance.
5.2. Theoretical and Practical Implications 5.2.1. Theoretical implications
This study contributed to the existing literature in several ways. Firstly, although there’re prior research that explored some other external constraints, the relationship
between output constraints of expected specifications and creativity lacked empirical evidence. This study proved the negative effect of expected specifications on creativity.
Secondly, Acar et al. (2019)’s cross-disciplinary integrative framework indicated that there was no empirical evidence on how output constraints affect creativity through motivational route. This study is the first one that empirically examines how output constraints affect creativity through a motivational route. Although the results showed no significant effect, it demonstrated that the debate should focus on the existence of that mediation effect, and whether there’re other motivational mediating mechanisms that truly make a difference.
Thirdly, this study also filled the gap where there’s little prior research that considered the moderators between output constraints of creativity and mediating mechanisms. The results showed no significant moderating effects of conscientiousness and openness to experience on the relationship between expected specifications and intrinsic motivation. Future research could consider other moderators.
Lastly, this study proved the positive effect of openness to experience on creativity, which was consistent with prior research (e.g. Sung & Choi, 2009). However, it also indicated an unexpected positive effect of conscientiousness on creative performance, which demonstrated that the debate should focus on the direction of that effect.
5.2.2. Practical implications
As outlined in the introduction, it’s important for managers to know how not to kill creativity at an individual level in practice, since creativity is crucial for an organization’s survival and development (Anderson, Potočnik & Zhou, 2014).
First, this study examined the negative effect of output constraints of expected specifications on creative performance. This indicates that in order to increase stuffs’
creativity, managers should reduce the externally imposed expected specification that is not a must, when dealing with a creativity task. Considering that clients might always make a lot of requirements on the output due to the information asymmetry, which is really common and not easily controlled by the firm that provides creative products or service, managers should put effort into direct or indirect communicating with those customers, for example, holding some small workshops and trainings about creativity for the clients, and reduce unnecessary external output constraints imposed on employees, especially in creative industries or when the level of creativity is a priority.
Second, the findings of this study didn’t find proof for significant mediation effect of intrinsic motivation or moderation effects of conscientiousness and openness to experience. Assuming that these relationships indeed do not exist, when faced with the negative effect of expected specifications on creativity, managers cannot compensate for this impact by adjusting the intrinsic motivation of employees, even if the employees are with high conscientiousness and openness to experience. It’s more important to take initiative to control that negative impact by reducing unnecessary external output constraints as mentioned above.
Third, this study also examined the positive effects of both conscientiousness and openness to experience on creative performance. This indicates that individuals scoring high in conscientiousness and/or openness to experience are more likely to have a higher level of creativity. Based on it, a recommendation for managers of a firm or an
organization, especially in creative industries, is to conduct a personality test among people who come to apply for a job, and to take the test results into consideration when recruiting an employee whose work requires high level of creativity.
5.3. Study limitations and future suggestions
There are some limitations regarding this study which provide directions for future research. First, as mentioned in the discussion of results, the quality of the data collected from the online experiment could be relatively low. Although it might be the most executable way under the influence of the Covid-19 epidemic, participants can muddle through the experiment, which leads to low quality of data collected. Thus, apparently, conducting an online experiment is less reliable than a field experiment where the environment in which participants take part in the experiment can be controlled. What’s more, due to that all the participants were selected via the online platform Amazon Mechanical Turk, there might be risk of unmotivated participants taking part in the experiment not seriously but just to earn money. Although the platform helped in diversifying samples, it could have negative effect on the quality of the data collected. It would be interesting and meaningful for future research to conduct a field experiment to do similar research and see if the results would be different.
Second, although initially measured by two four-item scales, there’s a small limitation that the two personality traits were measured by two and three items respectively after adjusting according to reliability analysis. Although it was considered to reduce the transient measurement errors and keep the focus of the participants at first, and the Cronbach’s alpha was indeed high after deleting two items of conscientiousness
![[Review of: J. Ingleson (2014) Workers, Unions and Politics: Indonesia in the 1920s and 1930s] - 22134379_170_04_s010_text](data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACH5BAEAAAAALAAAAAABAAEAAAICRAEAOw==)