Searching outside the box in creative problem solving: The role of creative thinking skills and domain knowledge

Journal of Business Research 81 (2017) 1–10

Contents lists available at ScienceDirect

Journal of Business Research journal homepage: www.elsevier.com/locate/jbusres

Searching outside the box in creative problem solving: The role of creative thinking skills and domain knowledge

MARK

Tamara Montag-Smita,⁎, Carl P. Maertz Jr.b a b

Department of Management, Ball State University, Muncie, IN, United States Department of Management, Saint Louis University, Saint Louis, MO, United States

A R T I C L E I N F O

A B S T R A C T

Keywords: Idea generation Creative outcome effectiveness Information type Creative thinking skill Domain knowledge

This study provides evidence for how factual (directly relevant for developing creative solutions) vs. range (indirectly relevant) information can be used to provoke idea generation and effective creative outcomes. Data were obtained from 127 staff, faculty, and students at a private Midwest university. The results show that the effect of type of information was moderated by the participant's creative thinking skill and domain knowledge. For individuals high in creative thinking skill, range information improved idea generation originality, which in turn enhanced creative outcome novelty but reduced outcome usefulness. Factual information, under various conditions, both helped and hindered creative outcome usefulness. Overall, presenting information during idea generation can improve creative outcome effectiveness. Managers should be careful to present the appropriate information at the appropriate time, however.

1. Introduction Few researchers or business practitioners can deny the pervasive importance of creativity in the workplace for the economic benefit of organizations, countries, and markets (e.g., Marrocu & Paci, 2012). Thus, it is not surprising that creativity has been a key topic within the R & D design/engineering (Burbiel, 2009), marketing (Smith & Yang, 2004), entrepreneurship (Ardichvili, Cardozo, & Ray, 2003), and management literatures. In short, scholars across business disciplines seek to better understand workplace creativity and practical ways by which it can be facilitated, and a small subset of researchers seek to understand ways in which creativity is facilitated outside of employee formal job requirements (e.g., Axtell, Holman, & Wall, 2006; Kesting & Ulhøi, 2010; Unsworth, 2001). These researchers point to a key question for business: how can managers support employees in developing creative ideas when it is not a formal requirement of their job? Existing research on idea generation fits this question well, because it often utilizes tasks in which limited domain knowledge is necessary to produce ideas (Jones & Kelly, 2009; e.g., Nijstad, Stroebe, & Lodewijkx, 2003). In expected creativity, where it is part of the job (e.g., artists, scientists, etc.), employees may need significant domain knowledge to make novel contributions. In unexpected creativity, however, employees typically have sufficient knowledge and skills to generate creative ideas without extensive training. One common tactic for enhancing idea generation is brainstorming

⁎

(Osborn, 1953). An assumption of brainstorming is that individuals will come up with more creative ideas when they build off previously generated ideas (i.e., engage in external information processing). Research in this area, however, has produced mixed results when examining the effect that ideas from others have on idea generation. Some research findings support the assumption that paying attention to others' ideas during brainstorming can improve idea generation (Brown & Paulus, 2002; Nijstad & Stroebe, 2006). On the other hand, research also shows that the presence of other's ideas can reduce idea generation productivity (Diehl & Stroebe, 1987). While production blocking, described as the delay in idea generation based on group members taking turns to express their ideas, is one possible explanation for these mixed results (e.g., Diehl & Stroebe, 1987; Nijstad et al., 2003), we test two additional factors that may help explain these inconsistencies in the existing research: information type and individual characteristics. First, we suspect that the type of information available during idea generation will impact the creativity of the ideas generated, and we focus on two types: factual and range. With the vast array of information that is available to a person at any given time (e.g., visual, verbal, etc.) and a person's limited capacity to absorb such amounts of information (Malhotra, 1982), individuals choose the type of information on which to focus during idea generation tasks. Therefore, one key distinction is to classify information as either directly relevant to problem solutions (factual) or indirectly relevant (range) for the specific

Corresponding author at: 2000 W. University Ave., Ball State University, Department of Management, WB 205, Muncie, IN 47306, United States. E-mail address: [email protected] (T. Montag-Smit).

http://dx.doi.org/10.1016/j.jbusres.2017.07.021 Received 4 November 2016; Received in revised form 28 July 2017; Accepted 29 July 2017 0148-2963/ © 2017 Elsevier Inc. All rights reserved.

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

one cluster can easily be connected, meaning people will generate many ideas quickly when drawing from one semantic cluster (i.e., fluency; Nijstad et al., 2003). Building on this idea, we expect that external information will increase idea generation fluency to the extent that it can be easily integrated with or prompts recall from easily accessible idea clusters. SIAM also proposes that idea generation results from cognitive processes such as conceptual combination or the use of analogies (e.g., Kohn, Paulus, & Korde, 2011). As information (both external and internal) is activated in working memory, people form new associations between those pieces of information. As each new association comes into working memory, it potentially brings along additional search/ focus cues, which may then activate new semantic clusters. This process is repeated until new associations are exhausted using that search cue. At this point a person may activate a new search cue to repeat this process with different information. When people shift to a new semantic cluster they are more likely to develop ideas from multiple semantic categories, which represents greater flexibility (Nijstad et al., 2003). Moreover, the ideas generated later in this process tend to be more original (Kohn, Paulus, & Choi, 2011). Building on these arguments, we expect that external information will increase idea generation flexibility and originality to the extent that it can be integrated within one's semantic network and expands that semantic network. SIAM does not make direct predictions about idea generation elaboration. However, elaboration requires focus in working memory on one semantic cluster in detail, and searching for how the idea can be explained and clarified through more detail. Given that this is similar to within category fluency (e.g., Nijstad et al., 2002), it would seem that any information could aid in detail elaboration, but this would be especially true of information within one semantic category.

problem (Mumford, Baughman, Supinski, & Maher, 1996). This mirrors the idea of focusing on information that is “in the box” or “outside the box.” Based on this distinction, we suspect each information type will have a unique effect on idea generation. Second, we know that there are individual differences in cognitive processing during idea generation (e.g., Friedman & Forster, 2001). Two cognitive processes seem to be drawn upon during idea generation: procedural memory/knowledge and declarative memory/knowledge (Jin, Kwon, Jeong, Kwon, & Shin, 2006). In the current context, procedural knowledge is most reflected in creative thinking skill in idea generation, and declarative knowledge is one's level of task domain knowledge. Both creative skill and domain knowledge have both been theorized and found to improve idea generation (Amabile, 1983), but research has failed to examine how these individual characteristics influence idea generation. Thus, we examine the moderators under which information condition each characteristic is most helpful for creativity. Third, the creativity research seems to be divided into at least two areas: research examining idea generation as the dependent variable (e.g., Kohn, Paulus, & Korde, 2011; Nijstad et al., 2003), and research examining creative outcomes and products as the dependent variables (e.g., Oldham & Cummings, 1996). These constructs are theoretically distinct but causally-related ideas (Montag, Maertz, & Baer, 2012). According to Amabile's Componential Theory of Creativity (1983), people take nascent ideas (such as those generated during brainstorming) and evaluate, revise, and refine them to develop a product or solution. The nascent ideas are different from creative outcomes, because the latter “final products” are refined ideas that are intended for implementation or end use. Thus, idea generation behavior affects creative outcome effectiveness (COE; Montag et al., 2012), but these are distinct. In the current paper, we separate and test the relationship between these two constructs. In the following sections, we discuss the process of individual idea generation and the role that external information plays in this process. Then we describe our expectations for effects of factual and range information on idea generation. Four unique facets of idea generation (i.e., fluency, flexibility, originality, and elaboration) have long been theorized and measured in the literature (Guilford, 1957; Torrance, 1966). Fluency is the raw number of distinct ideas generated, flexibility is the extent to which these ideas represent different semantic categories/content areas, originality is the extent to which ideas are different from those suggested by other generators, and elaboration is the extent of detail and clarification provided in describing each idea. These different aspects of idea generation have been correlated with activity in different parts of the brain (Chávez-Eakle, Graff-Guerrero, García-Reyna, Vaugier, & Cruz-Fuentes, 2007), highlighting the importance of examining unique relationships with each facet. Thus, our hypotheses make specific predictions for the different facets of idea generation, tested simultaneously. From there we develop the rationale for the moderating effects of both creative thinking skill and domain knowledge. Finally, we explore linkages between the idea generation facets and COE.

3. Hypothesis development We suggest that different types of information are interpreted by individuals differently. Information can be classified into several types. In fact, creativity researchers have examined multiple types of information that have been shown to influence the creative process. Mumford and colleagues (Friedrich & Mumford, 2009; Mumford et al., 1996), for example, found that people who produced high quality and original solutions (high COE) spent less time on irrelevant information and more time on inconsistent and factual information. In this paper, we focus on factual and range information, which best approximates the concept of focusing inside vs. outside “the box”, respectively. Factual information is defined as information that is directly relevant for coming up with solutions to a creative problem (Mumford et al., 1996). This information is typically explored early in the idea generation process, and can also be used to verify or evaluate ideas. Exposure to factual information, just like common ideas generated in a brainstorming session, should spur many associations within a limited semantic network (Nijstad et al., 2002), which should directly increase idea generation fluency (Kohn, Paulus, & Choi, 2011; Nijstad & Stroebe, 2006). Elaboration is defined as extending and clarifying an idea through adding detail. Factual information, usually within a given sematic category but differing by sub-category, could allow more details to be added to a particular idea, improving elaboration more than having no such information. Factual information should aid individuals in generating responses from a similar category, which reduces the total number of categories from which they produce ideas (Brown & Paulus, 2002). Thus, factual information reduces idea generation flexibility. Likewise, factual information may cause individuals to get stuck thinking within a limited range of solutions (Runco, 2004) due to functional fixedness, described as a mental block against seeing or using something in a new way to solve a problem (Duncker, 1945), thereby reducing idea generation originality.

2. Theoretical background Idea generation is the process in which individuals use divergent thinking to develop ideas intended to solve non-algorithmic problems. We draw from the Search for Ideas in Associative Memory (SIAM; Nijstad & Stroebe, 2006; Nijstad, Stroebe, & Lodewijkx, 2002; Nijstad et al., 2003) model of idea generation to understand the information processing that occurs during idea generation. According to SIAM, information is stored in memory in semantic clusters or images (Nijstad et al., 2003). Information within a cluster is strongly related, and clusters of information are weakly related to other clusters. In the idea generation process people first search their long-term memory for solutions using the problem definition as a search cue (Nijstad & Stroebe, 2006). During this search, they activate idea clusters, and ideas from

Hypothesis 1a. Factual information will increase idea generation 2

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

fluency and elaboration compared to no information.

memory and are better able to give attention to and utilize external information (Chávez-Eakle et al., 2007). In particular, based on the increased working-memory demands of integrating range information and generating flexible ideas (as outlined in SIAM; e.g., Nijstad et al., 2003), we expect creative thinking skill to enhance the use of range information to develop flexible and original ideas.

Hypothesis 1b. Factual information will decrease idea generation flexibility and originality compared to no information. Range information is not directly relevant for coming up with a creative solution (Mumford et al., 1996). In other words, range information may require a more distant or abstract conceptual linkage to the task/problem at hand. It may seem irrelevant at first, but range information is typically explored after the obvious options are exhausted (Nijstad & Stroebe, 2006), so it may aid the idea generation process by stimulating novel associations. Range information may cause individuals to think of a wider range of solutions (i.e., flexibility), because range information cues thoughts allowing individuals to overcome functional fixedness (Nijstad et al., 2002). In fact, presenting seemingly unrelated ideas has been found to trigger more original category labels (Kohn, Paulus, & Korde, 2011; Mobley, Doares, & Mumford, 1992). Thus, range information should aid individuals in generating original responses to a creative problem (Kohn, Paulus, & Choi, 2011). Indirectly, range information may increase idea elaboration within a category because increased flexible and original thinking about a given idea should increase one's ability to elaborate that idea's details, creating richer imagery of the idea. Supporting this idea, Sosik, Kahai, and Avolio (1998) found higher idea elaboration for teams with transformational leaders, who promote consideration of different viewpoints (i.e., range information). Nevertheless, range information should also take longer to process due to its less direct relevance to the problem solution. Thus, this type of information may hinder a person's ability to produce a large number of distinct ideas (Kohn, Paulus, & Choi, 2011).

Hypothesis 4. Level of creative thinking skill will moderate the relationship between information type and idea generation, such that positive effects are enhanced and negative effects are attenuated when people are high in creative thinking skill. Domain-relevant knowledge includes facts, principles, technical skills, and opinions about issues and topics within a task domain. Creative performance involves connecting diverse pieces of information within one's working memory (Nijstad & Stroebe, 2006), and people who have higher levels of knowledge relevant for the task will have more information available with which to associate in their working memory. Additionally, domain knowledge can improve idea generation performance by increasing one's comprehension of new domain-relevant information (Hambrick & Engle, 2002). Thus, it would be expected that domain knowledge influences the creative thinking process through information processing efficiency. People use their domain knowledge to evaluate and synthesize new information or ideas (Amabile, 1996), so people with greater domain knowledge may be better able to integrate external information into their current framework of ideas than those with low domain knowledge. Moreover, they are likely better able to discriminate between information that is helpful for generating new ideas, and information that is not. However, individuals high in domain knowledge may find focusing on external factual information redundant such that processing the information actually distracts from or otherwise hinders their ability to generate original ideas, compared to those who are low in domain knowledge or those who do not receive factual information (e.g., Luo & Toubia, 2015).

Hypothesis 2a. Range information will increase idea generation flexibility, originality and elaboration compared to no information. Hypothesis 2b. Range information will decrease idea generation fluency compared to no information.

Hypothesis 5. Level of domain knowledge will moderate the relationship between information type and idea generation, such that positive effects are enhanced and negative effects are attenuated when people are high in domain knowledge.

We further reason that the combination of presenting both types of information during idea generation could allow individuals to simultaneously draw on the strengths of each type of information just mentioned, especially because each type of information is consistent and does not contradict one another (cf. Friedrich & Mumford, 2009; Mumford et al., 1996).

Finally, we expect idea generation to enhance COE. However, a limited number of studies have examined the relationship between these two constructs, and the current theory is vague regarding the impact of each facet (i.e., fluency, flexibility, originality, and elaboration) of idea generation on COE. Thus, while we expect an overall positive relationship between idea generation and COE (Amabile, Barsade, Mueller, & Staw, 2005), we explore the specific relationships at a facet level. Moreover, recent evidence shows that COE is best captured by examining its dimensions (i.e., novelty and usefulness) independently (Sullivan & Ford, 2010). Thus, to understand relationships with COE more precisely, we examine the formative dimensions of novelty and usefulness separately as well.

Hypothesis 3. Factual and range information in combination will increase idea generation performance (across all four dimensions) compared to all other conditions. The above hypothesized relationships may depend on one's task relevant skills and abilities. Creative thinking skill is an ability to think or process information in a manner that leads to new ideas (Hong & Milgram, 2010). Substantial research evidence shows that creative thinking skill predicts future creative accomplishments (Cramond, Matthews-Morgan, Bandalos, & Zuo, 2005), and an ability to generate creative ideas within specific tasks and domains (Hong & Milgram, 2010). This ability can also enhance the positive effects of contextual factors that facilitate creative outcomes, as well as attenuate the negative effects of such factors detrimental to creative outcomes (e.g., unsupportive organizational climate; Choi, Anderson, & Veillette, 2009). Given that attending to external information should improve creative outcomes, we generally expect that creative thinking skill will enhance this positive effect. Specifically, creative thinking skill should influence how well a person can integrate and build from information presented during creative idea generation. Evidence suggests that individuals higher in creative thinking skill are better able to identify and distinguish relevant information from irrelevant information (Davidson & Sternberg, 1984), but they do not necessarily filter out the irrelevant stimuli (Carson, Peterson, & Higgins, 2003). Moreover, they hold greater amounts of information in working

4. Method 4.1. Participants A total of 127 staff, faculty, and students from a private Midwestern university in the US completed the two-part study (42% male; 40% employed at the University; 48% attended the University full-time and 22% part-time). As an incentive, five participants were randomly selected to receive $50, and two were awarded $50 for having the highest rated creative proposals. Some students were also given extra credit for participation. Participants were randomly assigned to one of four conditions: control (n = 28), factual information (n = 28), range information (n = 35), and factual + range information (n = 36). 3

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

several participants were rated low and responses mentioned by only one individual rated high on a scale of one to seven (Carson et al., 2003). The inter-rater reliability was 0.84, and thus the originality score was calculated as an average across raters and responses for each case. An average count of the number of words used to describe each idea was calculated for each participant in order to determine elaboration.

4.2. Design and procedure The design was a two-by-two (factual information x range information), between-subjects experiment. Participation occurred in two online sessions using Qualtrics survey software. In session one, participants filled out demographic information and completed the domain knowledge measure and a standard creative thinking task. The results of the creative thinking task were used to determine the participant's creative thinking skill. In session two, which occurred approximately one day after session one (i.e., the link for session two was emailed to participants 24 h after session one), participants were randomly assigned to one of the four conditions while completing the creative task.

4.3.2. Creative outcome effectiveness ratings The Consensual Assessment Technique was used to measure COE based on extensive evidence supporting its validity for measuring creative outcomes (see Amabile, 1983, 1996). As Amabile states, “creativity is something that people can recognize and often agree on, even when they are not given a guiding definition (Amabile, 1983, p. 360).” Moreover, both experts and trained non-experts can agree on what is creative (Amabile, 1996). Thus, four trained graduate students (blind to experimental condition) rated the outcomes of the tasks (i.e., the proposals).2 These graduate students evaluated the proposals on a number of dimensions, including novelty and usefulness. Novelty was defined as relative newness of the idea compared to the existing practices of Student Life, and usefulness was defined by ability to usefully implement the idea. Inter-rater reliability for novelty was 0.84 and for usefulness was 0.71. These numbers suggest consistency between raters; thus, the hypotheses were tested using averaged ratings.

4.2.1. Creative task Participants were asked to write a proposal to improve student life/ involvement at their University. In order to write the proposal, participants were first asked to generate ideas. For this portion of the task, participants were told to censor themselves as little as possible, allowed to write incomplete thoughts, and asked to write down anything that came to mind and not delete responses once they were written. After generating a list of at least three ideas, they were instructed to choose one idea and expand upon it for their proposal. 4.2.2. Information manipulation Subjects were randomly assigned to one of four conditions. In the control condition, subjects were provided no external information to help them complete the task. In the second condition, they received factual (directly relevant) information for the task, which read, “Here is some information about student life at [University] that may help you with this task.” This was followed by a bullet point list of facts found on the University's website. In the third condition, subjects received range (indirectly relevant) information for the task, which read, “Here is some information about [University] and [city] that may help you with this task,” followed by a bullet point list of facts found on the University's website. In the fourth condition, subjects received both factual and range information. Half of the bullet points were randomly selected from conditions two and three so that the fourth condition provided individuals with equal total amount of information as conditions two and three. Because the timing of information presentation can affect creative performance (Friedrich & Mumford, 2009), all subjects were provided the external information at the beginning of idea generation.

4.3.3. Creative thinking skill To measure creative thinking skill subjects completed a generic creative thinking task at time one. Building on work by Guilford and colleagues (Guilford, 1975; Wilson, Guilford, Christensen, & Lewis, 1954), participants were asked to come up unique responses to the prompt “things that have wheels”. Responses were measured according to the four dimensions of idea generation: fluency, flexibility, originality, and elaboration. Fluency, flexibility and elaboration were measured in the same fashion as “idea generation performance”. For originality we assigned a score to each idea generated based on the infrequency of the response. For example, “car” or “auto” was the most frequent response, so this response was scored as a “1”. A highly infrequent response, on the other hand, was scored as a “7”. Once each response was scored, we calculated the average score for each participant. To calculate the creative thinking skill score for each participant, we then standardized and averaged the four facets.

4.3. Measures

4.3.4. Domain knowledge Relevant domain knowledge was measured with participant's selfreport responses to four items. The items included: I know a lot about student involvement, I would need to contact the office of Student Life to know what opportunities for student involvement are available (reverse coded), I could easily come up with new ways to improve student involvement, and I could explain opportunities for student involvement to a new student. Cronbach's Alpha reliability among the four items was 0.81, which is above acceptable standards, so we averaged the four items as a scale score.

4.3.1. Idea generation We measured four facets of idea generation: fluency, flexibility, originality, and elaboration (Guilford, 1957; Torrance, 1966). The number of ideas each participant generated determines fluency (e.g., Kohn, Paulus, & Choi, 2011; Kousoulas, 2010; Nijstad et al., 2002). Participants were requested to generate at least three ideas, and they were given a maximum of twenty slots in which they could write ideas. Four graduate student raters1 (blind to experimental condition) determined flexibility. First the authors determined a list of categories for each task. The raters verified this list after they became familiar with the ideas. Then each rater independently counted the number of categories within which each participant listed ideas (e.g., Nijstad et al., 2003). Inter-rater reliability between raters was above recommended cutoffs (0.92), and thus flexibility counts were averaged across raters. To measure the originality of ideas generated, four graduate student raters rated each response for originality. Originality was defined in reference to other participants, such that responses mentioned by

4.3.5. Time on task Given the potentially harmful effects of time pressure, participants were not given a time limit while completing this task. The number of seconds spent completing the creativity task was recorded using the Qualtrics software (average time spent = 17.3 min or 1040 s). This number was included as a control variable to account for the effect of 2 Based on our desire to reduce bias and enhance objectivity we took a number of additional measures beyond training our raters. First, we utilized two male raters and two female raters. Second, while all the raters were graduate students, two raters earned their undergraduate degree at the same institution. Moreover, they ranged in their number of years enrolled at the university (2–6 years). Finally, we used four raters, which goes beyond the typical standard of two to three raters often used in the research area.

1 Graduate student raters received a one-day training on rating practices. Additionally, they received training from the director of Student Life at the University to learn more about student involvement activities already in existence at the University that they may not have been familiar with.

4

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

5.1. Effects of information on idea generation

individual variation in intrinsic task motivation, which has been noted as an important theoretical predictor of creative idea generation (Amabile, 1983). We believe time on task is a good measure of intrinsic task motivation, because individuals who intrinsically enjoy working on the creative task would be more likely to spend more time on the task. Moreover, as discussed in Eisenberger and Cameron (1996), time on task is a common way to measure intrinsic motivation given that it is a measure of free time spent on the task. We believe our study meets these qualifications, because participants were not required to spend a certain amount of time on the task and thus the variation in time spent should, in part, reflect one's intrinsic interest in the task. Finally, we suggest time on task is a better assessment of intrinsic motivation than a generic scale of intrinsic motivation because it is directly related to the specific task, and while someone may enjoy creative tasks, they may not have enjoyed this particular task because of the task content.

To test Hypotheses 1–3, we used model 1 of the PROCESS macro, which tests simple moderation. In this case, factual information was entered as the predictor variable (X) and range information was entered as the moderator (M). This model was tested separately for each idea generation measure. Thus, four models were tested (see Table 2 below). First, we examined the effect of factual information on idea generation. Results indicate that presentation of factual information decreased idea generation fluency (B = − 0.84, p < 0.01), but have no effect on elaboration. Thus, Hypothesis 1a was not supported. Factual information also decreased idea generation flexibility (B = −0.50, p = 0.05), but factual information did not affect originality. Thus, Hypothesis 1b was only partially supported. Next, we examined the main effect of range information on idea generation. As expected, there was a significant positive effect of range information on average idea originality (B = 0.61, p = 0.02). However, no effect was found for flexibility or elaboration. Consequently, Hypothesis 2a was partially supported. Hypothesis 2b predicted that range information would decrease fluency, and the results confirm this hypothesis (B = −0.68, p < 0.01). Together, these results suggest that range information can help improve idea generation quality, but this may come at the expense of reducing the quantity of ideas produced. To test Hypothesis 3, which predicted that the presentation of both factual and range information would improve idea generation across all four dimensions compared to all other conditions, we examined the interaction of factual and range information on idea generation. We found a significant interaction effect on fluency (B = 0.78, p = 0.01). However, the interaction suggested that those who received no information produced more ideas than those who received factual, range, or both types of information. This is counter to expectation, and no other interactions were significant. Thus, Hypothesis 3 was not supported.

4.4. Analytic procedure All hypotheses were tested using the PROCESS macro in SPSS (Hayes, 2013) with 5000 bootstrap samples for bias corrected bootstrap confidence intervals. We chose to test our hypotheses using regression (rather than ANOVA) for two reasons. First, according to Hayes (2013), “mathematically, factorial analysis of variance is identical to the regression-based procedure (p. 211).” In other words, results are identical when testing them using an ANOVA compared to a regression. Second, given that our hypotheses included tests of conditional indirect effects, and our moderator variables are continuous, the regression analysis allowed us to test all our hypotheses. As Hayes (2013) points out, the main advantage of regression analysis over ANOVA is that its more flexible and allows for analysis of any type of variable, including dichotomous and continuous.

5. Results

5.2. Conditional effect of creative thinking skill

Table 1 shows the bivariate correlations among the study variables. Notably, type of information is unrelated to creative thinking skill (rs = 0.02–0.04, ns), but factual information is weakly correlated with domain knowledge (r = 0.20, p < 0.05). The four idea generation measures range in correlations with one another from null (r = − 0.05, ns) to strongly positive (r = 0.74, p < 0.05), with the strongest correlation between fluency and flexibility. The relationship between outcome novelty and outcome usefulness is also noteworthy in that they are strongly negatively correlated (r = − 0.51, p < 0.01). In this scenario novelty seems to come at the cost of usefulness and vice versa.

To test Hypothesis 4, we used Model 11 of the PROCESS macro, which tests for conditional indirect effects (or moderated moderated mediation). In this case, factual information was entered as the predictor variable (X), range information was entered as the first moderator (W), and creative thinking skill was entered as the second moderator (Z; variable is not mean-centered). Using model 11, we also tested the indirect effects of information on the COE dimensions (novelty and usefulness), and the idea generation measures were entered simultaneously as mediators (M). Table 3 below presents the results of the 3-way interaction on each idea generation measure, and Table 4 presents the effect of idea generation on COE.

Table 1 Bivariate correlations between all study variables.

1. Factual information 2. Range information 3. Creative thinking skill 4. Domain knowledge 5. Time on taska 6. IG: fluency 7. IG: flexibility 8. IG: originality 9. IG: elaboration 10. Outcome novelty 11. Outcome usefulness

M

SD

1

2

3

4

5

6

7

8

9

10

0.09 3.69 1.05 3.02 2.97 3.03 2.97 4.25 5.17

0.87 1.46 0.93 0.92 0.99 0.99 0.95 1.13 1.02

0.01 0.04 0.20⁎ − 0.07 − 0.22⁎ − 0.17 − 0.06 − 0.05 0.05 0.04

0.02 − 0.09 0.08 − 0.14 − 0.04 0.22⁎ 0.10 0.05 − 0.07

0.32⁎⁎ 0.27⁎⁎ 0.33⁎⁎ 0.24⁎⁎ 0.03 0.20⁎ 0.25⁎⁎ − 0.14

0.04 0.08 0.03 0.11 − 0.06 0.16 − 0.02

0.23⁎⁎ 0.21⁎ − 0.05 0.38⁎⁎ 0.05 0.05

0.74⁎⁎ 0.00 0.42⁎⁎ − 0.02 − 0.01

0.03 0.38⁎⁎ 0.05 − 0.02

0.18⁎ 0.27⁎⁎ − 0.32⁎⁎

0.04 0.10

−0.51⁎⁎

N = 127. a Value represents thousands of seconds. ⁎ p < 0.05. ⁎⁎ p < 0.01.

5

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

Table 2 Effect of information type in the four facets of idea generation. Fluency

Control variables Creative thinking Domain knowledge Time on task Main effects Factual info Range info Factual × range

Flexibility

Originality

Elaboration

B

SE

t

p

B

SE

t

p

B

SE

t

p

B

SE

t

p

0.29 −0.00 0.00

0.09 0.05 0.00

3.19 − 0.06 2.02

0.00 0.95 0.05

0.23 − 0.01 0.00

0.11 0.06 0.00

2.19 − 0.21 1.74

0.03 0.83 0.09

0.01 0.10 −0.00

0.11 0.06 0.00

0.07 1.58 − 0.89

0.94 0.12 0.38

0.17 − 0.07 0.00

0.10 0.06 0.00

1.68 − 1.31 3.89

0.10 0.19 0.00

−0.84 −0.68 0.78

0.22 0.21 0.29

− 3.83 − 3.22 2.64

0.00 0.00 0.01

− 0.50 − 0.26 0.31

0.26 0.24 0.34

− 1.96 − 1.06 0.90

0.05 0.29 0.37

−0.04 0.61 −0.26

0.26 0.25 0.35

− 0.17 2.42 − 0.73

0.86 0.02 0.47

0.04 0.17 − 0.11

0.24 0.23 0.32

0.19 0.76 − 0.33

0.85 0.45 0.74

R2 = 0.25 F (6120) = 6.65, p < 0.01

R2 = 0.12 F (6120) = 2.74, p = 0.02

R2 = 0.08 F (6120) = 1.75, p = 0.11

R2 = 0.18 F (6120) = 4.27, p = 0.00

N = 127. Bolded text depict statistically significant effects.

We examined the 2-way interaction coefficients with domain knowledge. In this case, domain knowledge did not moderate the effects of information on any facet of idea generation. We also examined the 3way interaction coefficients. Here, there was a significant 3-way interaction for fluency, and for flexibility the 3-way interaction approaches statistical significance, with p = 0.06. Thus, we examined the simple slopes to determine where the conditional effects are significant. Simple slopes analysis revealed that factual information has a negative effect on fluency (estimates = − 0.59 to −0.96, CIs = − 0.02, − 1.52) except when range information is also presented and individuals are low in domain knowledge (estimate = 0.47, CI = 0.09, 1.02), here the effect nearly flips to positive. For flexibility, the combination of factual and range information has a positive effect when individuals are low in domain knowledge (estimate = 1.02, CI = 0.02, 2.02), but all other conditions were non-significant.

We examined the 2-way interaction coefficients with creative thinking skill. In this case, creative thinking skill moderated the effect of range information on idea generation elaboration such that range information enhances elaboration when individuals are high in creative thinking skill (estimate = 0.70, CI = 0.06, 1.34). We also examined the 3-way interaction coefficients. None of these coefficients are significant according to the traditional 0.05 cut-off for p. However, for both flexibility and elaboration, the 3-way interaction approaches statistical significance, with p = 0.06. Thus, we examined the simple slopes to determine where the conditional effects might be significant. Simple slopes analysis revealed that factual information has a negative effect on flexibility when range information is also presented and individuals are high in creative thinking skill (estimate = − 0.68, CI = − 1.34, − 0.01); in all other conditions, factual information has no effect on flexibility. For elaboration, factual information similarly has a negative effect when range information is also presented and individuals are high in creative thinking skill (estimate = − 0.73, CI = − 1.34, − 0.13), but all other conditions were non-significant.

5.4. Direct effects and conditional indirect effects on COE Next, we examined the effect of idea generation on outcome novelty as well as the indirect effect of information (see Tables 4 & 6). The average originality of the ideas produced during idea generation had a positive effect on outcome novelty (B = 0.29, p = 0.01). No other aspects of idea generation had a significant effect on outcome novelty. Range information did, however, have an unconditional indirect effect (estimate = 0.14, CI = 0.03, 0.31) and a conditional indirect effect on

5.3. Conditional effect of domain knowledge To test Hypothesis 5, we again used Model 11 of the PROCESS macro. In this case, domain knowledge was entered as the second moderator (Z; variable is not mean-centered). Table 5 presents the results of the 3-way interaction on each idea generation measure.

Table 3 Results of regression analysis from independent variables to mediators, creative thinking skill as moderator. Fluency

Control variables Domain knowledge Time on task Main effects Factual info Range info Creative thinking 2-Way interactions Factual × range CT × factual CT × range 3-Way interaction CT × F × R

Flexibility

Originality

Elaboration

B

SE

t

p

B

SE

t

p

B

SE

t

p

B

SE

t

p

−0.01 0.00

0.05 0.00

− 0.23 2.20

0.82 0.03

− 0.02 0.00

0.06 0.00

− 0.35 1.76

0.73 0.08

0.10 −0.00

0.06 0.00

1.60 − 0.86

0.11 0.39

− 0.10 0.00

0.06 0.00

− 1.67 4.14

0.10 0.00

−0.82 −0.67 0.40

0.26 0.21 0.16

− 3.76 − 3.19 2.55

0.00 0.00 0.01

− 0.51 − 0.27 0.18

0.25 0.24 0.19

− 1.99 − 1.10 0.98

0.05 0.27 0.33

−0.03 0.62 0.11

0.26 0.25 0.19

− 0.13 2.49 0.57

0.90 0.01 0.57

0.05 0.16 0.15

0.23 0.22 0.17

0.21 0.73 0.87

0.83 0.47 0.39

0.82 −0.18 0.18

0.29 0.24 0.25

2.79 − 0.74 0.74

0.01 0.46 0.46

0.37 0.19 0.38

0.34 0.28 0.29

1.08 0.66 1.32

0.28 0.51 0.19

−0.19 0.13 0.04

0.35 0.29 0.29

− 056 0.46 0.14

0.58 0.65 0.89

− 0.05 − 0.08 0.55

0.31 0.26 0.26

− 0.15 − 0.29 2.12

0.88 0.77 0.04

−0.35

0.34

− 1.02

0.31

− 0.74

0.40

− 1.87

0.06

−0.60

0.40

− 1.49

0.14

− 0.69

0.36

− 1.90

0.06

2

R = 0.28 F (9117) = 5.13, p < 0.01

2

R = 0.15 F (9117) = 2.34, p = 0.02

N = 127, CT = creative thinking skill, F = factual information, R = range information. Bolded text depict statistically significant effects.

6

2

R = 0.12 F (9117) = 1.79, p = 0.08

2

R = 0.24 F (9117) = 4.03, p = 0.00

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

Table 4 Direct and indirect effects on COE. Novelty

Mediators to DVs IG: fluency IG: flexibility IG: originality IG: elaboration Indirect effect of factual info IG: fluency IG: flexibility IG: originality IG: elaboration Indirect effect of range info IG: fluency IG: flexibility IG: originality IG: elaboration

Usefulness

B

SE

t/LLCI

p/ULCI

B

SE

t/LLCI

p/ULCI

− 0.18 0.16 0.29 − 0.01

0.17 0.15 0.10 0.13

−1.10 1.07 2.76 −0.08

0.27 0.29 0.01 0.90

−0.08 −0.05 −0.37 0.24

0.15 0.13 0.09 0.11

− 0.51 − 0.35 − 4.02 2.21

0.58 0.74 0.00 0.03

0.05 − 0.05 − 0.04 0.00

0.08 0.06 0.06 0.02

−0.08 −0.22 −0.17 −0.03

0.24 0.04 0.07 0.07

0.03 0.02 0.04 −0.01

0.05 0.04 0.07 0.04

− 0.05 − 0.06 − 0.08 − 0.09

0.18 0.13 0.20 0.06

0.05 − 0.02 0.14 − 0.01

0.07 0.05 0.07 0.03

−0.04 −0.18 0.03 −0.11

0.26 0.04 0.31 0.03

0.02 0.00 −0.16 0.24

0.05 0.03 0.08 0.11

− 0.03 − 0.03 − 0.35 − 0.21

0.20 0.10 −0.05 0.03

Note. Indirect effects were tested with a 95% Confidence Interval (CI). N = 127, IG = idea generation. Bolded text depict statistically significant effects.

presented and creative thinking skill is high (estimate = −0.18, CI = −0.53, − 0.01), Here, receiving both types of information decreased idea elaboration for individuals high in creative thinking skill, which decreases outcome usefulness. However, this effect appears to be limited to presentation of factual information, because when range information is presented without factual information, we observed a positive effect on usefulness (via elaboration) for those high in creative thinking skill (estimate = 0.18, CI = 0.01, 0.48). We also examined the indirect effect of range information on outcome usefulness. We found that range information has negative indirect effect on outcome usefulness via idea generation originality (estimate = −0.16, CI = −0.35, − 0.05). The negative effect of range information on outcome usefulness via idea generation originality is true for those at high levels of creative skill (estimates = −0.25, CI = −0.62, − 0.02). However, for those low in creative thinking skill range information only had a negative effect when it was paired with factual information (estimate = −0.32, CI = −0.71, − 0.07).

outcome novelty via idea generation originality such that range information has a positive indirect effect (estimate = 0.25, CI = 0.05, 0.60) when factual information is present and creative thinking skill is low. Moreover, range information has a positive indirect effect via originality when creative thinking skill is high (estimate = 0.19, CI = 0.01, 53). When examining the direct and indirect effects on outcome usefulness (see Table 4), we found that originality has a negative direct effect (B = − 0.37, p < 0.01) and elaboration has a positive direct effect (B = 0.24, p = 0.03). Factual information did not have an unconditional indirect effect on outcome usefulness, but it did have a conditional indirect effect via idea generation originality and elaboration. Specifically, factual information has a positive effect via originality when range information was also present and creative thinking skill is high (estimate = 0.26, CI = 0.02, 0.63). In other words, receiving both types of information hindered idea generation originality for people high in creative thinking skill, which increases outcome usefulness as a result. Moreover, the presentation of factual information has a positive effect via elaboration when range information is also presented and creative thinking skill is low (estimate = 0.15, CI = 0.02, 0.43), but a negative effect on outcome usefulness when range information is also

6. Discussion In line with our expectations, the results broadly suggest that

Table 5 Results of regression analysis from independent variables to mediators, domain knowledge as moderator. Fluency

Control variables Creative thinking Time on task Main effects Factual info Range info Domain Knowledge 2-Way interactions Factual × range DK × factual DK × range 3-Way interaction DK × F × R

Flexibility

Originality

Elaboration

B

SE

t

p

B

SE

t

p

B

SE

t

p

B

SE

t

p

0.30 0.00

0.09 0.00

3.33 2.57

0.00 0.01

0.24 0.00

0.11 0.00

2.22 2.21

0.03 0.03

0.01 −0.00

0.11 0.00

0.10 − 0.76

0.92 0.45

0.18 0.00

0.10 0.00

1.76 4.07

0.08 0.00

− 0.95 − 0.79 0.05

0.60 0.53 0.11

− 1.58 − 1.49 0.48

0.12 0.14 0.63

−0.77 −0.74 −0.02

0.71 0.63 0.13

−1.09 −1.18 −0.17

0.28 0.24 0.87

−0.09 0.70 0.13

0.74 0.66 0.13

− 0.12 1.07 0.98

0.90 0.29 0.33

− 0.10 0.35 − 0.02

0.66 0.59 0.12

−0.15 0.60 −0.18

0.88 0.56 0.86

2.37 0.05 0.04

0.83 0.15 0.14

2.86 0.48 0.30

0.01 0.63 0.77

2.07 0.07 0.15

0.97 0.17 0.17

2.13 0.40 0.88

0.04 0.69 0.38

0.11 0.01 −0.03

1.0 0.18 0.17

0.11 0.05 − 0.15

0.91 0.95 0.88

0.76 0.03 − 0.05

0.91 0.16 0.16

0.84 0.20 −0.31

0.40 0.84 0.76

− 0.41

0.21

− 2.01

0.05

−0.47

0.24

−1.94

0.06

−0.09

0.26

− 0.36

0.72

− 0.22

0.23

−0.95

0.36

2

R = 0.30 F (9117) = 5.69, p < 0.01

2

R = 0.16 F (9117) = 2.55, p = 0.01

N = 127, DK = domain knowledge, F = factual information, R = range information. Bolded text depict statistically significant effects.

7

2

R = 0.08 F (9117) = 1.20, p = 0.30

2

R = 0.20 F (9117) = 3.18, p < 0.01

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

Table 6 Conditional indirect effects on COE. Novelty B Conditional indirect effect No range, low CT No range, high CT With range, low CT With range, high CT Conditional indirect effect No range, low CT No range, high CT With range, low CT With range, high CT Conditional indirect effect No factual, low CT No factual, high CT With factual, low CT With factual, high CT Conditional indirect effect No factual, low CT No factual, high CT With factual, low CT With factual, high CT

Usefulness SE

of factual info (mediator: originality) − 0.04 0.09 0.02 0.10 0.04 0.13 − 0.20 0.12 of factual info (mediator: elaboration) − 0.00 0.04 0.00 0.04 − 0.01 0.10 0.01 0.11 of range info (mediator: originality) 0.17 0.13 0.19 0.13 0.25 0.13 − 0.03 0.10 of range info (mediator: elaboration) 0.00 0.05 − 0.01 0.11 − 0.00 0.06 0.00 0.05

LLCI

ULCI

B

SE

LLCI

ULCI

−0.25 −0.14 −0.21 −0.49

0.12 0.29 0.34 − 0.02

0.05 −0.04 −0.05 0.26

0.11 0.13 0.17 0.15

− 0.16 − 0.31 − 0.44 0.02

0.30 0.21 0.27 0.63

−0.12 −0.07 −0.28 −0.20

0.05 0.08 0.15 0.27

0.03 −0.01 0.15 −0.18

0.06 0.06 0.10 0.12

− 0.07 − 0.14 0.02 − 0.53

0.16 0.13 0.43 −0.01

−0.02 0.01 0.05 −0.26

0.49 0.53 0.60 0.03

−0.22 −0.25 −0.32 0.04

0.15 0.15 0.16 0.12

− 0.59 − 0.62 − 0.71 − 0.19

0.03 −0.02 −0.07 0.31

−0.09 −0.24 −0.16 −0.09

0.14 0.23 0.07 0.11

−0.07 0.18 0.05 −0.00

0.06 0.12 0.08 0.08

− 0.23 0.01 − 0.05 − 0.18

0.02 0.48 0.31 0.16

Note. Indirect effects were tested with a 95% Confidence Interval (CI). N = 127, CT = creative thinking skill. Bolded text depict statistically significant effects.

allow people to reap the benefits of each type. In the current study, this was only true when predicting idea generation flexibility for individuals low in domain knowledge. In all other instances, the effect of combined information was not better than any of the other conditions on the creativity criteria. In fact, these individuals rarely outperformed the individuals that were given no information. To explain this finding, we reason that being presented with both types of information may create greater cognitive load by causing people to split attention and devote less cognitive resources to idea generation behavior itself. Seeing the two varying types of information could cause the person to redirect cognitive resources toward understanding and interpreting these differences. The two types of information may have also sent “mixed signals” to participants concerning the type of creative solutions being sought, leading to resources being directed away from idea generation toward resolution of these signaled goals, and possibly even problem redefinition. Consequently, managers should consider what is more important for COE (novelty or usefulness) when facilitating and structuring creative tasks for employees. Our findings have two additional research implications. First, information seems to affect different sub-criteria of creative outcomes differently. Originality of idea generation seems to have opposing effects on novelty and usefulness, with elaboration having effects only on the usefulness dimension. Specifically, originality of idea generation had a positive effect on novelty but a negative effect on usefulness. This supports the rarely explored proposition that improving creative outcomes may involve trade-offs. This confirms the need for creativity researchers to examine the determinants of each formative dimension of COE separately. This raises the larger question of whether there are creative outcomes for which novelty or usefulness is weighted more heavily than the other by the end user (Montag et al., 2012). Our negative correlation between the dimensions raises the compelling question, “Are there some types of creative outcomes, where novelty and usefulness are negatively correlated, such that more novelty or usefulness beyond a point means less COE?” That is, can the creative solution be too novel to be practical? This finding and interpretation challenges the assumption of most measures that more of both novelty and usefulness means more COE. Too much novelty being associated with less COE has recently been demonstrated in an in-role R & D task and setting (Criscuolo, Dahlander, Grohsjean, & Salter, 2017). Through this stream of research, we may learn about different criterion profiles of COE

introducing information during creative tasks can facilitate both idea generation and COE. Moreover, this research shows that the type of external information matters greatly. Rather than offering simple prescriptions for information interventions that promote extra-role creativity, our research uncovers a more nuanced story. First, our findings suggest that no information is actually better than introducing any information for idea fluency (i.e., number of ideas) during idea generation, which aligns with research from Svensson, Norlander, and Archer (2002). This makes sense given that viewing ideas from others can cause production blocking (Nijstad & Stroebe, 2006). The effect, however, appears to be limited to fluency; the presence of information did not have a universally negative effect on the other aspects of idea generation or COE. This calls into question research that has relied solely on fluency as a measure of “creativity”, particularly given that fluency did not correlate with idea originality (see also, Baruah & Paulus, 2008) or COE. In terms of the specific types of information, factual information seems to have mixed effects, and some of these effects seem to be dependent on creative thinking skill. For people high in creative thinking skill, factual information can increase COE usefulness by reducing idea generation originality, but it can also reduce usefulness by reducing idea elaboration. For individuals low in creative thinking skill, factual information increases elaboration and COE usefulness. Based on these results, factual information is most likely to increase COE usefulness, but the mechanism through which this happens seems to depend on level of creative thinking skill. Range information, on the other hand, significantly enhances idea generation originality, which in turn enhances COE novelty and generally reduces COE usefulness. This aligns with previous research from Kohn and colleagues (Kohn, Paulus, & Choi, 2011; Kohn, Paulus, & Korde, 2011), who also found that divergent stimuli (e.g., brainstormed ideas) enhance the novelty of ideas generated. Extending this research here, we find that average originality of ideas generated influences COE. This suggests that “outside the box” information stimulates a greater level of novelty when generating and refining creative ideas. Thus, information that can be related in indirect, unconventional ways to the task holds the greatest promise for outcome novelty. However, this same process may reduce outcome usefulness. With regard to presenting both factual and range information, it seemed reasonable that presenting multiple types of information would 8

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

individuals to self-report their perceived level of knowledge. Given that this assessment method could be prone to error, future research could examine the domain knowledge moderator with a direct, objective measure. Finally, in our 2 × 2 design, we controlled the amount of information across the three experimental conditions, resulting in half the factual and half the range information being presented in the combined condition. One could argue that there may not have been enough of each kind of information. Thus, we acknowledge that our results may have been different had we presented the full amount of both factual and range information in this condition, although not controlling for information amount would have created an alternative explanation and further risked information overload in the mixed information condition. Despite these limitations, this study advanced understanding of the role of information types in extra-role creativity tasks and the individual characteristics that moderate these relationships. We hope that this will inspire more such research across organizational settings.

across situations and creative problem types. Managers may eventually be able to learn how to influence the aspect of COE most needed in the situation/problem type. Second, although we focused on information effects through idea generation, idea evaluation and problem definition are other key behaviors in the creative process. It is likely that information dimensions as well as creative thinking skill and domain knowledge exert their influence through problem definition and idea evaluation as well. Domain knowledge, arguably stored information, has been associated with effective idea evaluation (Friedrich & Mumford, 2009). These findings suggest that having factual information at hand benefits idea evaluation, which in turn, leads to higher COE. Overall, future creativity research should address how externally-presented information type affects other creative behaviors in the process. 6.1. Practical implications The overarching practical implication of this research is that inexpensive and practical information interventions hold real promise as a way to improve creativity, especially among those already skilled in creative thinking tasks. Here we support this in creative tasks “for the masses” that can be applied at most any organization, like requests from all stakeholders for creative solutions to improve employee morale, operating efficiency, or customer growth. In fact, this direct challenge, facilitated by an explicit goal to be creative, a small extrinsic incentive for participation (as in the current study), and some targeted informational support from the organization, should ensure more engagement in problem-solving than generic suggestion programs. For example, as seen in the area of “crowdsourcing” ideas in marketing practice and research (Luo & Toubia, 2015), suggestion programs can be structured to organize and provide stimulating information. Within such tasks, both factual or range information domains could be defined for most any creative problem-solving task and samples from these domains would cost little to distribute selectively through a project website, social media account, or emailed links. This research has practical implications for the types of examples managers may provide employees. If managers are looking for highly novel solutions, our research suggests that providing employees with example ideas or other resources that represent the typical and easily accessible factual information will not help to achieve the goal of novelty. On the other hand, if a manager can present employees with resources that provide information that is more distally related to the issue (range information), this may improve outcome novelty. Interestingly, we found that the blending of these two types of information did not have a positive effect on creativity, so managers should aim to be selective in the type of information they provide, rather than overloading employees' working memory capacity with too many types of information.

References Amabile, T. M. (1983). The social psychology of creativity: A componential conceptualization. Journal of Personality and Social Psychology, 45(2), 357–376. Amabile, T. M. (1996). Creativity in context. Westview Press. Amabile, T. M., Barsade, S. G., Mueller, J. S., & Staw, B. M. (2005). Affect and creativity at work. Administrative Science Quarterly, 50(3), 367–403. Ardichvili, A., Cardozo, R., & Ray, S. (2003). A theory of entrepreneurial opportunity identification and development. Journal of Business Venturing, 18, 105–123. http://dx. doi.org/10.1016/S0883-9026(01)00068-4. Axtell, C., Holman, D., & Wall, T. (2006). Promoting innovation: A change study. Journal of Occupational and Organizational Psychology, 79, 509–516. http://dx.doi.org/10. 1348/096317905X68240. Baruah, J., & Paulus, P. B. (2008). Effects of training on idea generation in groups. Small Group Research, 39, 523–541. http://dx.doi.org/10.1177/1046496408320049. Brown, V. R., & Paulus, P. B. (2002). Making group brainstorming more effective: Recommendations from an associative memory perspective. Current Directions in Psychological Science, 11(6), 208–212. Burbiel, J. (2009). Creativity in research and development environments: A practical review. International Journal of Business Science and Applied Management, 4(2), 35–51. Carson, S. H., Peterson, J. B., & Higgins, D. M. (2003). Decreased latent inhibition is associated with increased creative achievement in high-functioning individuals. Journal of Personality and Social Psychology, 85(3), 499–506. http://dx.doi.org/10. 1037/0022-3514.85.3.499. Chávez-Eakle, R. A., Graff-Guerrero, A., García-Reyna, J. C., Vaugier, V., & Cruz-Fuentes, C. (2007). Cerebral blood flow associated with creative performance: A comparative study. NeuroImage, 38, 519–528. http://dx.doi.org/10.1016/j.neuroimage.2007.07. 059. Choi, J. N., Anderson, T. A., & Veillette, A. (2009). Contextual inhibitors of employee creativity in organizations the insulating role of creative ability. Group & Organization Management, 34(3), 330–357. Cramond, B., Matthews-Morgan, J., Bandalos, D., & Zuo, L. (2005). A report on the 40year follow-up of the Torrance Tests of Creative Thinking: Alive and well in the new millennium. The Gifted Child Quarterly, 49(4), 283–291. Criscuolo, P., Dahlander, L., Grohsjean, T., & Salter, A. (2017). Evaluating novelty: The role of panels in the selection of R & D projects. Academy of Management Journal, 60(2), 433–460. Davidson, J. E., & Sternberg, R. J. (1984). The role of insight in intellectual giftedness. The Gifted Child Quarterly, 28(2), 58–64. Diehl, M., & Stroebe, W. (1987). Productivity loss in brainstorming groups: Toward the solution of a riddle. Journal of Personality and Social Psychology, 53(3), 497–509. Duncker, K. (1945). On problem-solving. Psychological Monographs, 58(5), 1–113. Eisenberger, R., & Cameron, J. (1996). Detrimental effects of reward reality or myth? American Psychologist, 51(11), 1153–1166. Friedman, R. S., & Forster, J. (2001). The effects of promotion and prevention cues on creativity. Journal of Personality and Social Psychology, 81(6), 1001–1013. Friedrich, T. L., & Mumford, M. D. (2009). The effects of conflicting information on creative thought: A source of performance improvements or decrements? Creativity Research Journal, 21, 265–281. http://dx.doi.org/10.1080/10400410902861430. Guilford, J. P. (1957). Creative abilities in the arts. Psychological Review, 64(2), 110–118. Guilford, J. P. (1975). Varieties of creative giftedness, their measurement and development. The Gifted Child Quarterly, 19(2), 107–121. Hambrick, D. Z., & Engle, R. W. (2002). Effects of domain knowledge, working memory capacity, and age on cognitive performance: An investigation of the knowledge-ispower hypothesis. Cognitive Psychology, 44(4), 339–387. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis. New York: The Guilford Press. Hong, E., & Milgram, R. M. (2010). Creative thinking ability: Domain generality and specificity. Creativity Research Journal, 22(3), 272–287. http://dx.doi.org/10.1080/ 10400419.2010.503535. Jin, S. H., Kwon, Y. J., Jeong, J. S., Kwon, S. W., & Shin, D. H. (2006). Increased information transmission during scientific hypothesis generation: Mutual information

6.2. Limitations This study is not without limitations. First, our study has a small sample size, which may limit our ability to detect some effects, resulting in Type II error. However, with a sample of 127, power analysis suggests that we can detect an effect of 0.06 (assuming power of 0.8 and a p-value of 0.05), which gives us confidence that Type II error was limited, and likely only applies to very small effects. Second, we utilize a partly student sample working on a problem at a university, which may limit the generalizability of our study. However, the task is nearly identical to that which could be done in any organization where stakeholders suggest creative ideas for how to improve employee engagement or customer experience. Likewise, we did not detect differences within our sample based student or employee status. Thus, the creative process undertaken by students does not seem to differ from that of employees within the organization. Second, we did not measure domain knowledge with a pure test of knowledge. Instead, we asked 9

Journal of Business Research 81 (2017) 1–10

T. Montag-Smit, C.P. Maertz

factors at work. Academy of Management Journal, 39(3), 607–634. http://dx.doi.org/ 10.2307/256657. Osborn, A. F. (1953). Applied Imagination: Principles and Procedures of Creative Problem Solving. New York: Charles Scribner's Sons. Runco, M. A. (2004). Creativity. Annual Review of Psychology, 55, 657–687. http://dx.doi. org/10.1146/annurev.psych.55.090902.141502. Smith, R. E., & Yang, X. (2004). Divergence toward a general theory of creativity in advertising: Examining the role of. Marketing Theory, 4, 31–58. http://dx.doi.org/10. 1177/1470593104044086. Sosik, J. J., Kahai, S. S., & Avolio, B. J. (1998). Transformational leadership and dimensions of creativity: Motivating idea generation in computer-mediated groups. Creativity Research Journal, 11(2), 111–121. Sullivan, D. M., & Ford, C. M. (2010). The alignment of measures and constructs in organizational research: The case of testing measurement models of creativity. Journal of Business and Psychology, 25, 505–521. http://dx.doi.org/10.1007/s10869-0099147-8. Svensson, N., Norlander, T., & Archer, T. (2002). Effects of individual performance versus group performance with and without de Bono techniques for enhancing creativity. The International Journal of Creativity & Problem Solving, 12(2), 15–34. Torrance, E. P. (1966). Tests of creative thinking. Personnel Press. Unsworth, K. (2001). Unpacking creativity. Academy of Management Review, 26(2), 289–297. Wilson, R. C., Guilford, J. P., Christensen, P. R., & Lewis, D. J. (1954). A factor-analytic study of creative-thinking abilities. Psychometrika, 19(4), 297–311.

analysis of multichannel EEG. International Journal of Psychophysiology, 62, 337–344. http://dx.doi.org/10.1016/j.ijpsycho.2006.06.003. Jones, E. E., & Kelly, J. R. (2009). No pain, no gains: Negative mood leads to process gains in idea-generation groups. Group Dynamics: Theory, Research, and Practice, 13(2), 75–88. http://dx.doi.org/10.1037/a0013812. Kesting, P., & Ulhøi, J. P. (2010). Employee-driven innovation: Extending the license to foster innovation. Management Decision, 48(1), 65–84. http://dx.doi.org/10.1108/ 00251741011014463. Kohn, N. W., Paulus, P. B., & Choi, Y. (2011). Building on the ideas of others: An examination of the idea combination process. Journal of Experimental Social Psychology. http://dx.doi.org/10.1016/j.jesp.2011.01.004. Kohn, N. W., Paulus, P. B., & Korde, R. M. (2011). Conceptual combinations and subsequent creativity. Creativity Research Journal, 23(3), 203–210. http://dx.doi.org/10. 1080/10400419.2011.595659. Kousoulas, F. (2010). The interplay of creative behavior, divergent thinking, and knowledge base in students' creative expression during learning activity. Creativity Research Journal, 22(4), 387–396. http://dx.doi.org/10.1080/10400419.2010. 523404. Luo, L., & Toubia, O. (2015). Platforms and customizing the task knowledge. Journal of Marketing, 79(September), 100–114. http://dx.doi.org/10.1509/jm.13.0212. Malhotra, N. K. (1982). Information load and consumer decision making. Journal of Consumer Research, 8, 419–430. Marrocu, E., & Paci, R. (2012). Education or creativity: What matters most for economic performance? Economic Geography, 88(4), 369–401. Mobley, M. I., Doares, L. M., & Mumford, M. D. (1992). Process analytic models of creative capacities: Evidence for the combination and reorganization process. Creativity Research Journal, 5(2), 125–155. Montag, T., Maertz, C. P., & Baer, M. (2012). A critical analysis of the workplace creativity criterion space. Journal of Management, 38(4), 1362–1386. http://dx.doi.org/ 10.1177/0149206312441835. Mumford, M. D., Baughman, W. A., Supinski, E. P., & Maher, M. A. (1996). Process-based measures of creative problem-solving skills: II. Information encoding. Creativity Research Journal, 9(1), 77–88. Nijstad, B. A., & Stroebe, W. (2006). How the group affects the mind: A cognitive model of idea generation in groups. Personality and Social Psychology Review, 10(3), 186–213. Nijstad, B. A., Stroebe, W., & Lodewijkx, H. F. M. (2002). Cognitive stimulation and interference in groups: Exposure effects in an idea generation task. Journal of Experimental Social Psychology, 38, 535–544. Nijstad, B. A., Stroebe, W., & Lodewijkx, H. F. M. (2003). Production blocking and idea generation: Does blocking interfere with cognitive processes? Journal of Experimental Social Psychology, 39, 531–548. http://dx.doi.org/10.1016/S0022-1031(03)00040-4. Oldham, G. R., & Cummings, A. (1996). Employee creativity: Personal and contextual

Tamara Montag-Smit is an Assistant Professor of Management in the Miller College of Business at Ball State University. She earned her Ph.D. from Saint Louis University in Industrial and Organizational Psychology. She conducts research on workplace creativity and employee compensation practices. Her research has been published in several outlets including Journal of Management and Human Relations. Carl P. Maertz, Jr. is the Mary Louise Murray Endowed Professor of Management in the John Cook School of Business at Saint Louis University. He earned his Ph.D. from Purdue University's Krannert Graduate School of Management. His research on work-family conflict, cross-cultural adjustment, turnover, selection fairness, and workplace creativity has appeared in Academy of Management Journal, Psychological Bulletin, Organizational Behavior and Human Decision Processes, Journal of Applied Psychology, Journal of Management, Journal of Organizational Behavior, Industrial Relations & many others. He has also consulted with organizations in various areas of management for over 25 years, specializing in the improvement of employee work performance and retention.

10