The Emergence of Problem Construction at the Team-Level

Abstract

The present study progresses the understanding of the role problem construction plays in team creative problem-solving. Analyzing recorded teamwork sessions revealed that teams spent the great majority of their time engaged in problem construction behaviors (53.56% of data) compared to idea generation (28.03%), or idea evaluation (18.56%). A social sequence analysis did not reveal any statistically significant sequences of processes utilized by teams. We discuss the breakdown of specific behaviors within each class, provide qualitative descriptions of team processes, and discuss how problem construction was employed and observed at the team level.

Keywords

qualitative research creativity teamwork innovation creative teams

Organizations are increasingly emphasizing the importance of creative problem-solving to remain competitive in dynamic economic work environments (Shalley & Gilson, 2004). Creative individuals may generate socially useful ideas when designing new products, conceive efficiencies to provide production advantages, and introduce novel ideas to emerging or niche markets (Anderson et al., 2004). Moreover, access to new markets arising from innovations may facilitate organizational long-term success in fast-moving work environments (Anderson et al., 2014). Indeed, implementing creative solutions to workplace problems has been identified as one of the most critical skills for 21st century organizations (National Research Council, 2012). Simultaneously, organizations have responded to the dynamic nature of modern work environments with an increased reliance on teams (West et al., 2004). Teams are well-situated to address complex problems that tend to be ill-defined or contain conflicting and ambiguous information or assumptions (Reiter-Palmon et al., 1997). Teams may be particularly proficient at solving complex problems creatively, as individuals working together may be able to draw from a larger pool of past problem-solving experience to generate more potential solutions (Osborn, 1953; Puccio et al., 2005).

Due to the complexity and multifaceted nature of the creativity construct, various perspectives on creativity have arisen. In one prevalent taxonomy, Rhodes (1961) proposed the Four P’s of creativity, or four perspectives of creativity, referring to the person, press, product, and process. Notably, processes associated with engagement in creative endeavors may span several domains. For example, scholars have previously examined the interplay of social, interpersonal, and/or affective processes that influence the quality or originality of proposed problem solutions, as well as how such variables impact the beneficiaries, or receivers of creativity (Amabile, 1996; Runco, 2003; Zhou et al., 2019). Alternatively, the scope of creative processes may be as granular as the cognitive operations occurring within or between individuals during personal or shared creative problem-solving (Simon & Newell, 1971). Within the last 20 years, scholars have argued that such cognitive processes driving creativity are often overlooked (Drazin et al., 1999; Unsworth, 2001). Creative efforts tend to be primarily operationalized via divergent thinking measures (e.g., “how many problem solutions were generated”) during empirical investigations. This approach to the study of creativity disproportionately emphasizes the product perspective. That is, the use of divergent thinking tasks situates the outputs of a singular creative process as a global measure of creative proficiency rather than operationalizing the totality of cognition driving creativity (Reiter-Palmon et al., 2019). While there are competing theoretical and empirical strengths to each perspective, the present study applies the cognitive process perspective to the examination of teams engaged in creative efforts.

Under a cognitive process perspective of creativity, creative problem-solving is viewed as a sequence of cognitive activities in which an individual, or team, uses divergent and/or convergent thinking styles to identify the most effective solutions to a problem (T. M. Basadur & Basadur, 2013). Notably, problem-solving is an analytical process that involves decision-making. For example, a problem solver might ask themselves, “why is this a problem worth solving,” or “which problem among a set of problems should be solved.” Indeed, a key decision for team members early in the creative problem-solving process is how the team will define and commonly understand the presenting problem. Unfortunately, there is limited team-level research regarding how teams construct problems, or what cognitive activities underly the endeavor of arriving at a shared problem representation (Reiter-Palmon, 2014). Two notable challenges to this line of research hinder such inquiries.

First, there is a series of competing cognitive process models of creativity in extant literature (Amabile, 1988; Anderson et al., 2004). Lacking a consensus on the goals, operations, breadth, and sequence of cognitive processes prevents the nuanced understanding of corresponding behaviors, particularly in complex environments such as teams. However, a review of prominent process models revealed that three common core processes consistently emerge across creativity scholarship: (a) problem construction, (b) idea generation, and (c) idea evaluation (Reiter-Palmon & Illies, 2004). The second challenge to team-level inquiry is that problem construction processes tend to operate within individuals automatically and beneath conscious awareness when cues in the presenting problem scenario activate heuristics encoded in memory (Holyoak, 1984). Thus, it remains unclear how problem construction cognition emerges behaviorally during teamwork vis-à-vis cognitive processes, how problem construction processes are used by teams to solve problems, where problem construction exists in the sequence of processes during teamwork, or how frequently problem construction occurs in relation to other processes. Drawing from the above, the present study attempts to address such challenges by responding to calls for further investigations into which behaviors and cognition are integrated at the team level during creative problem-solving (Anderson et al., 2014; Rosing et al., 2018). Specifically, the present study aims to: (a) determine the frequency of core processes during creative problem-solving, focusing on the frequency of problem construction compared to other processes; (b) identify sequences of core creative processes commonly used, and the role of problem construction in these sequences; (c) qualitatively explore team interaction to identify how problem construction is applied during teamwork.

The Problem Construction Construct: Emergence and Research Gaps

Process models of creativity tend to situate cognitive processes operating to identify and define presenting problems early in the sequence of activities. Simon and Newell (1971) as well as Guilford (1967) refer to such mental operations as “problem structuring,” while M. S. Basadur (1994) uses the term “problem-finding.” Similarly, Amabile (1988, 1996) and Amabile and Pratt (2016) describe such operations as “preparation,” while Mumford et al. (1991) use the term “problem construction,” which we prefer for the present study. Problem construction refers to the cognition involved in the development of internal problem representations. Problem representations are cognitive structures that organize key features of the presenting problem, including the identification of constraints, goals, strategies, and parameters, as well as sensemaking of the current problem based on the activation of past problem-solving experiences embedded in memory (Holyoak, 1984; Reiter-Palmon & Robinson, 2009). Problem construction is conceptualized as strictly an individual-level phenomenon that operates automatically and beneath conscious awareness. Consequentially, individuals are often unsure of precisely when they are engaged in problem construction cognition during creative activities (Mumford et al., 1994; Reiter-Palmon & Murugavel, 2018). This represents a core limitation in previous efforts to examine processes driving creative efforts via prevalent methodologies, such as think-aloud tasks, where participants narrate their thought process during creative endeavors. That is, problem construction cognition is likely not captured by research tasks targeting purposeful thinking and thus have been omitted from previous analyses of creative processes (Boldt, 2019).

Following this limitation, the identification of team-level problem construction behaviors is difficult. However, some research suggests certain social phenomena occur in the context of teamwork which may cue engagement in problem construction cognition. For instance, teams have been shown to develop knowledge structures, referred to as shared mental models, comprised of task-relevant information or commonly understood elements (Cannon-Bowers et al., 1993). Shared mental models have been shown to reduce ambiguity and miscommunication within teams (Ward et al., 1999) and are associated with greater performance on creative tasks (Mumford et al., 2011). Similarly, teams are known to develop a shared understanding of presenting problems unique to the team’s membership. Cronin and Weingart (2007) propose teams form joint representations thought to be comprised of an aggregate of the team members’ individual problem representations. Notably, joint representations do not reflect all aspects of the individual team members’ representations. Cronin and Weingart argue it is impossible to fully understand other members’ past experiences or perceptions that contribute to problem representations, and thus unshared aspects of individual representations that remain unaccounted for will inevitably influence problem-solving. Cronin and Weingart also conclude that simply sharing individual representations will likely not result in a joint representation. Rather, competing representations, or varying definitions of the problem parameters, create disjointed representational gaps (rGaps) reflecting various contradicting elements amongst team members’ perspectives (Cronin & Weingart, 2007).

Paradoxically, empirical evidence suggests rGaps can benefit team creative outcomes. Weingart et al. (2008) demonstrated teams that explicitly address and overcome disjointed problem representations form stronger shared understandings of the presenting problem and display greater coordination during task work than teams that initially experienced narrow, or limited rGaps. The researchers argue that wide rGaps represent an opportunity for teams to engage in task-related conflict that scrutinizes implicit assumptions, promotes exploration and revaluation of opposing viewpoints, and increases the team’s analysis of the task at hand. Such interactions are key goals of the problem construction process and have been linked with greater team creative outcomes (Reiter-Palmon et al., 1997; Shalley & Gilson, 2004; West, 2002).

Considering such team-level phenomena, we argue problem construction emergence at the team-level is represented by behaviors indicating the formation of joint representations and/or shared mental models. Specifically, we focus on team interactions (i.e., verbal utterances) that serve to aggregate team members’ individual problem representations into one that is novel and unique to the team. For example, members may frame or describe aspects of the problem in a unique way during conversation. Researchers have identified framing as one common way that creative teams engage in sensemaking, clarifying, and explanation, potentially contributing to reduced rGaps (Ward et al., 1999; Zahedi & Heaton, 2017). A plethora of evidence suggests that restating, or summarizing, problems in unique ways is associated with the generation of higher quality and more original (i.e., creative) problem solutions (Mumford et al., 1996; Redmond et al., 1993; Reiter-Palmon et al., 1997). A member’s attempt to summarize the problem for the team during teamwork likely represents engagement in problem construction cognition. Additionally, discussion that shares personal experiences during problem framing may serve to align disjointed representations or increase the breadth and scope of knowledge available to the team, creating more robust shared mental models (Cannon-Bowers et al., 1993). It is important to note Cronin and Weingart’s (2007) finding that rGaps are likely to arise; thus, disagreements or discussion surrounding these behaviors likely indicates engagement in problem construction as well.

Unfortunately, there is limited research exploring how problem construction, and thus problem construction emergence, varies between teams with different levels of experience. Historically, researchers have situated team longevity as an antecedent for decreased levels of team innovation in empirical models. West and Anderson (1996) argued that tenured teams tend to become susceptible to groupthink and more likely to favor “tried-and-true” solutions rather than implement novel ideas. That is, teams with longer longevity tend to develop routinized work patterns and homogenized viewpoints due to their shared past experiences. Thus, it may be expected that tenured teams may omit or limit problem construction interactions during teamwork due to highly aligned problem representations, leading to less creative outcomes. Katz (1982) also demonstrated that research and development project teams with greater longevity become better at communicating with their fellow team members, but also become increasingly isolated from experts outside of their team. Thus, Katz argued that team longevity may become detrimental to team creative performance, as external communication is key for team innovation. Such interactions increase the breadth and depth of information available to the team, potentially leading to the adoption of novel methodologies or prompting the generation of unique solutions. However, the extant research is not conclusive on the effects of longevity on team creative performance. One prominent meta-analysis of 104 independent studies revealed that the mean corrected correlation between team longevity and team innovation was small (p = .02) and insignificant (Hülsheger et al., 2009). Still, both the impact and emergence of problem construction for team creativity regardless of team tenure remains under-studied.

Further, researchers debate where problem construction appears in the sequence of cognitive processes. For example, M. S. Basadur’s (1994) process model suggests that cognition flows linearly, with problem construction often, but not always, occurring at the beginning of the process while other processes co-occur, or operate as sub-processes of larger cognitive operations. Alternatively, Mumford et al. (1991) suggest that individuals may cycle back to early processes, such as problem construction, as a result of difficulties encountered during later processes. In one empirical study, Harvey and Kou (2013) documented a process sequence in teams where members alternated between idea generation and idea evaluation behaviors when the proposed product did not meet evaluation standards. Supporting Harvey and Kou, Boldt (2019) recently demonstrated that teams engaged in a sequence consisting of rapid idea generation followed by idea evaluation, which spurred additional idea generation. However, neither Harvey and Kou nor Boldt included the problem construction process in their framework for classifying interactions; thus, it is still not clear where problem construction fits into the sequence of processes, or how success/failure during problem construction impacts the order of cognitive processes.

Hypotheses

Notably, past research does suggest that certain core processes may naturally emerge more frequently than others. For instance, idea generation is the process most often associated with creative thought. Previous research on individual-level processes found that idea generation represented approximately 37% of think-aloud statements during creative efforts (Khandwalla, 1993). Additionally, research has documented “buildup” during idea generation, where one proposed idea stimulates further ideas that continually increase in number as additional solutions are proposed (Paulus & Yang, 2000). Thus, we anticipate:

H1.1: Teams will display more behaviors reflecting idea generation cognition, than behaviors reflecting either problem construction, or idea evaluation cognition; idea generation will emerge as the most observed category of behaviors.

Further, empirical studies investigating team-level problem construction (e.g., joint representations) tend to apply longitudinal designs that allow teams to formulate shared representations over an interval spanning weeks and multiple project subtasks (Weingart et al., 2010). Thus, it is unclear if ad hoc teams (as in the present study) will rely on their individual problem representations or dedicate time to engaging in team-level problem construction operations. Thus, we anticipate:

H1.2: Teams will display fewer behaviors reflecting problem construction cognition, compared to behaviors reflecting either idea generation, or idea evaluation cognition; problem construction will emerge as the least observed category of behavior.

In addition, we conduct a social sequence analysis to investigate the transition probabilities between coded behaviors. Provided that previous process models of creativity assume a linear pattern, we anticipated teams will utilize a sequential process model during problem-solving. Previous research has also observed teams alternating between idea generation and idea evaluation during creative teamwork (Boldt, 2019; Harvey & Kou, 2013). Thus, we will test the following hypotheses:

H2.1: A sequential analysis will reveal a linear process pattern emerges at a level greater than statistical chance: Teams will tend to start with problem construction, followed by idea generation, and concluding with idea evaluation.

H2.2: A sequential analysis will reveal an alternating process pattern emerges at a level greater than statistical chance: Teams will tend to start with idea generation, followed by idea evaluation, and return back to idea generation.

Method

Developing a Coding Scheme of Behaviors

The application of a new coding scheme was warranted given the specific nature of the present study’s unit of analyses (i.e., utterances reflecting creative cognition). Moreover, a review of the extant literature revealed a lack of existing coding schemes for utterances reflecting creative cognitive processes during team interactions. The present coding scheme uses a top-down etic approach to align verbal interactions between team members with creative core processes based off findings from relevant scholarship. Rather than taking an inductive approach, which may only be relevant to the present sample, the current study takes an etic approach in an attempt to develop a scheme which may be broadly applied to creative problem solving by other scholars using diverse samples as well. Additionally, the etic approach limits the bias of code developers by focusing on behaviors that have already been identified by existing literature. While we acknowledge the etic approach may not be gainful in identifying new behaviors, we argue our proposed coding scheme maintains enhanced validity by limiting our scope to those behaviors known to reflect engagement in certain cognitive processes relevant to creativity. The primary coding scheme is provided in Table 1.

Table 1.

Proposed Primary Coding Scheme.

Problem construction	Idea generation	Idea evaluation
1. Framing the problem through discussion	1. Suggestion of solutions/ideas	1. Discussions on pros and cons of ideas
2. Sharing past experiences relevant to the current problem	2. Discussions about modifications to solutions/ideas	2. Forecasting toward implementation concerns
3. Summarizing the problem		3. Debate/disagreement on which alternative to select
4. Discussions and disagreements about the framing of problems		4. Discussion on evaluation criteria

Problem Construction Codes

Based on our review of relevant literature discussed above, the codes used to indicate engagement in the problem construction class included: (a) framing the problem through discussion (Zahedi & Heaton, 2017), (b) sharing past experiences relevant to the current problem (Holyoak, 1984), (c) summarizing/restating the problem (Redmond et al., 1993), and (d) discussions and disagreements about the framing of problems (Cronin & Weingart, 2007).

Idea Generation Codes

Identifying behaviors reflecting idea generation is simpler than those reflecting problem construction. Since idea suggestion in teams must occur via verbal interactions. The codes used to indicate engagement in idea generation cognition included: (a) suggestion of solutions/ideas and (b) discussion about modifications to solutions/ideas (Paulus & Yang, 2000). The second code, “discussions about modifications to solutions/ideas,” was included to capture “buildup” of ideas that occurs when a suggestion prompts additional idea generation. However, the application of observational methods means it is impossible to determine if solutions are the result of being exposed to other members’ proposals, or the individual’s novel ideation. Thus, the present coding scheme captures buildup conservatively via statements where ideas are added to proposed solutions, or modifications are suggested.

Idea Evaluation Codes

The codes used to indicate engagement with idea evaluation cognition were: (a) discussions on pros and/or cons of ideas to limit the number of potential choices (Mumford et al., 2002), (b) forecasting toward implementation concerns (Mumford et al., 2002), (c) debate/disagreement on which alternative to select (Runco & Chand, 1995), and (d) discussion on evaluation criteria (Licuanan et al., 2007).

Procedure

Participants

To test our hypotheses, we recruited 60 students from a major Midwestern university to participate in a team creative problem-solving task. Conservative estimates of effect sizes in small groups (e.g., Rietzschel et al., 2006) have revealed that 20 teams are adequate to observe constructs of interest. Therefore, the 60 students were distributed into 20 teams consisting of three individuals per team. The students were recruited via the university’s SONA system for undergraduate psychology students. Participants were compensated for their participation in the study with extra credit toward their undergraduate psychology course. The sample consisted of 11 men (18.3%), 48 women (80%), and one subject who chose to not reveal their gender. Thirty-eight (63.3%) participants self-identified as White or Caucasian, eight (13.3%) identified as African American or Black, five (8.3%) identified as Asian American or Pacific Islander, eight (13.3%) identified as Hispanic or Latino, and one identified as Native American. The mean age of the sample was 21.4 years.

Stimulus Materials

We presented teams of participants with an ambiguous problem narrative and asked them to generate a solution to the situation. The drafting of the ambiguous problem followed the recommendations of Amabile (1988), calling for problem narratives that are structured enough to provide framework for subsequent creative thought, but open-ended enough to avoid limitations on creative solutions. The narrative followed the structure and format of problem narratives previously used in creativity research to ensure the drafted problem narrative fulfilled both criteria. The structure included the following elements: no obvious or clear solution, multiple methods of approaching and solving the presented problem, and substantial context to relate solutions to specific aspects of the problem narrative. The problem narrative was also selected to be relevant to college-aged participants to increase their familiarity of the problem domain. The problem focused on a college student, Tom, who has taken on more majors and extracurricular activities than his parents and academic advisor deem necessary.

The problem-solving team sessions took place in university classrooms. Participants were randomly assigned to sessions when multiple teams worked concurrently. After arriving at the pre-assigned location, participants were seated at a table with numbers in front of their seats to identify them in video recordings. Study materials in the room were placed in front of each participants’ seat, including: a copy of the problem narrative, an IRB-approved informed consent form, and an iPad used to record solutions. Team interactions during solution generation were video-recorded for data analysis. Study administrators read from a standardized script to provide instruction to the participants. Administrators first instructed participants to read the problem narrative before working as a team to generate a solution.

Video recordings of team interactions were coded at the utterance-level using the Mangold InterACT qualitative analysis software. Mangold InterACT was selected due to: (a) the capability of assigning nominal codes with timestamps on video data, (b) the functionality to provide each code with a time-range rather than a single point of occurrence, and (c) the capacity to conduct descriptive analytics on coded video data. Two trained raters classified utterances to determine engagement with the core processes of problem construction, idea generation, and idea evaluation. Raters were trained to follow best practices for qualitative research methodology (Tracy, 2013). During coding, raters viewed each team’s video recording in its entirety in order to understand the full context of team interactions. Raters were instructed not to “read-into” utterances or ascribe any implicit meaning to interactions beyond the words spoken as they would appear in a transcription. Raters did not use any keywords to indicate whether a certain code was present, but rather considered the content of utterances to make coding decisions. It is important to note that the start and end points of utterances can be ambiguous in natural language, especially during teamwork where interruptions are common. We do not view this common problem as a significant limitation to our study, and raters were instructed to use their best judgment in deciding when utterances began/ended. See Appendix for a list of example utterance data reflecting each code in the scheme.

Results

Coding Scheme Refinement

After data collection was complete, but prior to video coding, the coding scheme was refined by examining a random subsample of the collected data. The primary list of behaviors for each stage was re-examined to determine whether any behaviors clearly associated with the core processes were observed but were not included in the original coding scheme. Due to the intentionally broad scope of the initial list of codes, the review of the subsample revealed that the initial coding scheme described participants’ behaviors well. However, the scope of one code was expanded to accommodate observed behaviors that were unexpected. The problem construction code “Framing the problem through discussion” was revised to “Framing any aspect of the problem through discussion.” This distinction was made because some participants focused on specific competing pressures within the problem (e.g., “Is Tom really happy?”; “Do his parents know what’s best for him?”) rather than the overall problem scenario (i.e., “What should Tom do?”) during team interactions. Thus, this code was revised to more accurately describe the full range of problem construction behaviors observed. See Table 2 for the secondary coding scheme reflecting this change.

Table 2.

Secondary Coding Scheme.

Problem construction	Idea generation	Idea evaluation
1. Framing any aspect of the problem through discussion	1. Suggestion of solutions/ideas	1. Discussions on pros and cons of ideas
2. Sharing past experiences relevant to the current problem	2. Discussions about modifications to solutions/ideas	2. Forecasting toward implementation concerns
3. Summarizing the problem		3. Debate/disagreement on which alternative to select
4. Discussions and disagreements about the framing of problems		4. Discussion on evaluation criteria

After the secondary coding scheme was finalized, a second random sample consisting of 10% of the data (two team sessions) was created. Two trained raters coded the new subsample of videos for instances of behaviors reflecting the three core processes. The two raters displayed an intra-class correlation coefficient equal to .74, indicating the two raters were reliable judges of observed behavior (Shrout & Fleiss, 1979). The raters proceeded to code the full sample.

Social Sequence Analysis Results

Social sequence analyses determine whether specific sequences (i.e., patterns) emerge at a level greater than statistical random chance across sequential units. Specifically, emergent sequences at the class-level (i.e., patterns reflecting the three core processes) and emergent sequences at the code-level (i.e., patterns of specific codes within the three core processes) were examined. The first-, second-, and third-order transition probabilities were calculated. First-order transitions occurred when one class or code directly followed a previous class/code. Second-order transitions occurred when a class/code was followed by the next-plus-one class/code. Lastly, third-order transitions occurred when a class or code was followed by the next-plus-two class/code. Using the frequencies of transition as a basis, InterACT provided transition probabilities by dividing transition frequencies by the transition sums. However, transition probabilities are confounded by event base rates. To correct the error, InterACT provides a significance value known as the z-statistic (Bakeman & Gottman, 1997). Sequences with z-values above 1.96 and below −1.96, respectively, are considered significant at the p < .05 level.

Frequencies

To investigate which core creative cognitive processes emerged during teamwork, we first requested InterACT to generate a series of frequency tables at the class level. Specifically, we requested the total duration of coded behaviors in seconds (C), total instances of behaviors for each core process (F), the ratio of the instances of class codes to the total number of coded behaviors (F/K), the total duration of each class in seconds (T), and the ratio of the total duration of each class in seconds to the total duration of coded behaviors (T/C). Surprisingly, behaviors consistent with problem construction cognition were observed most frequently (F = 141, F/K = 53.41%, T = 1114.29, T/C = 52.56%), thus failing to provide support for Hypotheses 1.1 and 1.2. The idea generation class emerged as the second-most frequent core process (F = 74, F/K = 28.03%, T = 703.07, T/C = 33.16%). Finally, the idea evaluation class was the least observed core process of creativity (F = 49, F/K = 18.56%, T = 302.87, T/C = 14.28%). In addition to class-level frequencies, code-level frequencies were also requested. See Table 3 for the full list of frequency information at the class level.

Table 3.

Class-Level Frequency Information.

Class	Count (F)	Duration in seconds (T)	Appearance rate (F/K) (%)	Duration rate (T/C) (%)
PC	141	1,114.29	53.41	52.56
IG	74	703.07	28.03	33.16
IE	49	302.87	18.56	14.28

Note. Total duration of coded behaviors in seconds (C) = 2,120.23; Total number of codes (K) = 264. PC = problem construction; IG = idea generation; IE = idea evaluation.

Regarding problem construction, “Framing any aspect of the problem through discussion” appeared most frequently (f = 59; f/k = 41.84%; f/K = 22.35%; t = 469.64; t/C = 22.15%; t/c = 42.15%). Within the problem construction class, “Summarizing the problem” emerged as the second-most displayed behavior (f = 55; f/k = 39.01%; f/K = 20.83%; t = 410.76; t/C = 19.37%; t/c = 36.86%), followed by “Discussions and disagreements about the framing of the problem” (f = 18; f/k = 12.77%; f/K = 6.82%; t = 175.74; t/c = 15.77%; t/C = 8.29%). Lastly, “Sharing past experiences relevant to the current problem” was the least-observed problem construction behavior (f = 9; f/k = 6.38%; f/K = 3.41%; t = 58.16; t/c = 5.22%; t/C = 2.74%). See Table 4 for code-level idea generation and idea evaluation frequencies.

Table 4.

Code-Level Frequency Information.

Class	Code	Code count (f)	Overall code rate (f/K) (%)	Code rate within class (f/k) (%)	Code duration in seconds (t) (%)	Code duration within total duration (t/C) (%)	Code duration within class duration (t/c) (%)
PC	Framing any aspect of the problem through discussion	59	22.35	41.84	469.64	22.15	42.15
PC	Summarizing the problem	55	20.83	39.01	410.76	19.37	36.86
PC	Discussions and disagreements about the framing of problems	18	6.82	12.77	175.74	8.29	15.77
PC	Sharing past experiences relevant to the current problem	9	3.41	6.38	58.16	2.74	5.22
IG	Suggestion of solutions/ideas	62	23.48	83.78	570.00	26.88	81.07
IG	Discussions about modifications to solutions/ideas	12	4.55	16.22	133.07	6.28	18.93
IE	Discussions on pros and cons of ideas	35	13.26	71.43	204.44	9.64	67.50
IE	Debate/disagreement on which alternative to select	9	3.41	18.37	59.62	2.81	19.69
IE	Forecasting toward implementation concerns	3	1.14	6.12	9.04	0.43	2.99
IE	Discussion on evaluation criteria	2	0.75	4.08	29.76	1.40	9.83

Note. Total number of coded behaviors (K) = 264. Total number of PC codes (k) = 141. Total number of IG codes (k) = 74. Total number of IE codes (k) = 49. Total duration of coded behaviors in seconds (C) = 2,120.23 seconds. Total duration of PC behaviors in seconds (c) = 1,114.29. Total duration of IG behaviors in seconds (c) = 703.07. Total duration of IE behaviors in seconds (c) = 302.87. PC = problem construction; IG = idea generation; IE = idea evaluation.

First-Order and Second-Order Sequences

Considering the sequence of cognitive processes, we conducted a first-order social sequence analyses (i.e., lag1) at the class and code levels. Statistically significant transitions indicate that teams were likely (i.e., at a level greater than probabilistic chance) to engage with the latter process after displaying the former. Second-order transitions (i.e., lag2) at both levels were also examined. The second-order analysis provided information concerning the skip-level sequence, or the likelihood of an observed behavior that follows the next-plus-one class/code.

During the first-order social sequence class-level analyses, InterACT revealed various significant transitions. Unexpectedly, the sequence of problem construction followed by idea evaluation emerged at a level statistically greater than chance (f = 12, z = −3.97, p < .05). However, regarding the test of a linear sequence of processes, the sequence of problem construction followed by idea generation was not significant (f = 51, z = 0.59, ns), failing to provide initial lag1 evidence for hypothesis 2.1 predicting that teams engage in a linear process during creative problem. Interestingly, the sequence of problem construction followed by further problem construction did emerge at a level statistically greater than chance (f = 78, z = 2.26, p < .05). All first-order sequences that began with idea generation were not statistically significant (z < 1.96; z > −1.96), indicating that no pattern of processes beginning with idea generation emerged at a level greater than statistical chance. This finding fails to replicate the alternating sequence of processes observed by Harvey and Kou (2013), and thus did not support hypothesis 2.2. The class-level first-order sequences and their respective z-value are provided in Table 5. During the second-order class-level analyses, it was found that all observed second-order transitions did not emerge above and beyond chance (z < 1.96; z > −1.96), failing to provide additional lag2 evidence in support of both hypotheses 2.1 and 2.2. The second-order class-level sequences and their respective z-value are provided in the table below and in Table 6.

Table 5.

Class-Level First-Order Transition Frequencies and their respective Significance Values.

	PC	IG	IE
PC	78 (2.26*)	51 (0.59)	12 (−3.97*)
IG	34 (−1.44)	24 (0.89)	16 (0.85)
IE	19 (−1.32)	8 (−2.02*)	16 (4.53*)

Note. Z-values are provided within parentheses. PC = problem construction; IG = idea generation; IE = idea evaluation.

z < −1.96. z > 1.96, p < .05.

Table 6.

Class-Level Second-Order Transition Frequencies and their respective Significance Values.

	PC	IG	IE
PC	81 (0.78)	41 (0.00)	17 (−1.08)
IG	36 (−1.14)	24 (0.47)	16 (0.96)
IE	23 (0.38)	14 (−0.62)	7 (0.27)

Note. Z-values are provided within parentheses. PC = problem construction; IG = idea generation; IE = idea evaluation.

z < −1.96. z > 1.96. p < .05.

Additional analyses were also conducted investigating code-level sequences to determine if any specific behaviors resulted in statistically significant transitions. The sequence of the problem construction code “Summarizing the problem” and the idea generation code “Proposing a solution” emerged statistically greater than chance f = 23, z = 3.10, p < .05). the problem construction code “Discussions and disagreements about the framing of problems” followed by further “Discussions and disagreements about the framing of problems” emerged as significant (f = 10; z = 6.13, p < .05). The sequence of the problem construction code “Framing the problem through discussion” followed by the problem construction code “Discussions and disagreements about the framing of problems” emerged above and beyond chance (f = 10; z = 3.33, p < .05). Other code-level sequences failed to exceed significance thresholds to establish that the sequence occurred at a level greater than random chance. For a complete list of all first-order sequences and their significance values, see Tables 7 and 8.

Table 7.

Code-Level First Order Transition Frequencies and Their Corresponding Significance Values.

	PC				IG		IE
	Framing	Sharing	Summarizing	Discussions/disagreement	Proposing	Modify	Pros/con	Forecast	Debate	Criteria
PC
Framing	8 (1.84)	1 (−0.43)	10 (0.77)	10 (3.33*)	11 (−0.87)	2 (−1.16)	3 (−1.31)	2 (0.68)	2 (0.31)	0 (−0.47)
Sharing	0 (−1.27)	2 (0.31)	0 (−1.4)	0 (−0.94)	2 (0.25)	1 (0.72)	2 (1.61)	0 (−0.27)	0 (−0.47)	0 (−0.15)
Summarizing	13 (1.84)	0 (−1.58)	11 (−0.04)	4 (−1.6)	23 (3.1*)	4 (−0.59)	2 (−1.79)	1 (0.42)	0 (−1.68)	0 (−0.55)
Discussions/disagreement	4 (−0.97)	0 (−1.01)	5 (−0.95)	10 (6.13*)	7 (−0.48)	1 (−1.01)	1 (−0.85)	0 (−0.61)	0 (−1.07)	0 (−0.35)
IG
Proposing	12 (0.86)	2 (0.03)	10 (−0.64)	2 (−0.62)	10 (0.25)	8 (1.33)	7 (0.17)	1 (0.35)	3 (0.62)	2 (1.75)
Modifying	2 (−0.29)	0 (−0.74)	6 (1.67)	0 (−1.58)	3 (−0.92)	3 (1.47)	1 (−0.75)	0 (−0.45)	2 (1.49)	0 (−0.26)
IE
Pros/cons	5 (−0.15)	3 (1.71)	6 (−0.12)	0 (−1.55)	2 (−1.33)	3 (0.66)	11 (4.51*)	0 (−0.57)	0 (−0.99)	0 (−0.33)
Forecasting	0 (−0.82)	0 (−0.29)	1 (0.5)	0 (−0.61)	2 (1.7)	0 (−0.49)	0 (−0.59)	0 (−0.17)	0 (−0.31)	0 (−0.1)
Debate	1 (−0.56)	0 (−0.5)	2 (0.05)	0 (−1.07)	1 (−0.95)	0 (−0.86)	3 (1.31)	0 (−0.31)	2 (0.44)	0 (−0.18)
Criteria	0 (−0.67)	0 (−0.23)	1 (0.98)	0 (−0.5)	0 (−0.81)	0 (−0.4)	1 (1.85)	0 (−0.14)	0 (−0.25)	0 (−0.08)

Note. Z-values are provided within parentheses. PC = problem construction; IG = idea generation; IE = idea evaluation.

z < −1.96. z > 1.96, p < .05.

Table 8.

Code-Level Second Order Transition Frequencies and Corresponding Significance Values.

	PC				IG		IE
	Framing	Sharing	Summarize	Discussions/disagreements	Proposing	Modifying	Pros/cons	Forecasting	Debate	Criteria
PC
Framing	12 (0.86)	0 (−1.38)	8 (−1.04)	10 (1.79)	12 (−0.09)	2 (−1.22)	4 (0.07)	0 (−0.84)	3 (1.13)	0 (−0.48)
Sharing	0 (−1.18)	3 (1.03)	0 (−1.25)	0 (−0.90)	0 (−1.36)	1 (0.81)	2 (1.76)	0 (−0.26)	0 (−0.45)	0 (−0.15)
Summarizing	10 (−0.68)	0 (−1.56)	15 (1.42)	6 (−0.23)	12 (0.74)	4 (−0.54)	3 (−1.39)	1 (0.44)	2 (0.77)	1 (1.84)
Discussions/disagreement	5 (0.04)	0 (−1.03)	5 (−0.7)	7 (2.47*)	8 (1.16)	2 (−0.35)	1 (−1.52)	0 (−0.62)	0 (−1.09)	0 (−0.36)
IG
Proposing	10 (0.15)	2 (0.05)	13 (0.42)	6 (−0.86)	13 (−0.24)	3 (−0.17)	10 (1.11)	1 (0.37)	1 (−0.93)	0 (−0.57)
Modifying	3 (−0.09)	0 (−0.73)	1 (−1.52)	1 (−0.77)	4 (0.04)	4 (1.49)	3 (0.9)	0 (−0.44)	1 (0.65)	0 (−0.25)
IE
Pros/Cons	6 (0.52)	2 (1.5)	6 (0.26)	1 (−1.36)	8 (−0.6)	2 (−0.09)	4 (0.66)	0 (−0.57)	1 (0.16)	0 (−0.33)
Forecasting	0 (−0.83)	0 (−0.30)	1 (0.57)	0 (−0.64)	0 (−0.96)	1 (1.65)	0 (−0.61)	1 (0.47)	0 (−0.32)	0 (−0.1)
Debate	2 (0.29)	0 (−0.52)	4 (1.83)	0 (−1.11)	1 (−0.87)	1 (0.37)	1 (0.01)	0 (−0.32)	0 (−0.55)	0 (−0.18)
Criteria	0 (−0.68)	1 (1.02)	0 (−0.08)	0 (−0.52)	1 (0.9)	0 (−0.41)	0 (−0.5)	0 (−0.15)	0 (−0.26)	0 (−0.08)

Note. Z-values are provided within parentheses. PC = problem construction; IG = idea generation; IE = idea evaluation.

z < −1.96. z > 1.96. p < .05.

Discussion

The present study furthers the cognitive process perspective of creativity by exploring how problem construction emerges in teamwork relative to other core processes. Quantitative analyses indicated that problem construction activities represented a core focus for teams engaged in creative efforts. Across teams, behaviors reflecting problem construction accounted for more than half (53.41%) of all coded interactions. Thus, the present observations failed to support our hypotheses that idea generation would emerge as the most recorded class of interactions. Rather, interactions reflecting engagement with idea generation operations accounted for only 28.03% of coded behavior, despite the “Suggestion of problem solutions” idea generation code emerging as the most observed behavior. Finally, idea evaluation was the least observed class of behaviors (18.56% of coded data). Notably, we were unable to replicate both a linear pattern of engagement in creative processes and a recurving pattern as described by past research (Harvey & Kou, 2013; Mumford et al., 1991). The finding that problem construction behaviors emerged as the most employed class of creative processes is surprising but provides ample opportunity to explore how problem construction emerged behaviorally at the team level. During problem construction interactions, teams tended to rely primarily on framing aspects of the problem (41.84% of problem construction codes; 22.35% of all coded behavior) and summarizing the problem in their own words (39.01% of problem construction codes). In contrast, sharing past experiences (6.38% of problem construction codes) was observed far less frequently.

Theoretical Implications

Examining the quantitative analysis of code frequencies alongside a qualitative exploration of team interactions reveals several observations: (a) problem construction represents a key activity for ad hoc teams, perhaps due to a lack of experience working together; (b) evaluation operations may play a significant role in problem construction emergence during teamwork; and (c) problem construction behaviors were recorded throughout the problem-solving sessions, co-occurring alongside other processes to serve multiple creative teamwork functions.

Problem construction at the team-level fundamentally represents an information sharing process whereby the team combines and reorganizes individual problem representations into one that is shared and common among members. Indeed, empirical examinations into information sharing, have demonstrated positive outcomes for creative team performance (Dong et al., 2016; Gong et al., 2013). Knowledge management scholars have proposed a diverse intra-team exchange of perspectives on facts or events may connect previously disparate knowledge structures or interpretations embedded in individual members’ schematic memory (Holyoak, 1984; Nahapiet & Ghoshal, 1998). Such connections or recombination of elements represent the creation of new ideas and new knowledge (i.e., generation cognition; Kogut & Zander, 1993). However, past research has shown that rGaps are expected to arise when multiple new concepts or understandings are generated in the context of problem-solving, and high-functioning teams must work to bridge novel and competing concepts without devolving into relational conflict (Weingart et al., 2010). In this study, utterances that served to vet, integrate, and/or select competing problem representations were coded under the problem construction category as “Discussion/disagreements on framing the problem” due to the relevance of addressing underlying rGaps. However, “Discussion/disagreements on framing the problem” utterances observed during team interactions appeared to pursue idea evaluation goals while simultaneously operating within the broader goal of problem construction: arriving at a consensus regarding the team’s shared problem representation. Consider the following observed team interaction (numbers index each utterance, letters identify different members):

A: So it’s basically like between limiting his majors and limiting his extracurricular activities.

B: Maybe if he can find a way to balance out?

A: But, so it says he wishes he had more time to relax and enjoy college but he could not be happier with his current situation. So he says he’s happy but wants a little bit more time. I don’t know if I would cancel the majors (major area of study).

C: Why?

A: Because, at the beginning it says he has a good chance of being offered a full-time position after graduation. I think the company will be more interested in his majors than extracurricular activities.

B: No, but he has an internship that’s why he’s gonna get a job with that company. So I think that job is gonna be locked in either way, because he has an internship with them. So I don’t think they’re interested in his majors.

A: You don’t think his company is interested in his college majors?

B: I mean I’m sure they’re interested in his major, so as long as he doesn’t give up the major that has to do with the job. If he has a major in two different things, they won’t care.

In this session, members of the team have identified an rGap, or misalignment among individual members’ interpretation of certain problem elements (Cronin & Weingart, 2007). That is, problem framing within the team during problem construction led to a brief conflict between members concerning how the team should consider certain elements within the problem narrative. The nature of the conflict appears to be that member A has framed the retention of both major areas of study as a key precedent for securing employment post-graduation. However, member B introduces new consideration to the conversation via the prediction (i.e., forecasting) that Tom’s internship in and of itself will likely result in employment; thus, member A’s framing statement (line 5) may not result in positive outcomes if the team adopts that position before engaging in solution generation. Interactions like the exemplar appear to reflect an evaluation subprocess occurring during shared problem structuring. Member B rejected the framing statement based on their personal forecasting toward expected outcomes (utterance 6) and attempted to reduce the pool of potential representations that could be adopted. Member B maintained that dropping majors would not impact Tom’s employment prospects and accepting the opposing problem frame by member A may result in a sub-optimal solution.

This example of an evaluation operation occurring within a broader creativity process supports the process model described by M. S. Basadur (1994). Basadur situates the creative process as a series of stages progressing from problem-finding (i.e., problem construction) to solution implementation. Within each stage exists a “mini-process” (p. 341) referred to as “ideation-evaluation,” whereby divergent thinking free of judgment (ideation) is followed by convergent thinking (evaluation) applied to refine or select the most ideal option. Under Basadur’s process model, ideation-evaluation during problem finding represents the attempt to first identify a host of problems, or problem elements, to solve and then reduce or refine the host to arrive at a single best selection. Moreover, the team’s disagreement appeared to result in a more nuanced shared problem representation in the example interaction. Member A challenged the notion that a company would not be interested in majors (line 7), and in response member B refined her competing problem frame such that maintaining the major relevant to the job became more important than maintaining unrelated majors (line 8). That is, the rGap observed between the two members’ proposals was bridged via discovering/acknowledging the limitations of member B’s framing and refining the representation accordingly. Thus, frame forecasting, or looking ahead toward the outcomes of accepting competing problem frames, may reflect a unique form of evaluation cognition operating during problem construction whereby problem representations, rather than solutions, are evaluated.

Under a different perspective, the above pattern of interactions could represent an instance where teams “pro-curved” (rather than recurved) to later stages of creative cognition to accomplish the goals of problem construction. Under the framework of Mumford et al.’s (1991) process analytic model of creative capacity, obstacles encountered during creative processes cause individuals to recurve to earlier stages where additional work is done to overcome those challenges. Mumford et al. propose a recurving linear process model in individuals where activity in early stages (e.g., problem construction) must be completed before progressing to later stages (e.g., evaluation). Such transitions have been supported by empirical findings by Harvey and Kou (2013). However, the problem construction process operates automatically, and individuals are not likely to develop competing problem representations. Thus, evaluation subprocesses during individual creative efforts may be unnecessary. It appears that teams can engage in evaluation and problem construction operations simultaneously as a result of obstacles to consensus during problem construction. Given that problem construction is considered the first processes during creative activities, there are no earlier operations to recurve toward during team rGap resolution. The observations of teams engaging with evaluation efforts within the context of problem construction may reflect a proclivity for teams to jump ahead (i.e., “pro-curve”) to advanced processes to overcome disagreements (i.e., obstacles) in early processes. However, the present social sequence analysis revealed insignificant code-level transition probabilities between problem construction and idea evaluation, possibly due to low co-occurrence or transition base rates.

Conversely, the problem construction code “Summarizing the problem” was a prominent problem construction behavior that co-occurred with other processes of creativity. The present study recorded team members frequently restating problem elements alongside their proposed solutions to emphasize which elements were being addressed. One member stated, “I would say since he has three majors, and they want him to go down to one major, but he’s happy where he’s at, maybe he can compromise and do two majors.” Other examples were “He has good grades and he’s succeeding, I don’t think anything needs to change” and “I think he should drop extracurricular activities since he’s being advised by both his parents and advisor.” The “Summarizing the problem” code was defined to include any restatement of problem elements in members’ own words (i.e., not reciting the narrative). Given that the generation of problem restatements has been strongly linked to problem construction activity, it was not anticipated that the restatement of problem elements would be observed during idea generation efforts (Redmond et al., 1993; Reiter-Palmon & Robinson, 2009). Indeed, restating or summarizing problem elements may be a more complex behavior than anticipated. Members may be engaging in problem construction cognition to align their proposed solution with the team’s shared problem representation. Notably, the social sequence analysis partially supports a problem construction to idea generation sequence at the code level: We demonstrate a significant first-order transition probability from summarizing the problem to proposing a solution. However, the transition failed to reach significance thresholds at the class level, suggesting the relationship between summarizing the problem and proposing a solution could represent a statistical artifact rather than an underlying sequence. Summarizing the problem during idea generation could also reflect members’ attempts provide logical validity to the solution being proposed. That is, members may co-present a solution alongside its corresponding problem to emphasize the merit of the solution and draw attention to which problem tensions are successfully being addressed. Thus, summarizing the problem could also represent a form of idea evaluation during teamwork, such that members are providing an advantage to adopting their idea by emphasizing relevant problem elements.

Similarly, it was not anticipated that sharing past experiences relevant to the current problem would represent an uncommon problem construction category (F = 9). The problem narrative was developed with the goal of presenting a familiar scenario to undergraduate participants. Given that past problem-solving experiences represent a key source of individuals’ problem representations, it is likely that past experiences impacted individual members’ problem conceptualization in this study (Holyoak, 1984). However, our results suggest that sharing key experiences may not be a common strategy that ad hoc team members use to arrive at joint representations. Rather, members’ past experiences likely influenced the content of framing statements, whereby members impart their opinions for the team’s consideration. Members that did display sharing behaviors tended to use their experiences to help select one of several competing solutions during idea evaluation operations. For example, one team was comparing two competing solutions: Tom transitioning one of his major areas of study into a minor, versus Tom dropping all majors unrelated to his desired career. One member supported the former solution by sharing her own approach for balancing her course load, stating,

For my example, my psych [sic] is like my fun class that I go to, that’s why I add it as a minor so I have a minor as my fun class, but I also have chemistry that’s not very fun at all, so there is like what I want to do, and then there is my fun class.

Similarly, a different team was discussing whether Tom dropping only one major would be enough, considering his many extracurricular activities. One member supported the proposal to only drop one major by sharing their own situation: “Yea, I’m a double major and I’m fine, I generally have plenty of time.” Teams tended to share past experiences during the evaluation of various solutions proposed. However, the present study observed few instances of sharing interactions, and thus how and why teams discuss their past experiences related to problem-solving remains unclear.

Problem construction also emerged as a means of unifying teams around a common goal. Specifically, “Summarizing the problem” appeared across teams as a means of initiating team interactions and encouraging other members to contribute. “Summarizing the problem” appeared to “break the ice” and initiate the ad hoc teams’ shared goal pursuit by outlining a common set of problem elements that require problem-solving efforts. Thus, problem construction activities may represent one way that novice teams overcome initial barriers to collaboration. Indeed, there is prior evidence that problem construction cognition both increases team satisfaction with the process/outcomes and reduces intra-team conflict (Reiter-Palmon & Murugavel, 2018). Franco (2007) found similar themes in a case-study of project teams within a European construction company. Franco concluded problem structuring techniques that emphasized differences in members’ initial problem appraisals facilitated team collaboration and problem-solving. However, Franco qualitatively analyzed structured interviews rather than performance on a problem-solving task. Thus, it is not clear if summarizing activities are associated with quantifiable changes in team processes. Moreover, empirical research is inconclusive regarding the impact of team-level problem restatements on team creative performance (Leonardi, 2011; Reiter-Palmon, 2011, 2017). While the present observations suggest summarizing behaviors are used by teams to effectively initiate and direct efforts, further empirical research is necessary to determine the implications for team performance.

Practical Implications

Managers, or other leaders of creative teams may find interest in the results of our quantitative analysis of ad hoc team behaviors. Specifically, our results suggest that novel teams spend a disproportionate amount of time engaged in problem construction behaviors compared to other core creative processes. Given that leaders are often focused on project timeline management and resource allocation, leaders should understand that newly formed or ad hoc creative teams may require a greater “initial investment” in the early processes of creative teamwork. That is, leaders who budget additional time for members to engage in problem exploration, consensus building, and problem representation refinement may set the stage for team members to converge on a more robust joint representation. Conversely, novel creative teams that are not provided ample opportunity could experience process loss in the form of misunderstandings, inefficiencies, and greater intra-team conflict (Reiter-Palmon & Murugavel, 2018).

The finding of a disproportionate emphasis on problem construction is also important for leaders of tenured project teams. As project teams may engage with problems within various domains, it is important that knowledge gained from past efforts is captured to prevent a large amount of time spent in problem construction at the commencement of each new initiative. We echo scholars recommending sensemaking activities such as formal debriefs at project conclusion (Allen et al., 2018; Salas et al., 2008). Debriefs (also referred to as after-action reports) provide teams with an opportunity to share their personal experiences working with the team, reflect on strategies or processes that worked, identify problems that reduced performance, and plan for continued success in the future. Implementing formal debriefing sessions has been shown to facilitate greater team effectiveness and may reduce the amount of time teams spend aligning on how to best approach problem-solving (Tannenbaum & Cerasoli, 2013).

Limitations and Future Directions

The present results should be considered in light of certain limitations. First, the present study examined the creative process of ad hoc teams of undergraduate college students. While the effect of team longevity on team creativity is inconclusive (Hülsheger et al., 2009), the generalizability of the present findings to experienced teams of employees may be limited. Indeed, the present effort reflects an investigation into the early stages of team formation without concern for how interactions may mature as the team gains experience. It is possible that different sequences may emerge at different points throughout a team’s lifecycle. Further replication of the present study is necessary to ascertain whether the pattern of frequencies observed across the present 20 teams is indicative of general team processes.

A second limitation is the inherent range restriction that occurs when applying a top-down framework to observed interactions (Tracy, 2013). That is, the present coding scheme only included behaviors previously identified by the extant creativity literature. It may be the case that other important team interactions reflect engagement with core creative processes. Thus, the relative emphasis on any one code or class may vary depending on the identification of other relevant interactions. Future researchers may apply a hybrid grounded theory approach to reveal novel clusters of behaviors fitting within the present core process framework. Finally, it should be noted that the present study descriptively analyzed how teams interacted during creative problem solving. While it appears that certain interactions were a focus of participant teams, no effort was made to link the emphasis on any one class of creative cognition to creative performance. Future studies may further this line of inquiry by comparing different observed sequences of processes or varying emphasis on certain behaviors to determine differential outcomes on the quality and originality of problem solutions. This area of research has large potential to delineate effective versus ineffective team practices regarding creative endeavors.

Future research efforts could also more explicitly explore the underlying motivation behind certain behaviors observed. Structured interviews with team members after problem-solving sessions could reveal insights into why members summarized the problem while providing solutions, or why some chose to share personal experiences during problem construction. These research avenues could provide robust context surrounding the role the problem construction plays from the perspective of team members, as well as provide insights regarding the perception of effective versus ineffective problem construction behaviors. Given that the present results suggest problem construction may be a focal point for newly formed teams, a greater understanding of how problem construction efforts impact both individual members and the team as a whole is warranted.

Footnotes

Appendix

Example Utterance Data.

Class	Code	Examples
PC	Framing	“He’s, like, he’s kinda like the personality type that is always driven toward doing something, staying busy, doing activity.” “Why would you ever need three majors?”
	Sharing	“In my experiences, you need to be committed to [fraternities], because if not, you can get in trouble.” “For my example, my psych [sic] is like my fun class that I go to, that’s why I add it as a minor so I have a minor as my fun class, but I also have chemistry that’s not very fun at all, so there is like what I want to do, and then there is my fun class.”
	Summarizing	“Well, he thinks that he should have more time to relax and enjoy things.” “So, he has three majors, several activities, and an internship, and the advisor wants him to have one major, and his parents want him to drop off some extracurriculars.”
	Discussions/disagreement	“I agree, I think he’s doing it because other people are telling him to, not because he actually wants to” “But that’s not the underlying issues because his advisor thinks that its necessary”
IG	Proposing	“I think he could drop it all down to one, instead of the double major.” “I think Tom should just keep doing him.”
IG	Modifying	“Drop one major and an extra two curriculars, depending on how much he really likes them.” “What if instead he dropped one major and two extracurriculars, depending on if he really likes them.”
IE	Pros/cons	“If he dropped it down to one major, major he could still be involved in some of the other extracurriculars, which would help him to. . .still be more involved around campus.” “Because he’s already been in [the organization] so it will still be on his resume for that number of years, so he doesn’t have to do it all four.” “Yea and in fraternities you do a lot of activities anyways so if he drops [another extracurricular] he’s already doing other things there.”
	Forecasting	“That would probably make him more qualified in the future.” “It would stress him out less, and he could enjoy things more.”
	Debate	“No, I think [only dropping] one would be better.” “Well, he had two good years of college where he did extracurriculars and stuff so I feel that he can chill on the last one or the last half”
	Criteria	“Well, I think it depends on if he wanted to go to grad school or not, because might have a potential opportunity to get a job after graduation.”

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Salvatore Leone

Author Biographies

Salvatore Leone is a PhD student at the University of Nebraska at Omaha and a Senior Data Scientist at a Fortune 50 telecommunications enterprise. His research focuses on creative problem-solving and teamwork, while his professional endeavors have focused on improving the employee experience, employee data analytics, machine learning, and automation.

Payge Japp is a MA/PhD student in the Industrial Organizational Psychology program at the University of Nebraska at Omaha (UNO). She earned a BS in Psychology in 2019 from UNO. Payge’s research interests include individual and team creativity, malevolent creativity, and creative problem-solving.

Roni Reiter-Palmon, PhD is a Distinguished Professor of Industrial/Organizational Psychology at the University of Nebraska at Omaha. Her research focuses on creativity and innovation in the workplace, cognitive processes that influence creative performance of individuals and teams, leading creative individuals, and development of creativity and leadership skills.

References

Allen

Reiter-Palmon

Crowe

Scott

(2018). Debriefs: Teams learning from doing in context. American Psychologist, 73(4), 504–516. https://psycnet.apa.org/doi/10.1037/amp0000246

Amabile

T. M.

(1988). A model of creativity and innovation in organizations. Research in Organizational Behavior, 10, 123–167.

Amabile

T. M.

(1996). Creativity in context: Update to the social psychology of creativity. Westview Press.

Amabile

T. M.

Pratt

M. G.

(2016). The dynamic componential model of creativity and innovation in organizations: Making progress, making meaning. Research in Organizational Behavior, 36, 157–183. https://doi.org/10.1016/j.riob.2016.10.001

Anderson

De Dreu

C. K.

Nijstad

B. A.

(2004). The routinization of innovation research: A constructively critical review of the state-of-the-science. Journal of Organizational Behavior, 25(2), 147–173. https://doi.org/10.1002/job.236

Anderson

Potočnik

Zhou

(2014). Innovation and creativity in organizations: A state-of-the-science review, prospective commentary, and guiding framework. Journal of Management, 40(5), 1297–1333. https://doi.org/10.1177/0149206314527128

Bakeman

Gottman

J. M.

(1997). Observing interaction: An introduction to sequential analysis. Cambridge University Press.

Basadur

M. S.

(1994). Managing the creative process in organizations. In Runco

M. A.

(Ed.), Problem finding, problem solving, and creativity (pp. 237–268). Greenwood Publishing Group.

Basadur

T. M.

Basadur

M. S.

(2013). Creative problem-solving style and individuals’ advice network formation and creative performance. Academy of Management Proceedings, 2013, 17222. https://doi.org/10.5465/ambpp.2013.17227abstract

10.

Boldt

(2019). Artistic creativity beyond divergent thinking: Analyzing sequences in creative subprocesses. Thinking Skills and Creativity, 34, 1–8. https://doi.org/10.1016/j.tsc.2019.100606

11.

Cannon-Bowers

J. A.

Salas

Converse

S. A.

(1993). Shared mental models in expert team decision making. In Castellan

N. J.

(Ed.), Individual and group decision making. (pp. 221–246). Lawrence Erlbaum Associates.

12.

Cronin

M. A.

Weingart

L. R.

(2007). Representational gaps, information processing, and conflict in functionally diverse teams. Academy of Management Review, 32(3), 761–773. https://doi.org/10.5465/amr.2007.25275511

13.

Dong

Bartol

K. M.

Zhang

Z. X.

(2016). Enhancing employee creativity via individual skill development and team knowledge sharing: Influences of dual-focused transformational leadership. Journal of Organizational Behavior, 38(3), 439–458. https://doi.org/10.1002/job.2134

14.

Drazin

Glynn

M. A.

Kazanjian

R. K.

(1999). Multilevel theorizing about creativity in organizations: A sensemaking perspective. Academy of Management Review, 24(2), 286–307. https://doi.org/10.5465/amr.1999.1893937

15.

Franco

L. A.

(2007). Assessing the impact of problem structuring methods in multi-organizational settings: An empirical investigation. Journal of the Operational Research Society, 58(6), 760–768. https://doi.org/10.1057/palgrave.jors.2602191

16.

Gong

Kim

T. Y.

Lee

D. R.

Zhu

(2013). A multilevel model of team goal orientation, information exchange, and creativity. Academy of Management Journal, 56(3), 827–851. https://doi.org/10.5465/amj.2011.0177

17.

Guilford

J. P.

(1967). Creativity: Yesterday, today and tomorrow. The Journal of Creative Behavior, 1(1), 3–14. https://doi.org/10.1002/j.2162-6057.1967.tb00002.x

18.

Harvey

Kou

C. Y.

(2013). Collective engagement in creative tasks: The role of evaluation in the creative process in groups. Administrative Science Quarterly, 58(3), 346–386. https://doi.org/10.1177/0001839213498591

19.

Holyoak

K. J.

(1984). Mental models in problem solving. In Anderson

J. R.

Kosslyn

S. M.

(Eds.), Tutorials in learning and memory: Essays in honor of Gordon Bower (pp. 193–218). W. H. Freeman.

20.

Hülsheger

U. R.

Anderson

Salgado

J. F.

(2009). Team-level predictors of innovation at work: A comprehensive meta-analysis spanning three decades of research. Journal of Applied Psychology, 94(5), 1128–1145. https://doi.org/10.1037/a0015978

21.

Katz

(1982). The effects of group longevity on project communication and performance. Administrative Science Quarterly 27(1), 81–104. https://doi.org/10.2307/2392547

22.

Khandwalla

P. N.

(1993). An exploratory investigation of divergent thinking through protocol analysis. Creativity Research Journal, 6(3), 241–259. https://doi.org/10.1080/10400419309534481

23.

Kogut

Zander

(1993). Knowledge of the firm and the evolutionary theory of the multinational corporation. Journal of International Business Studies, 24(4), 625–645. https://doi.org/10.1057/palgrave.jibs.8490248

24.

Leonardi

P. M.

(2011). Innovation blindness: Culture, frames, and cross-boundary problem construction in the development of new technology concepts. Organization Science, 22(2), 347–369. https://doi.org/10.1287/orsc.1100.0529

25.

Licuanan

B. F.

Dailey

L. A.

Mumford

M. D.

(2007). Idea evaluation: Error in evaluating highly original ideas. The Journal of Creative Behavior, 41(1), 1–27. https://doi.org/10.1002/j.2162-6057.2007.tb01279.x

26.

Mumford

M. D.

Baughman

W. A.

Threlfall

K. V.

Supinski

E. P.

Costanza

D. P.

(1996). Process-based measures of creative problem-solving skills: I. Problem construction. Creativity Research Journal, 9(1), 63–76. https://doi.org/10.1207/s15326934crj0901_6

27.

Mumford

M. D.

Lonergan

D. C.

Scott

(2002). Evaluating creative ideas: Processes, standards, and context. Inquiry: Critical Thinking Across the Disciplines, 22(1), 21–30. https://doi.org/10.5840/inquiryctnews20022213

28.

Mumford

M. D.

Mobley

M. I.

Reiter-Palmon

Uhlman

C. E.

Doares

L. M.

(1991). Process analytic models of creative capacities. Creativity Research Journal, 4(2), 91–122. https://doi.org/10.1080/10400419109534380

29.

Mumford

M. D.

Reiter-Palmon

Redmond

M. R.

(1994). Problem construction and cognition: Applying problem representations in ill-defined domains. In Runco

M. A.

(Eds.), Creativity research, problem finding, problem solving, and creativity (pp. 3–39). Essay, Ablex.

30.

Mumford

M. D.

Robledo

I. C.

Hester

K. S.

(2011). Creativity, innovation, and leadership: Models and findings. SAGE.

31.

Nahapiet

Ghoshal

(1998). Social capital, intellectual capital, and the organizational advantage. Academy of Management Review, 23(2), 242–266. https://doi.org/10.5465/amr.1998.533225

32.

National Research Council. (2012). Education for life and work: Developing transferable knowledge and skills in the 21st century. The National Academies Press. https://doi.org/10.17226/13398.

33.

Osborn

A. F.

(1953). Applied imagination. Scribner. https://doi.org/10.2307/1248180

34.

Paulus

P. B.

Yang

H. C.

(2000). Idea generation in groups: A basis for creativity in organizations. Organizational Behavior and Human Decision Processes, 82(1), 76–87. https://doi.org/10.1006/obhd.2000.2888

35.

Puccio

G. J.

Murdock

M. C.

Mance

(2005). Current developments in creative problem solving for organizations: A focus on thinking skills and styles. Korean Journal of Thinking and Problem Solving, 15(2), 43. https://doi.org/10.1142/9789812770868_0016

36.

Redmond

M. R.

Mumford

M. D.

Teach

(1993). Putting creativity to work: Effects of leader behavior on subordinate creativity. Organizational Behavior and Human Decision Processes, 55(1), 120–151. https://doi.org/10.1006/obhd.1993.1027

37.

Reiter-Palmon

(2011). Introduction to special issue: The psychology of creativity and innovation in the workplace. Psychology of Aesthetics, Creativity, and the Arts, 5(1), 1. https://doi.org/10.1037/a0018586

38.

Reiter-Palmon

(2014). Can we really have an integrative theory of creativity? The case of creative cognition. Creativity. Theories, Research, Applications, 1(2), 256–260. https://doi.org/10.15290/ctra.2014.01.02.13

39.

Reiter-Palmon

(2017). The role of problem construction in creative production. The Journal of Creative Behavior, 51(4), 323–326. https://doi.org/10.1002/jocb.202

40.

Reiter-Palmon

Forthmann

Barbot

(2019). Scoring divergent thinking tests: A review and systematic framework. Psychology of Aesthetics, Creativity, and the Arts, 13(2), 144–152. https://doi.org/10.1037/aca0000227

41.

Reiter-Palmon

Illies

J. J.

(2004). Leadership and creativity: Understanding leadership from a creative problem-solving perspective. The Leadership Quarterly, 15(1), 55–77. https://doi.org/10.1016/j.leaqua.2003.12.005

42.

Reiter-Palmon

Mumford

M. D.

O’Connor Boes

Runco

M. A.

(1997). Problem construction and creativity: The role of ability, cue consistency, and active processing. Creativity Research Journal, 10(1), 9–23. https://doi.org/10.1207/s15326934crj1001_2

43.

Reiter-Palmon

Murugavel

(2018). The effect of problem construction on team processes and creativity. Frontiers in Psychology, 9(2098), 1–8. https://doi.org/10.3389/fpsyg.2018.02098

44.

Reiter-Palmon

Robinson

E. J.

(2009). Problem identification and construction: What do we know, what is the future? Psychology of Aesthetics, Creativity, and the Arts, 3(1), 43. https://doi.org/10.1037/a0014629

45.

Rhodes

(1961). An analysis of creativity. The Phi Delta Kappan, 42(7), 305–310.

46.

Rietzschel

E. F.

Nijstad

B. A.

Stroebe

(2006). Productivity is not enough: A comparison of interactive and nominal brainstorming groups on idea generation and selection. Journal of Experimental Social Psychology, 42(2), 244–251. https://doi.org/10.1016/j.jesp.2005.04.005

47.

Rosing

Bledow

Frese

Baytalskaya

Johnson Lascano

Farr

J. L.

(2018). The temporal pattern of creativity and implementation in teams. Journal of Occupational and Organizational Psychology, 91(4), 798–822. https://doi.org/10.1111/joop.12226

48.

Runco

M. A.

(2003). Idea evaluation, divergent thinking, and creativity. In Runco

M. A.

(Ed.), Critical creative processes (pp. 69–94). Hampton Press.

49.

Runco

M. A.

Chand

(1995). Cognition and creativity. Educational Psychology Review, 7(3), 243–267. https://doi.org/10.1007/bf02213373

50.

Salas

Klein

King

Salisbury

Augenstein

J. S.

Birnbach

D. J.

Robinson

D. W.

Upshaw

(2008). Debriefing medical teams: 12 evidence-based best practices and tips. The Joint Commission Journal on Quality and Patient Safety, 34(9), 518–527. https://doi.org/10.1016/S1553-7250(08)34066-5

51.

Shalley

C. E.

Gilson

L. L.

(2004). What leaders need to know: A review of social and contextual factors that can foster or hinder creativity. The Leadership Quarterly, 15(1), 33–53. https://doi.org/10.1016/j.leaqua.2003.12.004

52.

Shrout

P. E.

Fleiss

J. L.

(1979). Intraclass correlations: Uses in assessing rater reliability. Psychological Bulletin, 86(2), 420–428. https://doi.org/10.1037/0033-2909.86.2.420

53.

Simon

H. A.

Newell

(1971). Human problem solving: The state of the theory in 1970. American Psychologist, 26(2), 145–159. https://doi.org/10.1037/h0030806

54.

Tannenbaum

S. I.

Cerasoli

C. P.

(2013). Do team and individual debriefs enhance performance? A meta-analysis. Human Factors, 55(1), 231–245. https://doi.org/10.1177/0018720812448394

55.

Tracy

S. J.

(2013). Qualitative research methods: Collecting evidence, crafting analysis, communicating impact. Wiley-Blackwell.

56.

Unsworth

(2001). Unpacking creativity. Academy of Management Review, 26(2), 289–297. https://doi.org/10.5465/amr.2001.4378025

57.

Ward

T. B.

Smith

S. M.

Finke

R. A.

(1999). Creative cognition. In Sternberg

R. J.

(Ed.), Handbook of creativity (pp. 189–213). Cambridge Press. https://doi.org/10.1017/cbo9780511807916.012

58.

Weingart

L. R.

Todorova

Cronin

M. A.

(2008, August). Representational gaps, team integration and team creativity [Paper presentation). 2008 academy of management meetings, Anaheim, CA.

59.

Weingart

L. R.

Todorova

Cronin

M. A.

(2010). Task conflict, problem-solving, and yielding: Effects on cognition and performance in functionally diverse innovation teams. Negotiation and Conflict Management Research, 3(4), 312–337. https://doi.org/10.1111/j.1750-4716.2010.00063.x

60.

West

M. A.

(2002). Sparkling fountains or stagnant ponds: An integrative model of creativity and innovation implementation in work groups. Applied Psychology, 51(3), 355–387. https://doi.org/10.1111/1464-0597.00951

61.

West

M. A.

Anderson

N. R.

(1996). Innovation in top management teams. Journal of Applied psychology, 81(6), 680–693. https://psycnet.apa.org/doi/10.1037/0021-9010.81.6.680

62.

West

M. A.

Hirst

Richter

Shipton

(2004). Twelve steps to heaven: Successfully managing change through developing innovative teams. European Journal of Work and Organizational Psychology, 13(2), 269–299. https://doi.org/10.1080/13594320444000092

63.

Zahedi

Heaton

(2017). A model of framing in design teams. Design and Technology Education, 22(2), 8–25.

64.

Zhou

Wang

X. M.

Bavato

Tasselli

(2019). Understanding the receiving side of creativity: A multidisciplinary review and implications for management research. Journal of Management, 45(6), 2570–2595. https://doi.org/10.1177/0149206319827088