Time to be Creative? Self-Regulation of Time in Creative Professions. Flora Beeftink

Transcription

1 Time to be Creative? Self-Regulation of Time in Creative Professions Flora Beeftink

2 A catalogue record is available from the Eindhoven University of Technology Library ISBN: NUR 771 Keywords: Creativity / Time / Self-regulation / Time management Printed by Universiteitsdrukkerij Technische Universiteit Eindhoven Cover design: Anne Beeftink Cover photo: Anne Beeftink 2008, Flora Beeftink, Eindhoven All rights reserved. No part of this publication may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without permission in writing from the author.

3 Time to be Creative? Self-Regulation of Time in Creative Professions PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de Rector Magnificus, prof.dr.ir. C.J. van Duijn, voor een commissie aangewezen door het College voor Promoties in het openbaar te verdedigen op woensdag 15 oktober 2008 om uur door Flora Beeftink geboren te Pannerden

4 Dit proefschrift is goedgekeurd door de promotoren: prof.dr. C.G. Rutte en prof.dr.ir. J.W.M. Bertrand Copromotor: dr. W. van Eerde

5 Acknowledgements And now this PhD project is completed. It took some effort, but I enjoyed working on it very much. Doing research about time management I was very aware of the limited time I had to complete this dissertation and felt slightly threatened by the milestones and deadlines of my project. This urged me to work hard at times, which prevented extreme time pressure near the final deadline. Following the rules of time management as one could say, and indeed I experienced the feeling of being in control of time and of my project, which decreased the amount of experienced work stress considerably. And with great peace of mind I spend almost all my Saturdays (in addition to my already stretched work week) working on my dissertation for the past few months... This sounds like I have been working on this project all by my self, but that was definitely not the case. Although this dissertation carries my name, the manuscripts on which the chapters of this dissertation are based are written in co-authorship with Wendelien van Eerde, Christel Rutte, and Will Bertrand. Not only were they my co-authors, they were my teachers and coaches that transformed me slowly but surely from a engineer with a passion for work psychology into a work psychologist with a background in engineering. They gave me enough guidance to be able to get me started and enough freedom to make my own choices by the time I was up to it. In addition, I thank Wendelien for her indefinite inspiration and for being there at all times. My appreciation also goes out to Christel for recognizing my potential, then, but also now. Furthermore, the studies in this dissertation highly depended on the willingness of people to participate. Therefore, I would like to thank everyone that took the time and effort v

6 to be part of my investigations. I am particularly grateful to the 25 architects that participated in my interview study for being the 'voice of practice' and explaining to me in detail what it is like to be a creative professional under deadline pressure. Although this qualitative study absorbed an incredible amount of my scarce time, I cannot imagine me having done my other studies without knowing in the back of my head what the most urgent issues were that these architects were confronted with on a day-to-day basis. Also, I would like to show my gratitude to the participants of my diary study in particular, for their continued efforts and commitment that were necessary for this study. Furthermore, I thank the organizations that allowed me to approach all these people: the BNA (Organization of Dutch Architects), the Stan Ackermans Institute (School for Technological Design), KreaNET (Network of Creativity Enhancers), Festina Lente Creative Productions (and their Seaday project), and ARCAM (Amsterdam Centre for Architecture). A big help were also my colleagues from the HPM group, first and foremost my roommate Eric van der Geer-Rutten-Rijswijk (see also the 'dearest friends' section below). Second, my colleagues from the 'U-gang' (Ad Kleingeld, Anniek van Bemmelen, Josette Gevers, Tanja Bipp, Daphne Dekker, Marieke van der Tooren, and Marieke Habraken) for providing me with ample time for incubation during the afternoon tea sessions, which was a boost both for my positive affect and daily creativity. Anniek also deserves my gratitude for checking my English and being a very dedicated and patient teacher. Since feeling well and happy is one of the best predictors of job performance, I owe my friends, family, and pets. I would like to thank Sander in particular, for allowing me to be myself at all times, and because he was so very much not interested in my work that it was very easy to put things in perspective when I tended to be stressful. Also a big thank you to Gitta, for getting me through the worst episodes of my life, for being a great sport when it comes to sharing the most fun parts of life, and for being even slightly more confident in me then myself. Eric, thank you is not enough: I miss you every day! Lasts me to thank Gemma Zonneveld, my mum, for being so incredibly proud of me that it is almost impossible not to be proud of myself as well, and Wouter en Anne, my brother and sister, because them being family means more to me than all friendships together. Flora Beeftink, July 2008 vi

7 Contents Acknowledgements v Chapter 1 1 Introduction Chapter 2 9 The effect of interruptions and breaks on insight and impasses: Do you need a break right now? Chapter 3 23 Working on something else for a while: Uncovering patterns of task engagement and disengagement in creative professions Chapter 4 51 Managing one's creativity? A diary study among R&D engineers on the effect of time management on positive affect and daily creativity Chapter 5 77 Being successful in a creative profession: The role of having an innovative style, self-regulation of design work, and self-efficacy Chapter General discussion vii

8 References 117 Summary 133 Samenvatting 137 About the author 141 viii

9 Chapter 1 Introduction In general, a creative profession requires engaging in cognitive creative activities, taking time to let ideas sink in (i.e., incubation time), and making sure that project milestones and deadlines are met, while monitoring the timeliness, efficiency, and quality of work tasks. Although managing one's time in creative professions seems therefore to be highly relevant, up to now, little research has been conducted to investigate its added value. This dissertation presents five studies (one experiment and four field studies) in which we address the self-regulation of time (i.e., time management) in creative work. Particular emphasis is put on incubation time, task discretion, pacing patterns, deadlines, and self-regulation behaviors such as prioritizing, planning, and monitoring creative work activities. Knowledge based work, and creative work in particular, is becoming increasingly important in today's economies (DeFlillippi, Grabher, & Jones, 2007). Creative work requires specific research attention, because it differs from other work: It concerns the development of ideas that are novel and potentially useful (Amabile, 1996). Past research on creativity at work focused predominantly on the influence of personal characteristics and the work environment on creative performance (Shalley, Zhou, & Oldham, 2004). For example, it has been found that it is helpful to creativity to have an innovative cognitive style (i.e., an 1

10 Chapter 1 orientation towards creative problem solving) (e.g., Tierney, Farmer, & Graen, 1999), or a high level of self-efficacy (i.e., the extent to which individuals believe that they can successfully execute the behaviors required to produce an outcome (Bandura, 1977)). However, many questions still remain to be answered. In this dissertation we focus on a topic that has received little research attention so far: the role of time in creative work (Drazin, Glynn, & Kazanjian, 1999), and, specifically, the role of self-regulation behaviors that are focused at managing one's time at work, in more popular terms: time management. We investigate how creative professionals manage their time at work, whether they do it differently from people in other professions, and whether it is beneficial when they make an effort to manage it well. We built on Control Theory of Self-Regulation (Carver & Scheier, 1982). This theory suggests that task attainment requires cycles of feedback loops, in which 'feedback' refers to information about task progress. The purpose of going through these feedback loops is to manage the process of task execution and adjust the course of action if necessary. (For a more detailed description of this theory we refer to Chapters 4 and 5 of this dissertation.) Time management implies going through these self-regulation feedback loops, but, in addition, puts a particular emphasis on time use. Thus, in the term 'time management', it is not the time that is managed, but rather the process of task attainment with a focus on time use (Claessens, Van Eerde, Rutte, & Roe, 2007). In the following, we will first define creative work and elaborate on the role of time in creative work. Then, we will describe the research questions that are addressed in this dissertation. We will end with providing the outline of the dissertation in which we briefly illustrate the content of the different chapters and studies. Creative Work In this dissertation we use Amabile's (1996) well known definition of creativity: Creativity concerns the development of ideas that are novel and potentially useful. Our definition of creative work is therefore: Work in which it is required to develop ideas that are novel and potentially useful. In the work of a research and development engineer, for example, designing a new (novel) product that must meet functional specifications (usefulness) requires considerable creative efforts. Moreover, we extend our definition of creative work 2

11 Introduction to work in which novel and potentially useful ideas are required to solve ill-defined problems. Ill-defined problems are problems for which a clear path to solution is unknown (Sternberg & Davidson, 1999) and for which creative problem solving is necessary to find a solution (Cropley, 1999). An R&D engineer, for example, does not know beforehand what specific activities are needed to develop a certain new product. Similar to the definition of Amabile (1996), our definition does not distinguish between incremental and more radical creative contributions, because, in our view, creativity is needed for both. Our definition excludes work with only opportunities for voluntary creativity. In other words, we exclude work in which creativity is optional and not part of one's main responsibilities (Unsworth, 2001), and only include work in which creativity is required. Thus, this dissertation concerns the work of creative professionals. Specifically, this dissertation focuses at the role of time in their work. Time in Creative Work Time is latently present in our daily lives, and therefore, also at work: It passes while we engage in work activities. Thus, it also passes while engaging in creative work activities. Although in this perspective the role of time is passive, time can also have a more dominant role in creative work. First, time can be part of the creative process. Incubation time literature suggests that some time away from a creative task (i.e., incubation time) aids creative problem solving (Wallas, 1926). After a period of time during which an individual has been doing something else, suddenly and unexpectedly an insight (i.e., a helpful idea) can come to mind ('eureka!'). Thus, temporary disengagement from a creative task has purpose, and in this perspective, (incubation) time can be considered part of the creative process. Second, time can provide structure to the work through the use of project milestones and deadlines. Orlikowski and Yates (2002), for example, referred to project milestones and deadlines as being temporal structures that "are created and used by people to give rhythm and form to their everyday practices" (p. 685). In other words, individuals use them to make decisions with regard to the allocation of their time to their work activities. 3

12 Chapter 1 A third perspective views time as a scarce resource (Ancona, Okhuysen, & Perlow, 2001, p. 515). This perspective focuses on the economic side of time at work, in which time represents a valuable commodity. This perspective concerns aspects such as timeliness, efficiency, and quality. Timeliness refers to norms for when a certain task should be completed, and puts restrictions on the amount of available time. Striving for efficiency means trying to do as much as possible in as little time as possible. Being short of time may result in quality-time trade-offs. Thus, in concepts as timeliness, efficiency, and quality, time is scarce and valuable and should be used as effectively as possible. When considering the role of time in creative work, these three perspectives are all relevant: Creative professionals use their time at work to execute creative tasks, or for incubation, while monitoring the quality, efficiency, and timeliness of task completion, within time structures of project milestones and deadlines. Taking this into account, one can realize that it may require considerable efforts to manage time at work in creative professions. However, up to now, little empirical evidence has been provided to verify this. Self-Regulation of Time in Creative Work Research Questions This dissertation addresses a number of research questions concerning time management in creative work. The first concerns the role of the 'self' in the self-regulation of time in creative work. Specifically, whether it makes a difference if individuals can regulate when to work on which tasks themselves, compared to others deciding that for them. In other words, whether the discretion to choose when to work on which tasks contributes to creative performance. This also includes the discretion to choose when to stop and work on something else for a while (i.e., taking incubation time). Alternatively, tasks could be assigned in an optimal way so that incubation may occur, but without the individual's input as to when this is needed (see Chapter 2 of this dissertation). A first research question is therefore: Research question 1: Are individuals more creative if they have personal discretion over how to use their time? 4

13 Introduction The next two research questions concern the role of pacing in creative work. In this denotation pacing refers to the allocation of efforts for the execution of tasks over time (Gevers, Rutte, & Van Eerde, 2006), in other words, to the self-regulation of work efforts over time. Researchers have distinguished different pacing patterns that describe how work efforts can be distributed over time within tasks or projects (Gevers, Mohammed, Baytalskaya, & Beeftink, 2008). For example, individuals may (a) work on tasks with a moderate, constant level of engagement, without periods of very high levels of task engagement, or periods of disengagement; (b) include an intermediate period of disengagement, which may be used as incubation time; (c) do most of the work right after de start of a task or project; or (d) do most of the work at a looming deadline. (More details about these pacing patterns can be found in Chapter 3.) In this dissertation we address two research questions with regard to the pacing of creative work: Research question 2: Do creative professionals distribute their work efforts differently over time compared to others working in less creative professions? Research question 3: How do creative professionals pace their projects over time exactly? The fourth and fifth research questions concern the usefulness of time management in creative work. Although, in general, the self-regulation feedback loop (and thus also time management) seems to be highly useful in the process of task completion, its use in creative work is not self-evident. The difficulty lies in the specific behaviors of the self-regulation feedback loop: (a) deciding what to do, (b) planning for action, (c) task execution, and (d) monitoring task progress (Carver & Scheier, 1998). Since creative work consists of novel, ill-defined tasks (Cropley, 1999), it comprises a high level of task uncertainty. For some behaviors in this feedback loop, and for planning in particular, ill-defined tasks present a dilemma (Ormerod, 2005): Task uncertainty makes planning ahead difficult. On the other hand, the self-regulation feedback loop may be particularly useful in creative work because of this high level of task uncertainty: Adjusting and making new plans regularly during the course of a creative task or project may be required. Therefore, our fourth and fifth research 5

14 Chapter 1 questions regard the benefit of time management and the corresponding behaviors of the self-regulation feedback loop to creative work. Research question 4: Is time management beneficial to creativity? Research question 5: To what extent are self-regulation behaviors (setting priorities, planning work activities, and monitoring time and task progress) related to the success of creative professionals? Outline of the Dissertation The chapters of this dissertation present studies in which different research methods were applied: An experiment, two surveys, an interview study, and a diary study. Table 1.1 provides an overview of the chapters, the studies, their methods, and the topics addressed. Although the chapters are certainly related and may partly overlap, each chapter provides a unique contribution to the literature, and can be read independently of the others. Chapter 2 describes an experiment among 109 students of industrial engineering and management science. In this experiment it was investigated whether task discretion is helpful to creativity. Specifically, we tested whether the discretion to choose when to work on something else for a while (incubation time) is beneficial to creativity, over and above the effect of incubation time when others decide for you when to switch tasks. We investigated whether this would be beneficial to the number of insights conceived to solve problems in the tasks at hand and to prevent impasses (i.e., feelings of being 'stuck'). Chapter 3 contains two field studies that both focused on pacing of creative work. The first is a survey study among 88 individuals working in various professions. The purpose of this study was to investigate whether creative professionals distribute their work efforts differently over time than other professionals, and, if so, whether they have a preference for specific pacing patterns. The second study is an in-depth interview study among 25 architects. In this study we explored pacing patterns in more detail, taking into account some of the complexities of working on creative projects, such as the influence of project milestones and deadlines and of working on multiple projects simultaneously. Chapter 4 presents a diary study among 68 R&D engineers. This study investigated the effect of time management in creative work, over and above the impact of 6

15 Introduction Table 1.1. Overview of chapters, research methods, and topics in the dissertation Topics Chapter Research method Self-regulation of work time Creativity Other topics 2. Experiment Task discretion Incubation time Insight Impasse 3. Survey Pacing patterns Creative professions Interviews Pacing patterns Project milestones and deadlines Multi-tasking Deadline postponement Quality-time trade-offs Creative projects 4. Diary study Time management Daily creativity Fluctuations in daily creativity Innovative cognitive style Daily affect 5. Survey Self-regulation: - Prioritizing - Planning - Monitoring Innovative cognitive style Being a successful designer Self-efficacy Being a successful businessperson 7

16 Chapter 1 having an innovative cognitive style. Specifically, it investigated whether individuals who have a tendency to use time management are more creative during their work days. Furthermore, it was examined whether potential benefits of time management can be explained through its effect on positive affect. So, whether time management allows individuals to be more creative because they feel better. Finally, this study also examined whether time management was related to a more stable creative performance over several workdays. Chapter 5 describes a survey study among 339 architects. The aim of this study was to establish the relevance of three self-regulation behaviors (setting priorities, planning work activities, and monitoring time and task progress) to design work. Specifically, we assessed to what extent self-regulation and innovative cognitive style were related to being a successful designer and businessperson, and the mediating role of self-efficacy in these relationships. The last chapter, Chapter 6, provides an overview and integration of the results of each of the studies. We conclude with suggestions for future research and recommendations for practice. 8

17 Chapter 2 The Effect of Interruptions and Breaks on Insight and Impasses: Do You Need a Break Right Now? * Some time away from a problem, or incubation time, is found to be beneficial to creative problem solving. But are interruptions as equally helpful as breaks? An experiment was conducted to gain more insight into the differences between imposed and self-initiated breaks, and their effects on creativity, specifically on impasses and insights. There were three experimental conditions, (a) a continuous condition, in which participants were not allowed to switch back and forth between tasks, (b) an interruption condition, in which participants had to switch tasks at a predetermined moment, and (c) a break condition, in which participants could switch tasks at their own discretion. Results showed that taking breaks at moments chosen at one's own discretion led to solving more insight problems and reaching fewer impasses than at moments that were chosen by others. Furthermore, compared to working continuously, interruptions led to fewer impasses, but not to solving more insight problems. * This chapter is based on: Beeftink, Van Eerde, & Rutte (in press). The effect of interruptions and breaks on insight and impasses: Do you need a break right now? Creativity Research Journal. 9

18 Chapter 2 At work, most knowledge workers can switch between different activities or take breaks at their own discretion, and they are also regularly subjected to interruptions initiated by others (Czerwinski, Horvitz, & Wilhite, 2004; Nandhakumar & Jones, 2001). Jett and George (2003) theorized that self-initiated breaks from work can have positive effects on creativity, whereas interruptions initiated by others may not bring about these positive effects. This proposition has not yet been tested. The current chapter describes an experiment to gain more insight into the differences between imposed and self-initiated breaks, and their effects on creativity, specifically on insights (i.e., the distinctive and apparently sudden realization of a strategy that aids in solving a problem (Sternberg & Davidson, 1999, p. 58)), and on impasses (i.e., the feeling of being "stuck" and not being able to solve a problem (Fleck & Weisberg, 2004, p. 998)). Insight Problem Solving Having a sudden insight, or "eureka moment", is a well known phenomenon in everyday life and in the literature (Sternberg & Davidson, 1999). Individuals suddenly and unexpectedly get a good idea that brings them a great step further in solving a problem (Wallas, 1926). To be more specific: Individuals may suddenly become aware of a new strategy that may solve the problem at hand (Sternberg & Davidson, 1999). An often cited creative problem solving model that includes sudden insights is that of Wallas (1926). It consists of four stages: (a) The preparation stage: gain in-depth knowledge about a problem, (b) the incubation stage: take some time away from the problem, (c) the illumination stage: suddenly and unexpectedly an insight comes to mind, and (d) the verification stage: verify the applicability of the idea. The underlying hypothesis of this model is that incubation time aids the problem solving process. In a review article about experimental research on the effect of incubation time, Dodds, Ward, and Smith (in press) conclude that this incubation effect exists: roughly 75% of the experiments show positive effects of incubation time. In addition to Wallas' model, Smith and Blankenship (1991) and Smith (1995) suggested that incubation time is only valuable in specific situations: "Problems that are solved immediately require no incubation [time]" (Smith & Blankenship, 1991, p. 63). They argued that individuals try to tackle problems with a certain problem solving strategy, 10

19 Do you need a break right now? and that it is very difficult for them to approach a problem with a different strategy when the first one does not seem to be appropriate. They refer to this block to successful problem solving as a cognitive state of fixation. In four experiments, Smith and Blankenship showed that incubation time was only helpful in situations in which individuals needed to overcome this state of fixation. The insight problem solving model shown in Figure 2.1 combines Wallas' model (1926) with the ideas of Smith and Blankenship (1991) and Smith (1995). First, individuals start working on a problem, which results in some problem progress. This is similar to what Wallas called the preparation stage. If the problem is tackled with the appropriate problem solving strategy, it can be solved at once. If not, individuals reach a cognitive state of fixation. They can keep trying to solve the problem, but the chance that they will find a suitable strategy to tackle the problem is very small and, most likely, the problem will remain unsolved. Another option is to take some incubation time. A sudden insight might occur that reveals a new strategy which may solve the problem. Start working on problem Problem progress Fixation Continue to work on problem Incubation Problem solved Problem remains Insight Figure 2.1. Insight Problem Solving Model The Effect of Interruptions and Breaks on Insights Both interruptions and breaks give individuals the opportunity to take some time away from a problem, in other words, to have some incubation time. As previous research suggests (Dodds et al., in press), either form of incubation time should improve problem solving performance. Therefore, the first hypothesis of this chapter states that individuals who are 11

20 Chapter 2 interrupted benefit from the incubation effect, and solve more problems than individuals who continuously work on these problems. Similarly, a second hypothesis is that individuals who can take breaks at their own discretion solve more problems than individuals who continuously work on these problems. Hypothesis 1: Individuals who are interrupted solve more insight problems than those who work continuously on them. Hypothesis 2: Individuals who can take a break at their own discretion solve more insight problems than those who work continuously on them. Not empirically tested before is the question whether there is a difference between the effects of interruptions and breaks on creative performance, specifically, whether breaks should be preferred over interruptions. Interruptions are generally seen as negative and disruptive whereas breaks are not (Jett & George, 2003). In line with this belief, time management training advises to avoid interruptions as much as possible (Green & Skinner, 2005; Macan, 1994). Some research confirms the proposed negative effects on task performance. For example, in a field study in a commercial telecommunications office, Eyrolle and Cellier (2000) found that operators needed more time to complete tasks that were interrupted compared to tasks that could be carried out without interruptions. They also conducted an experiment that revealed that interruptions also have a negative effect on the number of errors individuals make. Recently, Bailey and Konstan (2006) have confirmed both of these findings in a similar experiment. They found that interrupted individuals needed more time to complete their tasks compared to uninterrupted individuals, and that they made twice as much errors. But not all research finds negative effects of interruptions on task performance. In a field study among 46 bank managers, König, Kleinmann, and Höhmann (2004) found no relation between daily performance and interruptions. Zijlstra, Roe, Leonora, and Krediet (1999) conducted two experiments in which they found that interruptions caused individuals to work faster while maintaining the same level of quality. In conclusion: Former research suggests that, opposed to the positive effect of incubation, interruptions might also cause negative effects on task performance. Self- 12

21 Do you need a break right now? initiated breaks, on the other hand, may not bring about such negative effects because individuals can choose a convenient point in time to take them. Therefore, a third hypothesis proposes that individuals who can take a break at their own discretion outperform individuals who are interrupted. Hypothesis 3: Individuals who can take a break at their own discretion solve more insight problems than those who are interrupted. The Feeling of Being Stuck Conceptually new to this field of research is the distinction between the cognitive state of fixation and the psychological state of impasse, and the use of impasses as a dependent variable in creativity research. An impasse is defined as the feeling of being "stuck" (Fleck & Weisberg, 2004), and can be differentiated from the cognitive state of fixation in two ways: First, the psychological state of impasse is affective and incorporates negative emotions such as confusion and frustration (Fleck & Weisberg, 2004), whereas the cognitive state of fixation is free of such emotions because it occurs unconsciously. Second, a lagged effect might exist between the moment of fixation and the perception of reaching an impasse: After fixation, individuals might not realize immediately that it will be difficult to solve the problem. Instead, it might take some time before they feel "stuck". In an experiment by Segal (2004), for example, participants worked up to twenty minutes on the presented insight problem before they stated that they had reached an impasse. To be more precise, we suggest that, after individuals get into a state of fixation, negative emotions gradually increase over time. When confusion and frustration reach a certain level, individuals decide that they do not want to work on the problem anymore and give up. At this point in time they reach the psychological state of impasse. In previous research on the incubation effect, both the cognitive state of fixation and the psychological state of impasse were predominantly treated as a necessary precondition for the incubation effect to occur (Dodds et al., in press). In this chapter the two are conceptually separated: Fixation is treated as a necessary precondition, but impasse as an affective outcome of the creative process, worthwhile to be studied in more detail. 13

22 Chapter 2 The Effect of Interruptions and Breaks on Impasses Both interruptions and breaks might be of influence on the process of reaching an impasse. After individuals reach the cognitive state of fixation, interruptions or breaks may bring the process of building up negative emotions to a temporary halt. This halt might prevent these emotions from reaching the critical level of impasse. Assuming that individuals reach a state of fixation if they are not able to solve an insight problem right away (Smith, 1995; Smith & Blankenship, 1991), Hypothesis 4 states that individuals who are interrupted report impasses less often than individuals who continuously work on problems. Similarly, a fifth hypothesis is that individuals who can take a break at their own discretion report impasses less often than individuals working continuously. Hypothesis 4: Individuals who are interrupted report fewer impasses than those who work continuously on an insight problem. Hypothesis 5: Individuals who can take a break at their own discretion report fewer impasses than those who work continuously on an insight problem. The last hypothesis focuses on the question whether there is a difference between the effects of interruptions and breaks on impasses reached, and which is to be preferred in order to prevent an impasse. Here the emphasis lays on the affective influences of interruptions and breaks, and whether these add to the negative feelings leading to the impasse. Breaks are not expected to be a source of negative emotions, but interruptions may be. For example, in the experiment of Bailey and Konstan (2006), interrupted participants felt more annoyed and anxious than uninterrupted participants. Similarly, Zijlstra et al. (1999) found that interruptions had a negative effect on their participants' emotions. On the other hand, interruptions may also prevent feelings of boredom when individuals are working on simple tasks (Fisher, 1998). To summarize: It is expected that, although interruptions may reduce feelings of frustration, they may also be a source of negative emotions. Therefore, the sixth hypothesis states that individuals who can take a break at their own discretion feel fewer negative emotions than individuals who are interrupted, and as a consequence they reach the state of impasse less often. 14

23 Do you need a break right now? Hypothesis 6: Individuals who can take a break at their own discretion report fewer impasses than those who are interrupted. Method Participants One hundred and nine Dutch undergraduate students (12 female) participated in the experiment. The average age of the participants was 20.8, the youngest was 19, and the oldest 27 years old. Their major was in Industrial Engineering and Management Science and they were highly educated in math and statistics. Task The creative task for each of the students was to solve three insight problems. Different types of insight problems exist: verbal, mathematical, and spatial problems (Dow & Mayer, 2004). We decided not to use mathematical or spatial insight problems, such as the well known nine dot problem (Runco & Sakamoto, 1999; Sternberg & Davidson, 1999), because of the risk that these would be solved mathematically instead of through insight (for comments on hybrid insight problems see: Weisberg, 1995). This risk was particularly large in our student sample. Therefore, verbal insight problems were used: Three relatively difficult cryptic crossword clues 2.1. An example is: "not seeing the window covering (5 characters)" is "blind". The three word puzzles were selected and pre-tested during a pilot study among 36 undergraduate students with the same educational background. Pre-testing showed answer patterns typical for insight problems: Solutions were generally found right away, in the first minute (55%), or when participants came back to a word puzzle after working on other word puzzles (21%). Design The design consisted of three conditions. Participants in all conditions had a total of three times six minutes to work on the word puzzles. Participants in the continuous condition 2.1 In cryptic crosswords, each individual clue is a word puzzle in and of itself. In this case the clues are the insight problems, not to be mistaken with clues or hints that are sometimes used as intervention to aid insight problem solving (e.g., Chronicle, Ormerod, & MacGregor, 2001). 15

24 Chapter 2 were not allowed to switch back and forth between the word puzzles, which meant that they worked six minutes on puzzle A, then six minutes on puzzle B, and then six minutes on puzzle C (see Figure 2.2). In the interrupted condition participants had to switch to the next puzzle after three minutes, and came back to it after they had worked on the other two puzzles for three minutes each. The sequential ordering of the puzzles was A, B, C, A, B, C (see also Figure 2.2). Note that both the participants in the continuous and the interrupted conditions worked continuously on a puzzle during the first three minutes. Differences between these conditions occurred only after the third minute. In the break condition participants could switch back and forth between the puzzles at their own discretion. An example of a possible sequence of puzzles in this last condition is also shown in Figure 2.2. Continuous Interrupted Break A A A B C A B C B A B C B C B C C time Figure 2.2. Experimental Design, Task Sequences Procedure Each participant was accompanied by an observer. Beforehand, the observers handed out and verbally repeated written instructions regarding the goal and procedure of the assignment. The participants were instructed to notify their observer when a puzzle was solved, or when they had reached an impasse. Participants were debriefed at the end of the experiment and the solutions to the puzzles were sent to them by the next day. Measures The observers recorded solutions and impasses. Since the continuous condition and the interrupted condition differed only after the third minute of a puzzle's allotted time, they distinguished between solutions and impasses that occurred during the first half (first three minutes) and the second half (last three minutes) of a puzzle's allotted time. As is customary when an experimental task consists of multiple insight problems (Dodds et al., in 16

25 Do you need a break right now? press), the data of the three puzzles were combined prior to data-analysis: For each participant, we added the number of puzzles that had been solved during the first half of their allotted time. Similarly, the number of solutions per participant during the second half was calculated. The same procedure was used to calculate the number of impasses during the first and second halves of the puzzles' allotted time. Results Solutions In total, 85 word puzzles were solved, which means that, on average, participants solved less than one out of the three puzzles. Sixty-three puzzles were solved during the first three minutes that had been spent on them, and 22 during the last three minutes. Specifically, half of the puzzles (n = 44) were solved right away, during the first minute, confirming that these were insight problems: Either a solution was found immediately, or it was unlikely that it would be found at all. Means and standard deviations of the number of solutions per participant for each of the three conditions during the first three minutes (first half) and the last three minutes (second half) are presented in Table 2.1. Significant differences between the conditions are marked with different superscripts. Table 2.1. Means and Standard Deviations of Solutions and Impasses Solutions Impasses 1st half 2nd half 1st half 2nd half M SD M SD M SD M SD Continuous (N = 36).62 a a a a.95 Interrupted (N = 36).67 a a a b.69 Break (N = 36).51 a b a c.45 Note: Different superscripts indicate significant differences between conditions, p <.05 (one-tailed). 17

26 Chapter 2 Although most puzzles were solved during the first half of their allotted time, no differences were expected between the three conditions during this first half. A One-Way ANOVA confirmed this: No significant differences were found between the groups, F (2, 108) =.47, p =.63. A second One-Way ANOVA revealed differences between the groups during the second half of the allotted time, F (2, 108) = 4.63, p =.01, ω =.30. Scheffé's tests were conducted to evaluate pairwise differences among the conditions during this second half. Participants in the continuous condition did not solve more puzzles than participants in the interrupted condition, p =.87, ns. This means that Hypothesis 1 was not confirmed; interruptions were not related to more problems solved. In the break condition, all but one participant took breaks from the puzzles they were working on: They started to work on a puzzle, worked on one or more other puzzles temporarily, and then got back to the first one. On average, they switched between puzzles nine times. During the second half of the allotted time, participants in the break condition solved more puzzles than participants in the continuous condition, p =.010 (one-tailed). Hypothesis 2 was confirmed, which means that positive effects of breaks on problem solving performance were found. Furthermore, participants in the break condition solved more problems than participants in the interrupted condition, p =.035 (one-tailed). Hypothesis 3 was also confirmed: Individuals who could take a self-initiated break outperformed the interrupted individuals. Impasses Ninety-two impasses were reported, which means that, on average, almost all participants reached an impasse. 2.2 Thirty-four impasses were reported during the first half of the allotted time and 58 during the second half. The fact that more participants felt stuck during the second half supports our assumption that it takes some time to reach the psychological state of impasse. Means and standard deviations of the number of reported impasses per participant for each of the three conditions during the first and the second half of the allotted time are presented in Table One outlier was excluded, because this participant reported nine impasses. 18

27 Do you need a break right now? Again, effects were only expected during the second half of the allotted time. A One-Way ANOVA showed no differences between the three conditions during the first half of the allotted time, F (2, 108) =.18, p =.84 ns, and did reveal differences between the groups during the second half of the allotted time, F (2, 108) = 8.66, p <.000, ω =.42. Scheffé's tests were conducted to evaluate pairwise differences among the conditions during this second half. Participants in the interrupted condition reported fewer impasses compared to the participants in the continuous condition, p =.004 (one-tailed). This means that Hypothesis 4 was confirmed; interruptions reduced the number of reported impasses. The participants in the break condition also reported fewer impasses than the participants in the continuous condition, p =.001 (one-tailed). This means that Hypothesis 5 was confirmed; breaks also reduced the number of impasses. Furthermore, the participants in the break condition reached fewer impasses than the participants in the interrupted condition, p =.035 (one-tailed). Thus, Hypothesis 6 was also confirmed: Taking a break at one's own discretion had additional positive effects over being interrupted. Discussion The purpose of this study was to gain more insight into the differences between imposed and self-initiated breaks, and their effects on creativity, specifically, on insight and impasses. The results reveal two major differences: Individuals who took breaks at their own discretion (1) solved more problems and (2) reached fewer impasses than interrupted individuals. It was also found that interrupted individuals reached fewer impasses than individuals who worked continuously on problems. Insight Problem Solving These results add to the literature on insight problem solving and incubation time. First, the findings confirm the existence of the incubation effect suggested by Wallas (1926), in line with earlier research on this topic (Dodds et al., in press; see also Segal, 2004). Second, they are a unique contribution, because they imply that the incubation effect is primarily evident when individuals can take incubation time at their own discretion, and not when others determine the moment at which the task is interrupted. 19

28 Chapter 2 One possible explanation might be that a self-chosen moment to switch tasks allows individuals to end their cognitive activities on one task and start working on the next, whereas an interruption may not allow for such closure, and temporarily forces individuals to divide their cognitive resources (Norman & Bobrow, 1975). This may have negative influences on task performance (Bailey & Konstan, 2006; Eyrolle & Cellier, 2000). If so, the beneficial effect of the incubation period on the creative problem solving process might be undone by the negative effect of being interrupted. A second possible explanation might be that the incubation effect is stronger when individuals intentionally choose to take some incubation time. In a survey among 213 college professors, Wells (1996) found that those who intentionally set manuscripts aside for some time produced more published articles than those who did not. More research is needed to replicate our results and investigate the precise mechanisms responsible. The results have important implications for further research on the incubation effect. In previous research, interruptions, rather than self-initiated breaks, were used in experimental designs: Generally, participants were not allowed to choose when they switched to another task and back. Instead, they had to switch tasks at predefined moments (Dodds et al., in press). Exceptions are some early studies on incubation (e.g., Olton, 1979; Patrick, 1938) and a recent study by Christensen and Schunn (2005), in which participants were free to move back and forth between tasks. Other exceptions are experiments in which participants were allowed to switch tasks after they had reached an impasse (e.g., Butler & Thomas, 1999; Segal, 2004; Seifert, Meyer, Davidson, Patalano, & Yaniv, 1995). Considering our results, it might be preferable to incorporate breaks instead of interruptions in future experimental designs when studying the incubation effect. The Feeling of Being Stuck A second unique contribution of this chapter is that it is one of the first experimental studies in which impasses, or feelings of being "stuck" (Fleck & Weisberg, 2004), are considered to be an affective outcome of the creative process, and have therefore been measured as a dependent variable. The results of this study suggest that both interruptions and breaks help to prevent individuals from reaching an impasse, and that individuals who take breaks at their own discretion reach even fewer impasses than interrupted individuals. 20

29 Do you need a break right now? Presumably, the underlying processes of these positive effects are associated with emotions rather than cognitions. If individuals use a faulty problem solving strategy, negative feelings, such as frustration and confusion, might build up gradually and result in an impasse when they reach a certain critical level. Both interruptions and breaks might prevent the negative emotions from reaching this critical level. Emotions may also explain why breaks are to be preferred over interruptions. The division of cognitive resources over the main task and the second, interrupting task (Norman & Bobrow, 1975) can cause an increased level of stress (Kirmeyer, 1988) and negative emotions (e.g., Bailey & Konstan, 2006; Zijlstra et al., 1999). When these are added to the negative emotions caused by the unsolvable problem, interrupted individuals might reach the critical level of impasse at an earlier moment than individuals who are able to take breaks at their own discretion. In future research, the psychological state of impasse might be an interesting concept for further investigation. Specifically, it may be interesting to focus on both the psychological state of impasse and the cognitive state of fixation and see how these relate to each other, to performance measures, and to other psychological constructs. Fixation might be an important construct with regard to problem solving and creative performance, and negative psychological effects are to be expected from impasses. To gain more insight into the difference between fixation and impasse, one could conduct an experiment similar to that of Fleck and Weisberg (2004), in which participants were asked to think aloud while solving an insight problem. Fleck and Weisberg primarily focused on the processes leading to insight, but a future experiment could focus on faulty problem solving strategies (fixation) and the development of emotions leading to the feeling of being stuck. Interruptions are generally considered to have an unfavorable effect on performance (Jett & George, 2003). Therefore, time management training advises individuals to prevent them as much as possible (Green & Skinner, 2005; Macan, 1994). Although the laboratory results of the current study may not be fully generalizable to creative tasks in a work setting, they imply that in certain situations interruptions may be beneficial. When individuals need to work on a problem on which they do not make much progress, an interruption may prevent high levels of frustration and a feeling of being stuck. Yet, we found that taking a break at one's own discretion is preferable to being interrupted. 21

30

31 Chapter 3 Working on Something Else for a While: Uncovering Patterns of Task Engagement and Disengagement in Creative Professions * In two field studies we investigated the dynamics of task engagement in creative work. We set out to develop an archetypical pattern of task engagement and disengagement over time, by linking the literatures about flow (high task engagement), incubation time (disengagement from tasks), and task pacing behavior (the tendency to distribute work efforts over time in a certain way). The first study, a survey study among 88 employees, showed that individuals with more creative work more often used patterns that included intermediate periods of disengagement and patterns that included most engagement right before a deadline. The second study, an explorative interview study among 25 architects, was used to explore patterns of task engagement and disengagement in more depth. Several suggestions for further research are presented. A substantive amount of empirical evidence reveals that creativity at work is related to organizational innovation, effectiveness, and survival (Amabile, 1996). Moreover, the * This chapter is based on: Beeftink, Van Eerde, & Rutte (2008). Working on something else for a while: Uncovering patterns of task engagement and disengagement in creative professions. Manuscript submitted for publication. 23