ASSESSING THE ROLE OF PROCESSING STYLE IN
IDEA SELECTION
Master thesis, MSc Human Resource Management University of Groningen, Faculty of Economics and Business
April, 2012 NADINE SIERO Student number: 1933450 Oude Vlierweg 16 7731 SL Ommen Tel: +31(0)638704762 E-mail: s1933450@student.rug.nl
Supervisior: Dr. F.A. Rink Co-assessor: Prof. Dr. B.A. Nijstad
Acknowledgement: I want to thank everybody who supported me during my study period and a few I would like to mention in particular. First, my supervisor dr. Floor Rink and dr. Eric Rietzschel for the advice, support and useful feedback that helped me to improve my research skills during my master thesis. Second, Lieke Wigger and Reindert Dallinga for the
TABLE OF CONTENTS
ABSTRACT 4
INTRODUCTION 5
Selection Effectiveness 6
THEORETICAL FRAMEWORK EXPERIMENT 1 7
Processing Style, Selection Instruction and Idea Selection 7 Processing Style, Selection Instruction and Satisfaction 8
METHOD EXPERIMENT 1 10
Participants 10
Independent Variables 10
Control variables 10
Manipulation of selection instruction 11
Manipulation of processing style 11
Measures 12 Selection effectiveness 12 Questionnaire items 12 RESULTS EXPERIMENT 1 13 Manipulation Check 13 Descriptive Statistics 13
Test of Hypothesis 1: Selection effectiveness 14
Test of Hypothesis 2: Satisfaction 16
Supplementary Analyses 16
DISCUSSION EXPERIMENT 1 17
INTRODUCTION EXPERIMENT 2 18
THEORETICAL FRAMEWORK EXPERIMENT 2 19
Processing Style, Kind of Idea and Idea Rating 19
Creativity ratings 19
Good and relevant ratings 20
Processing Style and Idea Selection 21
METHOD EXPERIMENT 2 22
Participants 22
(In)dependent Variables 22
Control variables 22
Phase 1: Idea generation 23
Manipulation of processing style 23
Phase 2: Idea selection 24
Measures 24
Idea selection 24
Idea rating 25
RESULTS EXPERIMENT 2 25
Manipulation Check 25
Descriptive Statistics 25
Test of Hypotheses 1: Idea Rating 27
Test of Hypotheses 2: Idea Selection 28
Supplementary Analyses 29 Idea rating 29 Idea selection 30 DISCUSSION EXPERIMENT 2 30 GENERAL DISCUSSION 31 Findings 31
Limitations and Future Research 32
Conclusions and Practical Implications 35
REFERENCES 36
APPENDIX A: ITEMS POST-EXPERIMENTAL QUESTIONNAIRE 41
EXPERIMENT 1
APPENDIX B: FIGURES SIGNIFICANT MODERATORS 42
SUPPLEMENTARY ANALYSES EXPERIMENT 1
APPENDIX C: CODING SCHEME BRAINSTORM SESSION 47
EXPERIMENT 2
APPENDIX D: ITMES POST-EXPERIMENTAL QUESTIONNAIRE 49
ABSTRACT
Many scholars have studied idea generation, however idea selection has received little research attention so far. The goal of this study was to examine how selection effectiveness (i.e. selecting an idea that is feasible yet original at the same time) could be improved. Central concept in the two experiments was the effect of processing style – the way people personally look at or attend information - on idea selection. Hypotheses could not be confirmed since global processing - a focus on general rather than specific features- did not result in the selection of high quality ideas. Selection effectiveness was only improved by giving participants specific creativity instructions. Supplementary analyses revealed that this relationship became even stronger when people were in a positive mood or had a promotion focus.
INTRODUCTION
‘We think good ideas to death, when we should be acting them to life’- Brian G. Jett
There is a growing interest by both academicians and management practitioners in creating work environments that support and nurture employees’ creativity (Shalley & Perry-Smith, 2001). Most theorists have defined creativity as the development of ideas about products, practices, services or procedures that are novel and potentially useful to the organization (Amabile, 1996; Shalley, Zhou & Oldham, 2004; Zhou & Shalley, 2003). The use and development of creative ideas allows organizations to adjust to shifting market conditions, respond to opportunities, and thereby, to adapt, grow and compete (Nonaka, 1991; Oldham, 2002). It is important to distinguish creativity from innovation. Creativity refers to the development of novel, potentially useful ideas. Only when the ideas are successfully implemented at the organization or unit level they would be considered innovation (Amabile, 1996; Mumford & Gustafson, 1988). Importantly creativity does not only exist out of idea generation. After this initial phase of creativity, people have to make a selection of these ideas for further development and eventual implementation within organizations (Nijstad & De Dreu, 2002).
Selection Effectiveness
Selection effectiveness depends on the quality of selected ideas which is usually based on the combination of ratings of novelty (i.e. originality: the extent to which an idea is considered innovative, new, imaginative, surprising, novel or at least novel to the individual and departs from existing paradigms and practices) and appropriateness (i.e. feasibility: the degree to which an idea is considered relevant to the topic at hand; e.g. Amabile, 1996; Diehl & Stroebe, 1987; Mumford, Blair, Dailey, Leritz & Osburn, 2006; Puccio & Cabra, 2012; Rietzschel, Nijstad & Stroebe, 2010; Sternberg & Lubart, 1999; Stroebe, Nijstad & Rietzschel, 2010). The degree of originality and feasibility of an idea depends on the situation: in some circumstances management might consider incremental ideas as desirable, whereas in other circumstances more radical ideas might be valued (Woodman, Sawyer & Griffin, 1993). Although, in particular situations conventional solutions are very useful and effective, brainstorming is especially used when conventional solutions are not available or will not work (Rietzschel et al., 2009). Therefore feasibility alone is often not sufficient - originality is certainly a requirement. Unfortunately though, research shows that people prefer to select feasible ideas over creative ideas because people tend to favour practical and realistic solutions to problems (e.g. Dillon, Graham & Aidells, 1972; Harari & Graham, 1975 in Rietzschel, 2005). The natural tendency to focus on feasibility is due to the belief that feasibility and originality are incompatible (Rietzschel, 2005; Rietzschel et al., 2010). Therefore selecting an idea that is both feasible and original is difficult. I will study whether the selection of high quality ideas depends on an important situational condition, that is, the presence of specific selection instructions and on one important individual difference characteristic: a person’s processing style .
Giving participants the explicit instruction to select creative ideas resulted in the selection of more creative and original ideas compared to giving participants the instruction to select the best idea (default instruction). People with the default instructions give more weight to feasibility and desirability. In other words there seems to be a trade-off between originality and feasibility: the more participants tried to select creative or original ideas, the less likely they were to select feasible or desirable ideas. In fact, those people who were instructed to select the most creative ideas reported to be less satisfied with their final choice than the people who were instructed to select the best ideas (Rietzschel et al., 2010).
meaning that this may depend on people’s processing style. A processing style may be described as the way we personally look at or attend to information (Förster, 2009). In a global processing mode, individuals are focused on the forest rather than the trees, they use broad mental categories, focus on general rather than specific features and represent knowledge in more abstract terms. In a local processing mode people focus on the trees rather than the forest, they use narrow mental categories, focus on details and represent knowledge in specific and concrete terms (De Dreu, Baas & Giacomantonio, 2010; Förster 2009; Förster, Friedman & Liberman, 2009; Nijstad, De Dreu, Rietzschel & Baas, 2010). Research shows that a global processing mode is more beneficial to creativity, e.g. people primed with a global processing mode create more atypical exemplars for a number of categories (Förster & Dannenberg, 2010; Friedman, Fishbach, Förster, & Werth, 2003). The same study also shows that those people came up with more creative titles for a cartoon and unusual uses for a brick (Förster & Dannenberg, 2010; Friedman et al., 2003).
Importantly though, the above research has only linked processing modes to idea generation (e.g. Förster & Dannenberg, 2010; Förster & Friedman, 2008 in Förster et al., 2009; Friedman et al., 2003). Because factors are not necessarily identical for the different phases of creativity – that is, idea generation and idea selection (Rietzschel et al., 2009) - it is important to study the influence of processing modes on idea selection separately. This master thesis contributes to this notion. In experiment 1 the moderating role of processing style in the relationship between selection instruction and selection effectiveness and satisfaction will be studied. In study 2 the relationship between processing style and idea selection will be studied again, but this time, idea rating will be investigated.
THEORETICAL FRAMEWORK EXPERIMENT 1 Processing Style, Selection Instruction and Idea Selection
Abstract categories used by global processors are broader and naturally more inclusive. This increases the chances of inclusion of the target in existing knowledge structures (Förster, Liberman & Shapira, 2009). Additionally this processing style is associated with a promotion focus, characterized by a focus on advancement, growth, accomplishments, exploration and a risky response bias (Crowe & Higgins, 1997; Friedman & Förster, 2001). Accordingly, people in a global focus experience creative ideas as less deviant and risky. I therefore propose that they are likely to select a creative idea even in the default situation where they are simply instructed to select the best possible idea – an instruction which on average leads people to prefer feasibility over creativity.
Because of their negative attitude towards creativity, people with a local processing style are not likely to choose an original idea when they are primed with a default instruction. Those people are more likely to have a prevention focus and are thus concerned with security, safety, responsibility and are likely to have a conservative response bias (Crowe & Higgins, 1997). The core concepts of local processors are: clear, close and concrete. They see novelty as threatening and therefore focus on familiarity (Förster & Dannenberg, 2010). Instead of being creative they are more analytical. Local processors focus on details and use narrow categories, therefore it is more likely that they focus on feasibility. Importantly though, in line with the finding that people with creativity instructions are able to select original ideas (Rietzschel, 2005; Rietzschel et al., 2010), it will be assumed that people’s processing style will then have less influence on the extent to which people select feasible vs. original ideas. Taken together, I propose;
Hypothesis 1: The relationship between selection instruction and idea selection (feasibility and originality) will be moderated by people’s processing style. When receiving the default instructions, global processers are more likely to select ideas that are feasible and creative compared to local processers. When receiving the creativity instructions, both global and local focused people will select a feasible and original idea.
Processing Style, Selection Instruction and Satisfaction
define as the ‘best’ and their choice, they are satisfied. To illustrate: global processors rate an idea as good when it is creative. If they select an idea that is in their opinion creative, they are satisfied with their choice. Local processors are focused on feasibility and are satisfied – or experience fit - when they choose a feasible idea. In other words, people can make use of their own norms / standards when selecting the best idea and are therefore satisfied with their choice.
With creativity instructions however, it has been found that people tend to be less satisfied with their final decision - especially compared to the default instructions (Rietzschel et al., 2010). I argue that this is primarily the case for people with a local processing style. They are not open to creative ideas, and it is more likely that they focus on feasibility. When local processors are instructed to select a creative idea, they cannot experience a fit between their processing style and the task instruction. According to their processing style they would like to select a feasible idea, however the instruction is to select a creative idea. Because of this incongruence it is likely that people with a local processing style are unsatisfied with their idea selection. By contrast, global processors are more creative, they rate original ideas as less deviant and are therefore more likely to select a creative idea. In this situation there is a fit: people are instructed to select a creative idea and according to their global processing style they are also likely to select a creative idea. This is likely to result in satisfaction. Therefore the following hypothesis will be tested:
Hypothesis 2: The relationship between selection instruction and selection satisfaction will be moderated by people’s processing style. When receiving the default instructions, both local and global processes will be satisfied, but when receiving the creativity instructions, only global processers will be satisfied.
FIGURE 1
Proposed Model Experiment 1
METHOD EXPERIMENT 1
Participants
One hundred and twenty two undergraduate students of the University Groningen participated in this study: 45.1 % women and 54.9 % men. The age of the participants was ranging from 17-28 years old (M = 20.6, SD = 2.37). 95.9 % participants had the Dutch nationality, 3.3 % German and 0.8 % other. Since the experiment was in Dutch, only students who could properly understand the Dutch language could participate. Participants received course credits or € 4 for their participation. After the experiment had ended, participants were debriefed and thanked.
Independent Variables
The experiment had a 2 (processing style: global or local) x 2 (selection instruction: creative or default) factorial design. Participants were randomly assigned to one of the experimental conditions. All participants worked individually throughout the experiment. Upon arrival in the laboratory, participants received and read general instructions (which were identical for all conditions).After that they were seated in individual cubicles equipped with a chair, a desk, and a computer with keyboard.
Second, research shows the influence of regulatory focus on creativity: promotion focus leads to higher creativity than prevention focus (e.g. Friedman and Förster, 2000; 2001; 2002). Regulatory focus was assessed, by using 20 items based on Lockwood, Jordan and Kunda (2002). I aggregated these items to compute a total promotion focus score and a total prevention focus score. Also a dominant regulatory focus score was computed by subtracting scores on the prevention goal subscale from scores on the promotion goal subscale. Higher scores on this measure reflect relatively greater promotion than prevention focus. The data meet criteria required to satisfy a difference score analysis (Edwards, 1994; 1995; Lockwood et al., 2002). One item of the promotion items was deleted in order to increase the reliability. The Cronbach’s alpha for the nine promotion items was .74 and for the ten prevention items was .74.
Besides the effects of mood and regulatory as control variables, their moderating influence in the relationship between selection instruction and selection effectiveness is expected. People are able to select original ideas when explicitly instructed to do so (Rietzschel, 2005; Rietzschel et al., 2010). Research shows the positive effect of positive mood and promotion focus on creativity (e.g. Baas et al., 2008; Nijstad et al., 2010; Friedman and Förster, 2000; 2001; 2002). Therefore it is likely that there is a potentiating effect: the higher the positive mood or promotion focus, the stronger the positive association between selection instruction and selection effectiveness.
Manipulation of selection instruction. A list of 20 ideas was used, which were generated in previous experiments were participants brainstormed about possible ways to improve health (Rietzschel, Nijstad & Stroebe, 2007). The ideas were selected to cover the entire range, from ‘not at all’ to ‘very much’, of originality and feasibility scores. Participants were asked to select and make a rank order of three ideas - in decreasing order - for the municipality of their hometown on how to improve the health of the residents. The idea set was presented three times, with the ideas ordered randomly: the first time participants were asked to select an idea for position one, the second time for position two, etc. Half of the participants received the creativity instruction: select the three most creative ideas from this set (creativity condition) and half of the participants were simply instructed to select the three best ideas (default condition). Participants were told that there was no time restriction; they could take as much time as needed.
intended to manipulate processing style. Participants were presented with eight global letters made up of local letters, together these eight letters made up a word. A sample item is presented in figure 2. Half of the participants were instructed to focus on the large letters, together making the word ‘abstract’ (global condition) and half to focus on the small letters, together making the word ‘concrete’ (local condition).
FIGURE 2
Sample Item from the Navon Letter Task: Global = A, Local = B (Navon, 1977)
After completion of the manipulation task, the system was connected to the server, and participants selected three ideas, as described above.
Measures
Finally, participants were presented with a post-experimental questionnaire to measure selection criteria used and satisfaction.
Selection effectiveness. I computed the average originality, average feasibility and an aggregated measure (combining originality and feasibility) of the three ideas selected by each participant.
RESULTS EXPERIMENT 1 Manipulation Check
MANOVA was used to test the effects of the selection instruction, both items were significant, default F(1,120) = 6.22, p = .14) and creativity F(1,120) = 125.78, p = > .001). As intended participants in the default manipulation argued that they were expected to choose the best idea (M = 6.34, SD = 2.39) compared to participants in the creativity condition (M = 5.13, SD = 2.93). Opposed, participants in the creativity manipulation argued that they were expected to choose the most creative idea (M = 7.17, SD = 2.24) compared to participants in the default condition (M = 3.16, SD = 1.67). These results indicate that the manipulation was successful.
Descriptive Statistics
Table 1 presents the correlations of the conditions (processing style and selection instruction), average originality, feasibility and aggregated measure of participants’ selected ideas, as well as satisfaction with the selected ideas. As one can see, there are no significant relations between processing style and the other main study variables. This also applies to the relationship between satisfaction and the other study variables. However, the other correlations are consistent with previous research, as for instance the significant negative correlation between originality and feasibility of the selected ideas (r = -.83, p = < .01). Furthermore, there was a significant positive correlation between selection instruction and originality of the selected ideas (r = .46, p = < .01). Selection instruction was also significantly correlated, although negatively, with average feasibility (r = -.38, p = < .01).
TABLE 1
Correlations among Study Variables
Variable 1 2 3 4 5 6
Test of Hypothesis 1: Selection Effectiveness
Hypothesis 1 predicted that the relationship between selection instruction and idea selection (feasibility and originality) would be moderated by people’s processing style. I performed a one way ANOVA on selection effectiveness with selection instructions and processing style as independent factors and mood and regulatory focus as covariates. There was no significant interaction effect F(1,114) = 1.89, p = n.s. However, a significant main effect for selection instruction was found. As shown in table 2, participants with the creativity instruction selected, independent of their processing style, ideas of a higher aggregated measure (Mglobal = 3.03, SDglobal = .29; Mlocal = 3.08, SDlocal = .28) as did participants with the
default instruction (Mglobal = 2.88, SDglobal = .17; Mlocal = 2.84, SDlocal = .17; F(1,114) = 22.87,
p = > .001). Thus, hypothesis 1 can only partially be confirmed.
Also when ANOVA’s were performed with average originality and feasibility as separated dependent variables, or questionnaire items were used as dependent variables, no significant interaction effects were found. Again, results show significant main effects for selection instruction which is in accordance with earlier empirical findings (Rietzschel, 2005; Rietzschel et al., 2009; Rietzschel et al., 2010). Participants with the creativity instruction selected, independent of their processing style, ideas of higher originality (Mglobal = 2.47,
SDglobal = .91; Mlocal = 2.61, SDlocal = .83) as did participants with the default instruction
(Mglobal = 1.81, SDglobal = .58; Mlocal = 1.77, SDlocal = .60; F(1,114) = 33.47, p => .001). This
is corresponding to the results of the questionnaire items, as shown in table 3: participants with the creativity instruction found it more important to choose the most original (F(1,118) = 21.87, p = > .001), the nicest (F(1,118) = 11.93, p = > .001), most surprising (F(1,118) = 8.46, p = > .001) and creative idea (F(1,118) = 18.40, p = > .001) compared to participants with the default instruction. Besides in their opinion they choose the most original idea (F(1,118) = 17.78, p = > .001) and did not select some ideas because they were boring (F(1,118) = 8.45, p = > .001).
Contrasting, participants with the default instruction selected independent of their processing style, ideas of higher feasibility (Mglobal = 3.95, SDglobal = .52; Mlocal = 3.90, SDlocal
= .41) as did participants with the creativity instruction (Mglobal = 3.59, SDglobal = .44; Mlocal =
3.54, SDlocal = .40; F(1,114) = 22.00, p = > .001). This is also shown by the questionnaire
to select the least bad ideas (F(1,118) = 7.66, p = > .001), and most useful ideas (F(1,118) = 7.53, p = > .001) and did not select some ideas because they were bizarre (F(1,118) = 4.31, p = .04).
Thus, I can not confirm that global processers are more likely to select ideas that are feasible and creative at the same time than local processers when receiving the default instructions. Rather, creativity instructions contribute to the selection of creative ideas, whereas feasible ideas are selected when default instructions are given.
TABLE 2
Means and Standard Deviations for Selection Effectiveness
(Originality, Feasibility and Aggregated Measure) and Satisfaction of the Selected Ideas
Global Local Measure Creativity instruction Default instruction Creativity instruction Default instruction Average originalitya 2.47 (0.91) 1.81 (0.58) 2.61 (0.83) 1.77 (0.60) Average feasibilitya 3.59 (0.44) 3.95 (0.52) 3.54 (0.40) 3.90 (0.41) Average aggregated measurea 3.03 (0.29) 2.88 (0.17) 3.08 (0.28) 2.84 (0.17) Satisfactionb 7.65 (1.20) 7.56 (1.11) 7.43 (1.33) 7.52 (1.17)
Note.N = 122
a
Measures refer to average originality, feasibility and aggregated measure of the selected ideas. Minimum value = 1, maximum value = 5.
b Measure refers to participants’ satisfaction with selected ideas. Minimum value = 1, maximum value = 9.
TABLE 3
Means and Standard Deviations for Questionnaire Items
Measure Creativity instruction Default instruction
Important to select most original ideasa
5.58 (2.47) 3.68 (2.02) Important to select nicest ideas a 6.13 (2.14) 4.89 (1.87) Important to select most
surprising ideas a
5.32 (2.57) 4.11 (1.98) Important to select most
creative ideas a
5.82 (2.43) 4.16 (1.83) I my opinion I chose the most
original ideas a
5.93 (2.28) 4.32 (1.91) Did not select some ideas,
because they were boring a
5.23 (2.52) 3.98 (2.18) Important to select most
concrete ideas a
5.89 (1.88) 6.68 (1.72) Important to select most
familiar ideas a
3.50 (2.12) 4.32 (1.93) Important to select most
excellent idea a
6.42 (1.93) 7.08 (1.50) Important to select most
realistic ideas a
Important to select easy implementable ideas a
4.83 (2.20) 5.90 (1.97) Tried to select least bad ideas a 3.38 (2.34) 4.58 (2.41) I my opinion I chose the most
useful ideas a
6.35 (2.07) 7.23 (1.50) Did not select some ideas,
because they were bizarre a
4.68 (2.92) 5.76 (2.76)
Note. N = 122 (60 creativity, 62 default).
Standard deviations are presented in parentheses
The table is only classified by selection instruction, because processing style revealed no significant results
a
Maximum value = 9
Test of Hypothesis 2: Satisfaction
The second hypothesis was about the moderating effect of processing style in the relationship between selection instruction and selection satisfaction. When receiving the default instruction, both local and global processes are expected to be satisfied, but when receiving the creativity instructions, only global processers will be satisfied. A one way ANOVA, corrected for mood and regulatory focus, with satisfaction as dependent variable and selection instruction and processing style as independent variable revealed no significant interaction effect F(1,114) = .02, p = n.s. Thus, there were no differences between the conditions (global vs. local and creativity vs. default) with regard to satisfaction with the idea selection, therefore hypothesis 2 can also not be confirmed.
Supplementary Analyses
Although not hypothesized there were also several interesting significant correlations found between mood and regulatory focus and the questionnaire items. It is especially interesting that many items are significant for both, promotion focus and positive mood. E.g. ‘I found it important to choose the most ambitious ideas’ (rpromotion focus = .21, ppromotion focus = 0.02; rpositive mood = .24 , ppositive mood = > .001). Promotion focus and positive mood are likewise positive correlated (r = .36, p = > .001). The same holds for prevention focus and negative mood, e.g. ‘I found it important to choose the most safe ideas’ (rprevention focus = .34, pprevention focus = > .001; rnegative mood = 0.31, pnegative mood = > .001). Prevention focus and negative mood are likewise positive correlated (r = .45, p = > .001).
DISCUSSION EXPERIMENT 1
This experiment was designed to test the moderating influence of processing style in the relationship between selection instruction and selection effectiveness. It was found that the instruction to choose the most creative idea more often caused participants to select ideas that scored higher on the aggregated measure of originality and feasibility than the instruction to select the best ideas. When this measure was separated into creativity and feasibility, it became clear that participants with the creativity instructions selected ideas of higher originality whereas participants with the default instructions selected ideas of higher feasibility. The post task questionnaire confirmed that participants who received the creativity instruction found it important to choose original, surprising and creative ideas and those participants with the default instructions found it important to choose concrete, familiar, realistic and easy implementable ideas. These findings are consistent with previous research, showing that originality is clearly not a quality dimension that people take into account spontaneously (e.g. Puccio & Cabra, 2012; Rietzschel, 2005). When not provided with specific selection criteria, people tend to select feasible rather than original ideas (Rietzschel, 2005; Rietzschel et al., 2009; Rietzschel et al., 2010). Note however, that I also did not obtain effects with regard to participants’ satisfaction with their idea selection. This is important, given that if people are satisfied with the ideas chosen, this is likely to result in higher acceptance of new ideas and consequently implementation success (De Dreu & West, 2001; Mumford & Gustafson, 1988; West, 1990).
local processing style (e.g. De Dreu, Giacomantonio, Shalvi & Sligte, 2009; Friedman, Fishbach, Förster & Werth, 2003; Giacomantonio, De Dreu, Shalvi, Sligte & Leder, 2010), one being the Navon Letter task used in this experiment (global letters formed by the configuration of local letters, Navon, 1977). However, in this experiment participants got only eight letters in order to induce processing style, which may not have been strong enough. The failure of this manipulation could also account for the null finding on the satisfaction scale after the task had finished.
INTRODUCTION EXPERIMENT 2
Study 2 builds on Study 1 in a number of important ways. First, a different manipulation will be used to induce global vs. local processing style – in this study, I rely on abstract and concrete mindsets (Freitas, Gollwitzer & Trope, 2004). An abstract mindset (why) is associated with global processing and a concrete mindset (how) is associated with local processing (Förster & Dannenberg, 2010; Förster et al., 2004). Past work (Freitas et al., 2004) shows that the accessibility of the cognitive operations of considering an activity’s process (how/concrete) or purpose (why/abstract) can color how people construe newly encountered activities. This implies that the manipulation induces processing modes (local or global) that may transfer to subsequent task performance (i.e. idea rating and selection).
Second, in Study 1 a pre-generated list of 20 ideas was used from which participants selected three ideas, so it was focused on idea selection only. Although improving health is a well-known and actual topic, it is possible that participants could not relate to the ideas presented to them and therefore selection effectiveness was low. Besides, previous research about interpersonal and intrapersonal evaluation (i.e. judgment others’ ideas vs. judgment of one’s own ideas) showed that people are generally more accurate in evaluating the originality of their own ideas versus the ideas generated by others (Grohman, Wodniecka & Klusak, 2006; Runco & Smith, 1992). Thus, to rule out the possibility that selection effectiveness was low due to these reasons, participants will first brainstorm individually about possible ways to improve health and subsequently rate their own ideas on a number of dimensions in Study 2.
Puccio & Cabra, 2012). Idea evaluation begins with forecasting the likely outcomes and consequences of implementing an idea within a particular setting (Blair & Mumford, 2007; Doerner & Schaub, 1994; Mumford, Lonergan & Scott, 2002). Insights gained through forecasting lead to an appraisal of the idea. The consequences and expected outcomes are weighted against desired performance standards, and a decision is made whether to implement the idea as it is (i.e. select the idea), to drop the idea (i.e. did not select the idea) or to revise it (Blair & Mumford, 2007; Mumford, Lonergan & Scott, 2002; Puccio & Cabra, 2012). The first two operations, forecasting and appraisal, are associated with idea rating and the last one, decision making, is associated with idea selection. Study 2 will examine the first two operations, or this so-called idea rating. By asking participants to rate their ideas they come up on a number of dimensions, it is possible to examine whether the two processing styles cause people to rate their ideas differently. It could be that Study 1 did not yield significant results of processing style on the actual idea selection, because global and local processors rate feasible and creative ideas differently.
THEORETICAL FRAMEWORK EXPERIMENT 2
Processing Style, Kind of Idea and Idea Rating
Dean, Hender, Rodgers, and Santanen (2006) carried out an exhaustive analysis to define constructs and sub-dimensions that are useful in evaluating the creativity of ideas. Three constructs will be used in this experiment: creativity (novelty in terms of Dean et al. defined as ‘the degree to which an idea is original and modifies a paradigm’), good (definition according to the Oxford dictionary: ‘having the qualities that are desirable’) and relevance (in terms of Dean et al. (2006) defined as ‘the idea applies to the stated problem and will be effective at solving the problem’).
differences. They focus on separating targets from one another, breaking them into their constituent parts, belonging to different categories (Förster, 2009). Thus, globally processing two targets fosters a focus on similarities (inclusion in the standard’s category), whereas during local processing dissimilarities may become more prominent (excluded from the category; Förster & Dannenberg, 2010). Suppose that there are two ideas, a ‘normal’ and highly original idea. In a broad, global processing approach, the viewer ‘zooms out’ on the two ideas whereby their distance is decreased and the perceived similarity increased. Similarities are generated between incoming information (i.e. highly original idea) and stored knowledge (i.e. normal idea). Because of the focus on similarities, the highly original idea will be perceived as less rare and surprising and therefore rated lower on creativity. Contrasted, with a narrow local processing style, the viewer ‘zooms in’ on the two ideas whereby their distance and the perceived dissimilarity are increased. Due to the focus on differences between the normal and highly original idea, it is likely that the highly original idea is rated higher on creativity as by people with a global processing style. Therefore, the following hypothesis will be assumed:
Hypothesis 1a: Participants induced with a global processing style rate more original ideas as less creative compared to participants with a local processing style
Hypothesis 1b: Participants induced with a global processing style are more likely to rate more original ideas as good and relevant than participants with a local processing style
Hypothesis 1c: Participants induced with a global processing style are less likely to rate more feasible ideas as good and relevant than participants with a local processing style.
Processing Style and Idea Selection
As in Study 1, it is assumed that there is an individual difference in the extent to which people prefer feasible ideas over creative ideas, displayed by people’s processing style. The actual idea selection takes place after forecasting and appraisal of the available ideas (Mumford, Lonergan & Scott, 2002; Puccio & Cabra, 2012). Given that global and local processors are expected to rate original and feasible ideas differently, it is likely that these differences transfer correspondingly to the next cognitive process, the decision whether to select the idea, drop the idea or revise it (Blair & Mumford, 2007; Mumford, Lonergan & Scott, 2002; Puccio & Cabra, 2012).
Past research shows (Rietzschel et al., 2010) – but also the results of experiment 1- that feasibility, rather than originality, is a criteria that people spontaneously apply to idea selection. Therefore it is likely that people are, independent of their processing style, able to select feasible ideas. However, for originality this is different. The abovementioned characteristics of global processors makes it likely to assume that they are able to select high quality ideas (i.e. scoring high on both, originality and feasibility). Contrasted to local processors who should select ideas scoring low on originality, but high on feasibility. Therefore the following hypotheses will be assumed:
Hypothesis 2: Participants induced with a global processing style select an idea that scores high on originality and feasibility. For them, the association between originality and feasibility should be positive.
FIGURE 3
Proposed Model Experiment 2
METHOD EXPERIMENT 2
Participants
One hundred and sixty four undergraduate students of the University Groningen participated in this study (four students were excluded from the analyses, because the manipulation did not work successfully, so the analyses are based on one hundred and sixty students). 40.6 % of the participants were women and 59.4 % were men. The age of the participants was ranging from 17-29 years old (M = 20.2, SD = 2.14). 99.4 % participants had the Dutch nationality, the other 0.6 % German. Since the experiment was in Dutch, only students who could properly understand the Dutch language could participate. Participants received course credits or € 6,00 for participation. After the experiment had ended, participants were debriefed and thanked.
(In)dependent Variables
The experiment had a 2 (processing style: global or local) factorial design. Participants were randomly assigned to one of the experimental conditions. All participants worked individually throughout the experiment.
Control variables. As in experiment 1, prior to the start of this experiment, participants completed a personality questionnaire to measure mood and trait regulatory focus. Positive affect (PA) and negative affect (NA) (Watson et al., 1988) were measured with the same items, Cronbach’s alpha of .85 for both. Because of the rather low Crobach’s alpha
Processing style (global / local)
Idea rating (creative, good, relevant)
Idea selection (original / feasible)
Kind of idea (original / feasible)
H1
for promotion and prevention focus in study 1, a different translation was used based on Lockwood et al. (2002). The Cronbach’s alpha for the ten promotion items was .81 and the alpha for the ten prevention items was .79. Again a dominant regulatory focus score was computed by subtracting scores on the prevention goal subscale from scores on the promotion goal subscale (Edwards, 1994; 1995; Lockwood et al., 2002). See method experiment 1 for more information.
Phase 1: Idea generation. Participants took part in an individual brainstorm session for 2 x 2.5 minutes. The time (2.5 minutes) was determined by pretesting ten people and counting the number of ideas they generated within a certain amount of time. The mean number of ideas within 2.5 minutes in the pretesting stage was six ideas. The brainstorm question was: ‘What can residents of municipality Groningen do to maintain or improve their health?’ The first session in the experiment was meant to practice with brainstorming and become familiar with the topic. The following brainstorming rules were applied: (1) Quantity counts not quality, (2) Postpone and withhold your judgment, (3) Freewheeling is welcomed and (4) Combine and improve.
FIGURE 4
Diagram Directing Participants to Think Increasingly Abstractly (lef side) or Increasingly Concretely (right side) about Improving Study Results, Which Served as General Abstract versus Concrete Mindset Inductions (figure adapted from Freitas et
al., 2004)
Phase 2: Idea selection. After completion of this mindset induction task, participants were asked to rate their own ideas they came up with during the official brainstorm session. Each idea was separately presented on the computer screen. The participant answered successively three questions per idea: How creative do you rate this idea? How good do you rate this idea? How relevant do you rate this idea? In the last task participants were presented with all the ideas they came up with during the final brainstorm session and were asked to choose the idea that they ultimately found the best (idea selection task).
Measures
Finally, participants were presented with a post-experimental questionnaire to measure which criteria they used for idea rating and idea selection.
Idea selection. All the 1198 ideas, including the ideas selected by participants during the selection task, were coded by the researcher for originality and feasibility using an existing and validated coding scheme (see appendix C, Rietzschel et al., 2007). For both dimensions, a 5-point scale was used ranging from ‘not original [feasible]’ (1) to ‘highly original [feasible]’(5). An example of an unoriginal idea of this experiment is ‘stop smoking’ and an example of an original idea is ‘develop healthy cigarettes’. Likewise, an example of an unfeasible idea for this topic is ‘prohibit snack bars’, whereas a feasible idea would be ‘do more sports’. A second rater scored 10% of the ideas to determine interrater reliability. Agreement existed in 93.6 % of the cases for originality, and in 97.8 % of the cases for feasibility. Differences between the raters were solved by discussion, which resulted in one originality and one feasibility score for each unique idea.
Improve study results How ↓ How ↓ How ↓ Why ↑ Why ↑ Why ↑
Idea rating. Participants themselves also rated all the ideas they came up with during the second brainstorm session. They rated these ideas on three dimensions: creativity, good and relevance. For all dimensions a 7- point Likert scale was used, ranging from ‘not creative [good] [relevant]’ (1) to ‘highly creative [good] [relevant]’(7).
Questionnaire items. All 41 questionnaire items of the post-experimental questionnaire were answered on a 9 point Likert scale ranging from ‘Completely disagree’ (1) to ‘Completely agree’ (9). The adapted items of Rietzschel (2005) assessed a wide range of criteria that participants used for idea rating and idea selection (e.g. ‘It is more important that an idea is creative compared to relevant’). Therefore, all items were analysed separately rather than combined into one reliable scale in order to prevent for loosing data. Finally, we checked people’s processing style with two self-developed items: ‘I have been asked to answer some questions how I might improve my study results (local processing)’ and ‘I have been asked to answer some questions why I might improve my study results (global processing)’. All questionnaire items are included in appendix D.
RESULTS EXPERIMENT 2
Manipulation Check
I used a MANOVA to test the effects of the processing style manipulation on the checks. As intended, both items were significant, global F(1,158) = 11.35, p < .01) and local F(1,158) = 28.05, p < .01). Participants in the global manipulation argued that they were asked to answer some questions why they might improve their study results (M= 5.31, SD= 3.19) compared to participants in the local condition (M= 3.73, SD= 2.75). Opposed participants in the local manipulation argued that they were asked to answer some questions how they might improve their study results (M= 5.56, SD= 3.13) compared to participants in the global condition (M= 3.14, SD= 2.61). These results indicate that the manipulation was successful.
Descriptive Statistics
.61, p < .01; local r = .58, p < .01). The same holds for the correlation between creativity and good (global and local r = .19, p < .01). Yet only participants with a local processing style significantly agree with expert ratings about the dimension creativity (r = .16, p < .01).
Table 5 is comparable to table 4, however the originality ratings of experts is replaced by feasibility ratings of experts. As in previous research, the ratings of feasibility and originality are significantly, but negatively correlated, and this is the case independently of people’s processing style. That is, the relationship between people’s own creativity ratings and the feasibility scores of the expert (global r = -.31, p < .01; local r = -.13, p < .01). For both processing styles feasibility ratings of experts and relevance ratings by participants are positive correlated (global r = .11, p < .01; local r = .15, p < .01), besides ratings of participants for good and relevant are positive correlated (global r = .67, p < .01; local r = .51, p < .01). Interesting is the significant negative correlation between creativity ratings and relevance ratings of participants with a global processing style (r = -.19, p < .01) and the significant positive correlation of creativity ratings and good ratings of participants with a local processing style (r = .15, p < .01)
TABLE 4
Means, Standard Deviations, and Correlations Original Ideas
Global M SD 1 2 3 4
1. Originality rating experta 2.14 0.39
2. Creativity rating participantb 4.46 1.70 -0.07
3. Good rating participantb 5.31 1.21 -0.07 --.19**
4. Relevance rating participantb 5.29 1.34 -0.08 0-.06 -0.61**
Local M SD 1 2 3 4
1. Originality rating experta 2.18 0.43
2. Creativity rating participantb 4.71 1.69 -0.16**
3. Good rating participantb 5.06 1.43 0-.06 -0.19**
4. Relevance rating participantb 5.18 1.51 0-.05 -0.02 -0.58** Note. N = 582 original ideas (298 global, 284 local). ** p < .01
a
Measures refer to the average rating of the most original ideas by an expert. First, all ideas were scored on originality, scale 1-5. A median split was executed to divide the ideas in original and not original (median = 1). Because we were only interested in the most original ideas, ideas with score 1 were removed from further analyses. Minimum value = 2, maximum value = 5.
b
TABLE 5
Means, Standard Deviations, and Correlations Feasible Ideas
Global M SD 1 2 3 4
1. Feasibility rating experta 4.18 0.39
2. Creativity rating participantb 3.64 1.88 0-.31**
3. Good rating participantb 5.41 1.27 -0.06 -0.02
4. Relevance rating participantb 5.49 1.33 -0.11* -0.19** -0.67**
Local M SD 1 2 3 4
1. Feasibility rating experta 4.19 0.39
2. Creativity rating participantb 3.74 1.84 0-.13**
3. Good rating participantb 5.18 1.36 0-.05 -0.15**
4. Relevance rating participantb 5.43 1.39 -0.15** 0-.06 -0.51** Note. N = 924 feasible ideas (441 global, 483 local). ** p < .01, * p < .05
a
Measures refer to the average rating of the most feasible ideas by an expert. First, all ideas were scored on feasibility, scale 1-5. A median split was executed to divide the ideas in feasible and not feasible (median = 4). Because we were only interested in the most feasible ideas, ideas with score 1, 2 and 3 were removed from further analyses. Minimum value = 4, maximum value = 5.
b
Measures refer to average creativity, good and relevance ratings. Maximum value = 7.
Test of Hypotheses 1: Idea Rating
Hypotheses 1 predicted that the rating of ideas depends on the interaction between the kind of idea (original / feasible) and processing style. In all ANOVA analyses is corrected for mood and regulatory focus. Hypothesis 1a stated that participants induced with a global processing style rate original ideas as less creative compared to participants with a local processing style. A one way ANOVA with creativity ratings by participants as dependent variable and original ideas and processing style (global vs. local) as independent variables revealed no significant interaction effect (F(1,569) = .21, p = n.s.). Thus, there were no differences between the conditions with regard to the given creativity ratings, therefore hypothesis 1a cannot be confirmed.
and feasible ideas F(1,916) = 2.07, p = n.s.). Processing style did not moderate the relationship between kind of idea (original vs. feasible) and good ratings of participants. Opposed to what was expected, participants with a global processing style rate feasible ideas higher on good (M = 5.41, SD = 1.27) as did participants with a local processing style (M = 5.18, SD = 1.36; F(1,916) = 5.49, p = .02). Therefore, hypothesis 1B must be rejected.
Hypothesis 1c predicted that participants induced with a global processing style rate more original ideas as relevant, contrary participants with a local processing style rate more feasible ideas as relevant. H1c cannot be confirmed since two one way ANOVA’s revealed no significant interaction effects: original ideas F(1,569) = 1.47, p = n.s. and feasible ideas F(1,916) = .47, p = n.s.
MANOVA’s with the questionnaire items related to the criteria participants used for idea rating revealed no differences between the processing styles.
Test of Hypotheses 2: Idea Selection
The second hypotheses stated that the type of idea (original and/or feasible) selected after brainstorming depends on people’s processing style, although ANOVA shows no significant results (originality of the selected idea F(1,158) = 0.01, p = n.s.; feasibility of the selected idea F(1,158) = 1.37, p = n.s.), the direction of the means, in table 6, shows that the trade off between originality and feasibility is smaller for global processors compared to local processors. The difference between feasibility and originality for global processors is 2.29 and for local processors 2.46. Hypothesis 2a stated that participants induced with a global processing style select an idea that scores high on both dimensions, originality and feasibility. More specific there is a positive association between originality and feasibility for global processors. Contrary to this hypothesis, there is a significant negative correlation between originality and feasibility for global processors (r = -.542, p = >.01). In line with hypothesis 2b (participants induced with a local processing style choose an idea that scores high on feasibility and low on originality.), there is a negative significant correlation between feasibility and originality of the selected ideas by local processors (r = -.646, p = >.01). Therefore, hypothesis 2b is confirmed. So, you could argue that overall, people have a strong preference for feasibility over originality, but that this preference is less strong for participants induced with a global processing style.
TABLE 6
Means and Standard Deviations for Idea Selection
Measure Global Local
Originality score expert selected ideaa 1.61 (0.67) 1.60 (0.65) Feasibility score expert selected ideaa 3.90 (0.88) 4.06 (0.88) Note. N = 160 total (80 in both conditions).
a
Measures refer to average originality and feasibility scores of selected ideas rated by an expert. Maximum value = 5.
Supplementary Analyses
Idea rating. Although no specific hypotheses were formulated with mood and regulatory focus as moderators, some interesting significant relationships were found. As in previous research, participants with a high score on positive mood (7.9) rate original ideas higher on creativity (M = 6.33, SD = 1.03) as did participants with a low score on positive mood (2.3) (M = 5.80, SD = .45; F(1,569) = 3.89, p = .05). The importance of originality for people with a positive mood is also shown by the questionnaire items: they argue that it is more important that an idea is creative compared to relevant (r = .20, p = .01), excellent compared to relevant (r = .25, p > .001), and creative compared to excellent (r = .20, p > .001). Additionally they label a new idea quickly as ‘deviant’ (r = .19, p = .02). Analyses revealed that participants with a high score on positive mood (7.9) rate feasible ideas higher on good (M = 6.2, SD = 1.03) as did participants with a low score on positive mood (2.3) (M = 3.86, SD = 1.95; F(1,916) = 4.09, p = .04). And finally, participants scoring high on positive mood (7.9) rate feasible ideas higher on relevance (M = 6.30, SD = 1.34) as did participants with a low score on positive mood (2.3) (M = 4.14, SD = 2.34; F(1,916) = 5.89, p = .02).
Participants scoring high on negative mood (5.40) rate feasible ideas lower on relevance (M = 5.25, SD = 1.28) as did participants with a low score on negative mood (1.00) (M = 5.45, SD = 1.59; F(1,916) = 6.19, p = .01). Besides a significant effect was found for feasibility ratings of experts. Feasible ideas (score 5) are rated as more relevant by participants (M =5.82, SD = 1.38) than did less feasible ideas (score 4)1 (M = 5.37, SD = 1.34; F(1,916) = 14.19, p = > .001).
Only two items of regulatory focus revealed significant relationships. People with a prevention focus find it hard to rate ideas on the degree of excellence (r = .21, p > .001) and relevance (r = .16, p = .04).
Idea selection. The results of the questionnaire items show the effect of mood and regulatory focus on criteria used for idea selection. Participants with a positive mood found it important to choose the most feasible idea (r = .23, p > .001), the most surprising idea (r = .19, p = .02) and thought carefully about their choice (r = .25, p > .001). While participants with a negative mood found it important to choose the most ambitious idea (r = .19, p = .02) and surprising idea (r = .16, p = .04), but also to choose the most safe idea (r = .33, p > .001). Participants with a promotion focus thought it was important to choose the most useful idea (r = .16, p = .04) and excellent idea (r = .18, p = .03). There was a negative significant relation with ‘I made my choice as fast as possible’ (r = -.24, p > .001). As could be expected participants with a prevention focus thought it was important to choose the most safe idea (r = .26, p > .001).
DISCUSSION EXPERIMENT 2
This experiment was designed to get more insight into the cognitive processes associated with idea rating and subsequently idea selection. Because experiment 1 revealed no significant results with respect to idea selection, the processes before this actual selection were explored in more detail in this second experiment. It was predicted that global processors would rate original ideas as less creative, but as better and more relevant, opposed to local processors who rate original ideas as more creative, but feasible ideas as better and more relevant. However no significant results were found, implying that global and local processors do not differ in the creative, good and relevance ratings they give to original and feasible ideas.
increase in the practice session beforehand so that participants are already at the stage of generating unconventional and original ideas.
Another possible explanation why processing style did not have an effect on idea rating probably lies in the rating criteria used. An exhaustive literature review of Dean et al. (2006) showed that two broad dimensions are universally applied: novelty and usefulness. In the current experiment only three rating criteria were used: creative, good and relevant and no definitions of these criteria were given to the participants. Therefore it is possible that participants were quite effective in rating and selecting ideas according to their own definition of these criteria (Rietzschel et al., 2010). Interestingly, the exit questionnaire once again showed that participants prefer feasibility in the rating of ideas. There is a significant positive correlation between feasible ideas and relevance ratings of participants, but no relation between original ideas and relevance ratings of participants. Additionally, while good and relevance ratings were significant correlated, creative and relevance ratings were not. These results highlight that people associate feasibility with good and relevant, but originality not. The same pattern applies to idea selection as shown by hypothesis 2b: participants preferred feasibility over originality, although this preference for feasibility was less strong for global processors.
Finally, it is possible that participants were not aware of the importance of originality during idea selection, because in this study, they were simply instructed to select the idea they ultimately found the best. People perhaps perceived originality as irrelevant and may indeed only be able to select high quality ideas when explicitly instructed to take originality into account (see experiment 1 and other studies; Rietzschel, 2005, Rietzschel et al., 2009; Rietzschel et al., 2010)
GENERAL DISCUSSION Findings
Experiment 2 studied idea evaluation more fundamentally: processing style as moderator in the relationship between kind of idea and idea rating and subsequently idea selection.
Experiment 1 revealed that people are only able to select original ideas when explicitly instructed to do so, which is in accordance with previous empirical findings (Rietzschel, 2005; Rietzschel et al., 2009; Rietzschel et al., 2010). Both experiments show that people preferred feasibility over originality during idea rating and idea selection, independently of their processing style. Research of Blair and Mumford (2007) has shown that people are likely to reject ideas that are original, risky, or require detailed descriptions. People tend to like ideas consistent with social norms, easy to understand and bring desired outcomes immediately to a number of people.
Limitations and Future Research
Although I believe that the findings of the present experiments contribute to the brainstorming literature of idea selection, some limitations should be considered in drawing conclusions from the results. One possible limitation, which is applicable to any laboratory task, is the use of a student sample. Although these are often used in academic studies because they are easy to manage, the generalizability of the results to real-world creativity remains an open question (Blair & Mumford, 2007; West, 2002). On the other hand, a realistic topic was used, improving health, which is plausible to be a real-world problem. Although research has shown that people tend to be less creative with realistic and relevant problems (Dillon et al., 1972; Harari & Graham, 1975 in Rietzschel, 2005), there is limited added value in using unrealistic topics because in organizations idea generation and idea selection is also used to solve realistic problems.
selection according to their own standards. In other words, it possible that participants were, according to their own selection criteria, quite effective in rating and selecting ideas.
Third, in my studies, processing style could not explain people’s low selection effectiveness. Evidently, there are more important factors and conditions that possibly cause this effect on which will be elaborated below. Two individual differences variables – regulatory focus and mood – which were included as control variable showed in the supplementary analyses a significant moderating influence. The relationship between selection instruction and originality of selected ideas becomes even stronger when people are in a positive mood, have a promotion focus or score high on dominant regulatory focus. The same applies when the aggregated measure of originality and feasibility is the dependent variable. Research has already shown the beneficial effects of positive mood and promotion focus on creativity (e.g. Baas et al., 2011; Nijstad et al., 2010; De Dreu et al., 2010; Stroebe et al., 2010), but not explicitly on idea selection and in combination with selection instructions.
Another individual difference variable that has consistently shown to be detrimental for idea selection is risk aversion (Sternberg & Lubart, 1995; Puccio & Cabra, 2012). Risk aversion implies that people will tend to evaluate ideas, and supposedly reject ideas, that undue risk with regard to the outcomes (Mumford et al., 2006). Due to their novelty, highly original ideas limit the confidence that can be placed in forecasts. Consequently, people tend to discount the value and reject novel, and potentially risky ideas (Licuanan, et al., 2007; Blair & Mumford, 2007). Risk aversion of the participants in the two experiments was not measured. Therefore it is possible that processing style yielded not the expected effects, due to the stronger effects of risk aversion, i.e. participants induced with a global processing style but who are naturally risk-averse would not select original ideas, but instead feasible ideas. It would be worthwhile to examine this relationship in future research.
particular sub process but not for another. In other words variables may be in and of themselves necessary but not sufficient for creative outputs to emerge (De Dreu et al., 2010). An integrative framework which decomposes creativity into creative outputs and creative processes is the Dual Pathway to Creativity Model (DPCM; De Dreu, Baas & Nijstad, 2008; De Dreu et al., 2010; Nijstad et al., 2010;). The model assumes that there are two ways to achieve creativity: flexibility (switching to a different approach or considering a different perspective) and persistence (sustained and focused task-directed cognitive effort). The flexibility pathway is what Förster and Dannenberg (2010) refer to as global processing. Aligned with global processing the assumption is that broad thinking and flexibility in cognitive processing facilitates creative production. The persistence pathway is aligned with local processing, but based on another assumption: persistence is beneficial to creativity, but local processing detrimental (De Dreu et al., 2010). The assumptions of DPCM are tested only on studies that focused on ideation tasks (Nijstad et al., 2010). For further research it is recommended to study the effects of both, processing style and DPCM separately, on idea rating and selection but also their interrelatedness.
Conclusions and Practical Implications
REFERENCES
Amabile, T.M. (1983). The social psychology of creativity: a componential conceptualization. Journal of Personality and Social Psychology, 45, 357-376
Amabile, T. M. (1996). Creativity in context: Update to the social psychology of creativity. Boulder, CO: Westview Press.
Amabile, T.M., & Mueller, J.S. (2008). Studying creativity, its processes and its antecedents: An exploration of the componential theory of creativity. In: Handbook of Organizational Creativity. Eds. Jing Zhou and Christina Shalley. Mahwah, NJ: Lawrence Erlbaum Associates.
Baas, M., de Dreu, C.K.W., & Nijstad, B.A. (2008). A meta-analysis of 25 years of mood-creativity research: Hedonic tone, activation, or regulatory focus? Psychological Bulletin, 134, 779-806.
Baas, M., de Dreu, C.K.W., Nijstad, B.A. (2011). When prevetion promotes creativity: The role of mood, regulatory focus and regulatory closure. Journal of Personality and Social Psychology, 100, 794-809
Blair, C.S., & Mumford, M.D. (2007). Errors in idea evaluation: Preference for the unoriginal? Journal of Creative Behavior, 41(3): 197-222
Crowe, E., & Higgins, E. T. (1997). Regulatory focus and strategic inclinations: Promotion and prevention in decision making. Organizational Behavior and Human Decision Processes, 69, 117-132
Dean, D.L., Hender, J.M., Rodgers, & Santanen, E.L. (2006). Identifying quality, novel and creative ideas: constructs and scales for idea evaluation. Journal of the Association for Information Systems, 7(10): 646-699
de Dreu, C.K.W., Baas, M., & Giacomantonio, M. (2010). Processing modes and creativity: Why (not)? Psychological Inquiry, 21(3): 203-208
de Dreu, C.K.W., Baas, M., & Nijstad, B.A. (2008). Hedonic tone and activation in the mood-creativity link: Towards a dual pathway to mood-creativity model. Journal of Personality and Social Psychology, 94(5), 739-756
de Dreu, K.W., Giacomantonio, M., Shalvi, S., & Sligte, D.J. (2009). Getting stuck or stepping back: effects of obstacles and construal level in the negotiation of creative solutions. Journal of Experimental Social Psychology, 45(3), 542-548
de Dreu, C.K.W., & West, M.A. (2001). Minority dissent and team innovation: The importance of participation in decision making. Journal of Applied Psychology, 86(6): 1191-1201
