An examination of the impact of stimuli type and GSS structure on creativity: Brainstorming versus non-brainstorming techniques in a GSS environment

1

This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

This version of the referenced work is the post-print version of the article—it is NOT the final published version nor the corrected proofs. You may access the final published paper from the publisher’s website.

The current reference for this work is as follows:

J.M. Hender, D.L. Dean, T.L. Rodgers, and J. F. Nunamaker, Jr. (2002). "An Examination of the Impact of Stimuli Type and GSS Structure on Creativity: Brainstorming Versus Non-brainstorming Techniques in a GSS Environment," Journal of Management Information Systems, 18:4, spring, pp. 59-85.

My vita can be found at: http://marriottschool.byu.edu/directory/details?id=5305

2

An Examination of the Impact of Stimuli Type and GSS Structure on Creativity: Brainstorming Versus Non-brainstorming Techniques in a GSS Environment

Jillian M. Hender, Douglas L. Dean, Thomas L. Rodgers, and Jay F. Nunamaker, Jr.

Jillian M. Hender Tel: (44-1491) 571454 Henley Management College Greenlands, Henley-on-Thames Oxfordshire RG9 3AU UK Office: [email protected] Home: [email protected]

Douglas L. Dean Office (801) 378-1224 Fax (520) 378-5933 School of Accountancy and Information Systems 569 TNRB, PO Box 23081 Brigham Young University Provo, UT 84602-3081 [email protected] Thomas L. Rodgers Office (409) 845-3139 Fax (409) 845-5653 804 Merion Court, College Station TX 77845 [email protected] Jay F. Nunamaker, Jr. Office (520) 621-4475 Fax(520) 621-2433 Center for the Management of Information University of Arizona 430 McClelland Hall Tucson, AZ 85721-0108 [email protected]

3

Author Biographies: JILLIAN M. HENDER is a member of the Associate Faculty at Henley Management College,

UK. She received a B.Sc. in Physics from University College, London, and a Ph.D. and MBA from Henley Management College. She has worked in the computer industry as an analyst/programmer and consultant. Her research interests include creativity, innovation management, teams, group support systems, and information technology for collaborative work.

DOUGLAS L. DEAN is an Assistant Professor at the School of Accountancy and Information

Systems at Brigham Young University. He received his Ph.D. in MIS from the University of Arizona in 1995, and a Master of Accountancy with an emphasis in information systems from Brigham Young University in 1989. His research interests include creativity, collaborative tools and methods, and requirements analysis. His work has been published in Management Science, Journal of Management Information Systems, Group Decision and Negotiation, and IEEE Transactions on Systems, Man, and Cybernetics.

THOMAS L. RODGERS is an Assistant Professor at the Lowry Mays College and Graduate

School of Business at Texas A&M University. He received his Ph.D. in Management from the University of Arizona in 1999, and a Master of Science in Finance from Colorado State University in 1994. His research interests include creativity, team aspects of software engineering, and expertise knowledge transfer.

JAY F. NUNAMAKER, JR., Regents and Soldwedel Professor of MIS, Computer Science and

Communication, is director of the Center for the Management of Information, University of Arizona, Tucson. In 1996, Dr. Nunamaker received the DPMA EDSIG Distinguished IS Educator Award. The GroupSystems software resulting from his research received the Editor’s Choice Award from PC Magazine, June 14, 1994. At the GroupWare 1993 conference, he received the GroupWare Achievement Award along with recognition of GroupSystems as best of show in the GDSS category. In 1992, he received the Arthur Anderson consulting Professor of the Year Award. Dr. Nunamaker received his Ph.D. in systems engineering and operations research from Case Institute of Technology, an M.S. and B.S. in engineering from the University of Pittsburgh and a B.S. from Carnegie Mellon University.

Keywords: Creativity, idea generation, brainstorming, assumption reversal, analogy, group support system, idea quantity, idea quality, laboratory experiment

ISRL categories: AF0802, HA0301, AA0903, AI0701, AI0701, AA0301, GB0404, AD05, DC0701, EI0102, AD0515, EI0102, A104

Acknowledgements: An earlier version of this paper appeared in the Proceedings of the 34th Annual HICSS, January 2001. This research was supported by a grant from the Economic and Social Research Council, UK.

4

Abstract

Of the techniques available for idea generation with Group Support Systems (GSS), little

research attention has been given to techniques that challenge problem assumptions or that use

unrelated stimuli to promote creativity. When implementing such techniques with GSS, choices

must be made regarding how to configure the GSS to deploy the initial creative stimuli and to

present the pool of emerging ideas that act as additional stimuli.

This paper reports the results of an experiment that compares Electronic Brainstorming (few

unnamed rotating dialogues) with Assumption Reversals (many related stimuli, many named

dialogues, free movement among dialogues) and Analogies (many unrelated stimuli, many

named dialogues, free movement among dialogues).

Analogies produced creative, but fewer, ideas, due to use of unrelated stimuli. Assumption

Reversals produced the most, but less creative, ideas, possibly due to fragmentation of the group

memory and cognitive inertia caused by lack of forced movement among dialogues.

Introduction

Creativity is essential in the quest for competitive advantage in today’s world of quickly

changing technologies and dynamic competitors. Notwithstanding the obvious importance of

creativity, however, people are often poor idea generators. Substantial literature suggests that

when facing large, complex problems, people tend to think within a narrow, bounded subset of

the possible solution space rather than thinking creatively [7, 8, 32, 53]. When facing complex

problems, people can overlook as much eighty percent of the solution space and not even be

aware that they are doing so [21]. Santanen et al. [50] have developed a Cognitive Network

Theory (CNT), grounded in mechanisms of human cognition, that explain why this may be so.

5

The model indicates that human memory is organized as associations among related items or

concepts. These bundles of associated, called frames, are linked together in a cognitive network.

Frames are activated by stimuli, and limitations of human memory mean that a particular

stimulus tends to consistently activate the same series of frames. This may help explain why

problem solvers often fail to consider large portions of the solution space. Limited, conventional

stimuli tend to activate familiar patterns of associations resulting in people staying within

bounded, familiar areas of their cognitive networks. Many creative techniques are used to

introduce different types of stimuli with the intent to compensate for this limitation. These

stimuli act as fresh entry points into a person’s cognitive network. This activation of additional

frames may trigger production of ideas that otherwise would not be produced.

With the most commonly used technique, Brainstorming, the problem statement and the

ideas contributed by other group members act as the creative stimuli. Participants are encouraged

to “piggyback” on other ideas within the emerging pool of ideas. Because each person has

developed a different set of associations, a single stimulus often activates different associations

across individuals, hence exposure to others’ ideas activates additional frames.

Group Support Systems (GSS) have been used in organizations for more than a decade

because they are particularly well suited to idea generation [40] and collection. GSS enable

groups to interact electronically by allowing participants to contribute anonymous, typed ideas in

parallel, thus avoiding wasted group time that occurs in manual groups where only one person

can speak at time [16]. Much research has been published--both laboratory experiments [19, 20,

43] and field studies [39]--showing that Electronic Brainstorming (EBS) can be used to produce

ideas of a higher quantity and quality than verbal Brainstorming. With EBS, parallel

6

communication and anonymity lead to a faster growing pool of ideas, which act as secondary

stimuli for production of further ideas.

With both verbal Brainstorming and EBS, the stimuli are the problem statement and the

emerging pool of ideas. Recent research has found, however, that introducing additional stimuli

in addition to the problem statement and emerging pool of ideas can further increase creativity.

Stimuli that deliberately direct the thinkers’ attention to different parts of the solution space can

increase productivity on a number of measures. For example, Santanen et al. [50] found that

providing verbal and visual cues (questions derived from criteria for effective solutions) resulted

in the production of more unique ideas, and a higher concentration of unique ideas, than those

produced by traditional EBS. Dennis et al. [15] found that presenting three sub-problems instead

of a combined intact problem resulted in an increase in three combined quantity/quality

measures: number of unique ideas, total quality, and number of good ideas. However, no

differences were found in mean quality.

EBS is based on the creative technique of Brainstorming. However, many other techniques

are detailed in the literature [9, 54] that researchers have argued may increase creativity further

by means of different types of stimuli. This research extends previous research by comparing

two of these, Assumption Reversals and Analogies, with EBS. EBS was used as the baseline

since Brainstorming is by far the most common form of ideation used by practitioners, and has

been used in a substantial proportion of idea generation research studies, both in a GSS [11, 12,

30, 31, 44] and manual environment [28, 34].

In Assumption Reversals, participants list all the major assumptions about the problem,

reverse them, and use the reversed assumptions as stimuli to generate ideas. Because

assumptions form an artificial boundary around the problem, reversing the assumptions is

7

deliberately designed to break existing thought patterns. Cognitive Network Theory suggests that

if activation of frames is done in such a way that previously remote concepts or frames are

brought together (associated) in short-term memory, these uncommon associations will result in

more creative ideas. Assumptions are traditional associations within and across frames, which

may include cause-effect relationships and problem constraints. These limiting associations may

constrain the thinker to search only narrow, bounded aspects of the solution space. Methods that

disrupt and confront these traditional associations may therefore increase creativity by forcing

the thinker to access more uncommon associations.

Both Brainstorming and Assumption Reversals use stimuli that are related to the problem.

Authors [5, 17, 23, 35, 55] have argued that another category of techniques has the potential to

increase creativity beyond techniques that use related stimuli. These techniques, which include

Analogies, introduce unrelated stimuli, which are related back to the problem using forced

relationships. According to Cognitive Network Theory, forcing associations between remote

aspects of the cognitive network should result in the production of very creative ideas.

The next section provides additional theoretical background and motivation for the research

by using Cognitive Network Theory [50] to examine how the structures of both creative

techniques and GSS affect the quantity and creativity of ideas, culminating in the research

hypotheses. Next, the research methods are presented. The results and discussion then follow,

and the paper concludes with suggestions for future research.

8

Background

Definition of Creativity

Creativity could be defined in terms of an attribute of a person, a quality of a product, or as a

process [25]. For the purpose of this study we define creativity in terms of a creative process,

conducted by persons, the outcome of which can be measured in the form of ideas. This research

examines characteristics of creative processes because of the importance of finding useful

mechanisms to help people produce creative ideas.

The Cognitive Network Theory

In order to understand the means by which a creative idea may be generated, Santanen,

Briggs, and de Vreede [50] have formalized insights from cognitive psychology into the

Cognitive Network Theory (CNT).

CNT is based on the assumption that human memory exists in the mind as bundles of related

meanings. These bundles are called frames. Frames result from human experience as different

items and concepts become associated and thus become encoded as memory. Association among

concepts within frames can be organized according to a variety of principles, including, but not

limited to, perceptions of cause and effect, constraints associated with possible courses of action,

time sequence of events [52], and similarity or typicality of items [48, 57]. Every day we

perceive many stimuli from multiple sources. Since stimuli can be associated in many ways, the

same item may be a member of many frames. An item that exists in multiple different frames

may serve as a link among the frames. In other words, each frame is a node in the knowledge

web, and each concept in a frame is a link to the other frames of which it is a member. Thus,

9

frames are linked into a network or web of related meanings through concepts common to

multiple frames.

Knowledge webs are vast networks that represent our knowledge and experience. Due to the

sheer size of the networks, it is possible to manipulate only a small number of frames at any one

time. Such manipulation occurs in short-term memory, which can be thought of as the workspace

for information that is under active consideration at the moment [18]. Frames currently active in

short-term memory are referred to as salient. Stimuli cause frames to be activated. Activation is

the process by which a particular frame becomes salient. By traversing the links that connect

some activated frame to other frames within our knowledge networks, successive frames from

memory become activated. When two or more frames are simultaneously salient they are said to

be associated. As a result of different associations within different people’s knowledge webs, the

same stimulus may activate very different frames, that is, bundles of associations, for different

people.

There are two patterns of activation among frames. The first, an automatic spreading

activation, occurs without intention or conscious awareness [38]. If a particular stimulus

automatically and consistently activates the same sequence of frames on all occasions, this may

explain why problem solvers often fail to consider large areas of the solution space and think

primarily within bounded and familiar areas of their knowledge networks. The second is a

conscious, limited-capacity, spreading activation that depends on the context of the stimulus and

requires intention and conscious awareness. This suggests that perhaps an intervention that

provides a variety of stimuli from different contexts might activate new or different areas of our

knowledge networks.

10

Many researchers assert that the process of making new and unexpected associations

between previously unrelated frames often leads to the formation of highly creative solutions to

solve problems. For example, Mednick [32] defined the creative thinking process as "the forming

of associative elements into new combinations which either meet specified requirements or are in

some way useful. The more mutually remote the elements of the new combination, the more

creative the process or solution" (p. 221). Thus, according to Cognitive Network Theory,

Proposition 1: A creative idea is caused by the unexpected juxtaposition in working memory

of concepts that were previously remote from one another in the cognitive web. The more remote

the concepts used in the juxtaposition, the more creative the idea.

The working space within short-term memory, however, is relatively small. Humans are

typically capable of maintaining seven (plus or minus two) items in short-term memory at a

given moment [33]. Cognitive loads escalate rapidly as thinkers try to push beyond this limit. To

overcome this limit, people cluster related frames into more abstract frames or chunks [33], and

wield those bigger chunks as frames in short-term memory. The cognitive load of accessing

concepts within a given salient frame is low. But, there are cognitive costs associated with

swapping concepts in and out of short-term memory. It takes a great deal of cognitive effort to

bring concepts to mind that are distant on the cognitive web from salient concepts [18]. Thus,

Proposition 2: The cognitive load of traversing among frames in the cognitive web increases

in relation to how far apart they are on the cognitive web.

Attributes of Creative Techniques

All group creative techniques begin with the problem statement and use shared visibility of

the evolving idea pool as creative stimuli. Visibility of previously entered ideas supports

production of additional ideas since they provide entry points into the unique cognitive network

11

possessed by each group member. There are four categories of techniques that introduce stimuli

in different ways: 1) traditional brainstorming, 2) techniques that introduce additional related

stimuli in the form of prompts, 3) techniques that introduce additional related stimuli by the

exploration of associations, and 4) techniques that introduce additional stimuli that are unrelated

to the problem.

Brainstorming is the most basic and frequently used technique. It provides only the problem

statement itself and the emerging idea pool as stimuli. The second category of techniques

provides stimuli in the form of multiple questions that deliberately direct the attention of thinkers

to different facets of the solution space. By presenting the problem-solver with multiple starting

categories, different areas of their cognitive networks may be accessed. This avoids the narrow

activation patterns that occur when a single question is provided. Multiple questions help

participants more easily find areas of their cognitive webs where they can generate solutions,

resulting in the generation of more ideas, and possibly more creative ideas. Examples of these

techniques that have been implemented with GSS are Problem Decomposition [15] and Directed

Brainstorming [50]. Dennis et al. [15] found that groups who spent fifteen minutes on three sub-

problems produced more ideas and more good ideas than groups who spent forty-five minutes on

an intact problem. Santanen et al. [50] found that participants using Directed Brainstorming

produced more unique solutions than participants using traditional Brainstorming. In Directed

Brainstorming, participants were presented with a sequence of thirty six questions that reflected

effective solution criteria, such as “now give me a solution that is inexpensive to implement,”

and “now give me a solution that will catch the enemy by surprise,” etc.

A third category of techniques stimulates creativity by having thinkers explore their existing

associations between the problem and its causes, assumptions, attributes, etc. Examples of these

12

techniques are Causal Thinking [45], Assumption Reversals, and Attribute Listing [54]. Potter

[45] found that giving participants time to think of possible causes of the problem prior to

Brainstorming increased the number of ideas generated. Methods that explore associations can

be used to trigger activations regarding aspects of the problem. Assumption Reversals is one of

these methods that purposely attempts to access non-traditional associations. This method has

been used by practitioners both manually [22, 47, 55] and with a Group Support System [13], but

has not been evaluated in controlled experiments.

Phases of the Assumption Reversal idea generation technique are as follows:

1. List all the major assumptions about the problem.

2. Reverse each assumption in any way possible.

3. Using the reversals as stimuli, generate ideas.

Since assumptions stem from traditional associations about a problem, reversing them to

introduce non-traditional assumptions may increase creativity by forcing the thinker to activate

frames not traditionally associated with the problem. For example, when considering the problem

of improving restaurants, an assumption about restaurants may be “restaurants serve food.” By

reversing this assumption to read “restaurants do not serve food,” a different set of associations

may be activated, such as between restaurants and entertainment, restaurants and shopping, or

restaurants and social events.

The previous three categories of techniques use stimuli that are related to the problem. The

ideas are generated by free association, which occurs when participants follow a train of thought

and rely largely on chance and incubation [54]. In other words, free association allows people to

follow the automatic spreading of activations among frames that occurs without intention or

conscious awareness.

13

Researchers and practitioners have suggested that a fourth category of techniques that use

unrelated stimuli may increase creativity beyond related stimuli techniques by introducing

concepts during the creative process that are unrelated to the problem. These unrelated concepts

are then linked back to the problem as stimuli for idea generation. For example, participants may

be asked to forget about the problem, look at a picture or object that is totally unrelated to the

problem, and describe it in detail. Next, thinkers are asked to use these details, which are

unrelated to the problem, as stimuli to generate solutions for the problem. The forcing of

unrelated stimuli back to the original problem is called forced relationships. VanGundy [54] (p.

75) describes forced relationships as the “forcing together of two or more objects, products, or

ideas to produce new objects, products, or ideas.” The Analogies method, used in the current

study, is an example of this approach. An analogy is a statement of similarity, that is, drawing a

comparison regarding how objects, persons, situations, or actions are similar in process or

relationship to one another. The elicitation of analogy details that are unrelated to the problem is

followed by use of those unrelated details as stimuli to generate solutions for the problem.

Phases of the Analogies idea generation technique are as follows:

1. Decide the major principle represented by the problem.

2. Use the major principle to generate a list of analogies that are similar in concept.

3. Select any of the analogies that look interesting and describe each in detail. Elaborate

on the analogy by listing details, such as parts, functions or uses. While completing

this step, try to forget about the problem.

4. Force fit the analogy descriptions back to the original problem in order to suggest

ideas for solving the problem.

14

Many authors have emphasized the importance of analogy in creativity [4, 5, 17, 27, 42].

The method has been used both as a stand-alone technique, and as a part of formal techniques,

such as Synectics [46, 54]. The intent of the Analogies method and other similar methods [23,

35, 55] is to activate remote concepts in the cognitive web. Forced relationships force the

juxtaposition of these concepts in short-term memory, and if these concepts were previously

remote, then, according to Proposition 1, creative ideas will result. This supports the assertion,

made by many researchers and practitioners [35, 54], that techniques which use unrelated stimuli

and forced relationships will more likely produce novel ideas than techniques which use related

stimuli and free association.

However, the creative benefits of unrelated stimuli and force-fitting may come at the cost of

fewer ideas being produced. Techniques that use unrelated stimuli and forced relationships force

the juxtaposition of formally remote concepts in short-term memory. According to Proposition 2,

traversing among frames that are remote in the cognitive web causes considerable cognitive load.

Increased cognitive load should increase the time required to generate each idea, and therefore

reduce the number of ideas generated in a given time period. This proposition has not been

specifically investigated in past research, although the result of one study [23] in a manual

environment found that the Guided Fantasy technique, which uses unrelated stimuli and forced

relationships, produced more creative but fewer ideas than Brainstorming.

Attributes of Group Support Systems

GSS structures can be used to record ideas and to present other creative stimuli. The ability

to store previously generated ideas so that they can be made available to participants has been

termed group memory [40]. The ability to configure group memory within GSS makes it

possible to manipulate the exposure of each participant to the ideas generated by the other

15

participants in a variety of ways. When creative techniques are adapted for use with GSS, the

way that group memory is configured may vary for different techniques. Therefore, it is

necessary to examine the possible effects of these configurations on creative outcomes.

Group memory may be configured in two ways: by partitioning it into one or many

dialogues used to collect ideas, or by manipulating movement among the dialogues. Both

partitioning and movement are determined by the GSS tool used and the configuration of that

tool [6]. For example when using the EBS tool within GroupSystems, N+1 unnamed dialogues

are often used, where “N” is the number of participants. This tool supports automatic rotation of

dialogues; that is, once a participant submits an idea, the tool automatically presents him or her

with a different dialogue that shows a different portion of the idea pool. Other GSS tool

configurations may be used that offer a flexible number of dialogues and free, as opposed to

forced, movement between the dialogues.

The number of dialogues chosen and forced versus free movement can effect creative

outcomes. Two studies [1, 14] have shown that groups using multiple (N+1) dialogues and

forced movement generate more ideas than groups using a single dialogue. The latter study also

showed that this configuration led to an increase in both the number of good ideas, and the

number of novel ideas.

A single dialogue may focus on only a few narrow themes and thus miss addressing large

portions of the overall problem space that may contain many useful ideas. Multiple dialogues, on

the other hand, may focus on different themes--either because they started that way or evolved

that way--and are therefore more likely to provide a variety of stimuli that explore larger portions

of the overall solution space. Each dialogue may activate different parts of the cognitive web,

and thus lead to the generation of more, and more creative, ideas.

16

Another way to stimulate creativity through the use of multiple dialogues is to manifest a

creative stimulus as a name on each dialogue. Providing this type of stimuli may cause the

thinker to explore associations or to perceive stimuli that may not occur if only one or few

unnamed dialogues are used. In the present study, we used multiple named dialogues to present

thinkers with reversed assumptions and analogy details as stimuli.

Free movement versus forced movement among dialogues may also impact creative

outcomes. Free movement among dialogues allows a thinker to choose among a variety of

stimuli during idea generation. A thinker is thus able to focus on stimuli that are most interesting

or that seem to provide the most useful stimuli. However, this freedom can come at the cost of

paying attention to a narrower segment of stimuli. Free movement among dialogues may result

in cognitive inertia if individuals follow a train of thought as far as it goes, using it as a basis for

generating ideas, before attending to the ideas generated by other participants in different

dialogues. As Dennis et al. [14] (p. 205) note,

There is always the possibility that groups will not actively use multiple dialogues. Given a

choice about how many dialogues to use, and how many ideas to contribute to one dialogue

before moving to a new one, group members might tend to use their ‘traditional’ structures.

For example, if several dialogues were available to the groups, members might choose to

focus on only one or two dialogues and ignore others because this is cognitively simpler, or

because it is closer to learned social behaviors.

Cognitive inertia can lead to narrow spreading activation patterns and exploration of narrow

subsets of the problem solution space--that is, people may become stuck in a particular pattern of

thought and be unable to break free.

17

In contrast, forced movement between dialogues may promote production of creative ideas

for three reasons. First, force movement forces thinkers to see changing segments of group

memory, thus reducing the tendency toward cognitive inertia. Second, different dialogues may

focus on different topics or categories, leading participants to explore different areas of their

knowledge networks. Third, when a new dialogue is presented, ideas from the previous dialogue

may still be in a person’s short-term memory, and these may combine with items in the new

dialogue. Therefore, in accordance with Proposition 1, the likelihood of the unexpected

juxtaposition in short-term memory of two concepts that were previously distant from one

another on the cognitive web will be greater, and hence the likelihood of producing creative

ideas will be greater.

All three treatments evaluated in this study used multiple dialogues. The Assumption

reversals and Analogies treatments provided stimuli through many named dialogues. The EBS

method used N+1 unnamed dialogues with rotation.

Creative Technique Implementations

Three creative techniques were included in the present study: Brainstorming, Assumption

Reversals, and Analogies. GroupSystems for Windows by GroupSystems.com Corporation was

the software used to implement each of the three creative techniques. Tools and tool

configurations within GroupSystems were selected based on how well they supported each

creative technique.

Brainstorming

The Electronic Brainstorming (EBS) tool, which uses N+1 rotating dialogues, was used to

implement the Brainstorming treatment. Both Assumption Reversals and Analogies require a

18

tool that supports multiple named dialogues. Hence, for the purposes of comparison, EBS was

chosen since it supports the use of multiple dialogues, and has been used in the vast majority of

past research.

Assumption Reversals

The Categorizer tool was used to implement the Assumption Reversals treatment because it

supports named idea collection topics. Participants list the major assumptions about the problem

as topics in the tool. The technographer then, with input from participants, reverses the

assumptions. The text of each reversed assumption is visible as a node in the overall list of

reversed assumptions. Using the reversed assumptions as stimuli, participants then submit ideas

to solve the problem as comments behind a chosen node. Unlike the EBS tool, the Categorizer

tool does not support automatic rotation of dialogues upon idea submission, so participants were

free to move at will among the various dialogues.

Analogies

The Categorizer tool was also used to implement the Analogies treatment because it

supports the naming of idea collection topics and the multiple levels of hierarchy required to

implement this creative method. First, Categorizer allows participants to type in a list of

analogies as topics. The technographer then moves the analogies into category buckets within

which participants contribute short descriptions or details about the analogy, while forgetting

about the initial problem. Although the analogies are related to the problem, the analogy details,

by design, are unrelated to the problem. Participants then use these details as stimuli to suggest

ideas for solving the problem--that is, they force fit them back to the problem. They may choose

any detail about any analogy as a stimulus and enter the ideas as comments behind the detail.

19

Each analogy detail was visible as a named “dialogue,” in a list of dialogues. As with

Assumption Reversals, participants were allowed to move freely among the dialogues in order to

generate ideas.

Hypotheses

Creativity

Proposition 1: A creative idea is caused by the unexpected juxtaposition in working

memory of concepts that were previously remote from one another in the cognitive web. The

more remote the concepts used in the juxtaposition, the more creative the idea.

Analogies versus Assumption Reversals

Both Analogies and Assumption Reversals use many named dialogues and free, as opposed

to forced, movement between dialogues. Therefore any differences will be due to creative

technique effects only, specifically differences in type of stimuli. Assumption Reversals uses

related stimuli in the form of reversed assumptions to deliberately break common associations

with the aim of forming more uncommon ones. However, ideas are generated by free

association, which typically allows people to follow spreading-activations of their cognitive

webs, which is less likely to create unexpected juxtaposition of previously remote concepts in the

cognitive web. On the other hand, techniques based on unrelated stimuli and forced relationships,

such as Analogies, will very likely create unexpected juxtapositions, which give rise to creative

ideas, since they have mechanisms that deliberately bring remote concepts into short-term

memory and deliberately force them together. Therefore,

20

H1a: The average creativity score of the ideas generated by participants using the

Analogies technique will be higher than the average creativity score of the ideas

generated by participants using the Assumption Reversals technique.

Brainstorming versus Assumption Reversals

Creative Technique Effect: Apart from the pool of ideas generated, Brainstorming has only

the problem statement as a stimulus, whereas Assumption Reversals makes many reversed

assumptions available as stimuli. When the Assumption Reversals technique is implemented

using a GSS, the reversed assumptions will be shown as dialogue labels.

GSS Effect: Assumption Reversals, however, uses free movement between dialogues. In

contrast, Brainstorming, by using forced movement between dialogues, may interrupt any

tendency toward cognitive inertia and may force the combination of somewhat remote stimuli.

It was not known a priori which of the two effects would be stronger, but when designing

the study we assumed that the stimuli manipulation from the creative techniques would dominate

the GSS effects. Therefore,

H1b: The average creativity score of the ideas generated by participants using the

Assumption Reversals technique will be higher than the average creativity score of the

ideas generated by participants using the Brainstorming technique.

Analogies versus Brainstorming

Creative Technique Effect: Apart from the pool of ideas generated, Brainstorming has only

the problem statement as a stimulus, whereas Analogies has many analogy details as stimuli.

21

Participants are instructed to force these unrelated stimuli back to the problem in order to

generate ideas. In contrast, Brainstorming has no structure or procedure for forcing together

remote concepts. Hence, Analogies should produce more creative ideas than Brainstorming.

GSS Effect: Alternatively, Brainstorming, using forced movement between dialogues, could

produce more creative ideas than Analogies, which uses free movement. However, the forced

movement for Brainstorming would still only expose participants to somewhat remote but related

stimuli that would only be combined by chance, whereas with Analogies there is deliberate

exposure to very remote, unrelated, stimuli and a deliberate forcing together of the remote

stimuli with the problem. Therefore,

H1c: The average creativity score of the ideas generated by participants using the

Analogies technique will be higher than the average creativity score of the ideas

generated by participants using the Brainstorming technique.

[Table 1 here]

Quantity

Proposition 2: The cognitive load of traversing the cognitive web from one frame to

another is a function of the degree of separation of the frames within the cognitive web.

Increased cognitive processing should increase the time required to generate each idea, and

therefore reduce the number of ideas generated in a given period of time [37]. Therefore,

according to Table 1,

H2a: Participants using the Assumption Reversals technique will generate more ideas

than participants using the Analogies technique.

22

H2b: Participants using the Brainstorming technique will generate more ideas than

participants using the Analogies technique.

H2c: Participants using the Brainstorming technique will generate more ideas than

participants using the Assumption Reversals technique.

Methodology

Research Design

In order to test the hypotheses, a 3 x 1 single-factor experimental design was used. The

design included Creative Technique, as implemented in the GSS, as the independent variable,

and Quantity and Creativity of Ideas as dependent variables. There were twenty-seven groups in

all, nine groups in each treatment, and five participants in each group. Participants were

randomly assigned to groups, and groups were randomly assigned to different treatments. The

same GSS classroom was used for all sessions. In order to control for facilitator effects, the

treatments and subject instructions carefully followed a predefined script. Treatments were

administered to groups, but the unit of analysis for this study was the individual.

One hundred and thirty-five undergraduate students, enrolled in an introduction to business-

computer-systems class at a large United States university, participated in the experiment. A pre-

session questionnaire was given to each of the participants in order to gather demographic and

background data on the subjects and to determine whether any systematic biases had occurred in

their assignments to the three treatments. Since individuals vary in their ability to generate ideas

[29, 56], and this may confound results in idea generation experiments, it was necessary to

prevent systematic bias of ideational fluency ability for participants within treatment groups. A

23

Productive Thinking Test (PTT) [24], used to measure ideational fluency, was administered to

confirm that there were no differences across the treatment conditions.

Participants were asked to provide creative ideas for the following problem: “A restaurant

located next to campus is losing customers. What can the restaurant do to retain its customers?”

To support analysis of production by each participant, and yet retain a measure of

anonymity, a unique Participant Identification Number (PIN) was randomly assigned to each

participant. This PIN was entered into GroupSystems by each participant prior to idea generation

and was also entered on the questionnaires. This allowed ideas to be tracked to the contributor

without making the contributor’s name visible to other participants.

Having completed the Pre-session Survey and the PTT, participants were then provided with

a printed description of the technique they were going to use, as well as a scripted explanation of

the technique. A warm-up session, using GroupSystems, was then held to allow the participants

to become accustomed to both the technique and GroupSystems. Next, the restaurant problem

was distributed, and the session began.

The time allowed for actual idea generation was twenty minutes. However, differences in

time spent prior to idea generation existed due to inherent procedural differences in the three

techniques. The Assumption Reversals technique involved the listing of assumptions about the

problem, then the reversing of these assumptions, followed by idea generation. Five minutes

were allowed for the listing of assumptions, and twenty minutes for idea generation. The

Analogies technique involved the generation of analogies, the listing of details about the

analogies, followed by idea generation. Five minutes were allowed for the generation of the

analogies, five minutes for the listing of details, and twenty minutes for idea generation. The

24

facilitator controlled the process carefully. This procedure is similar to that used by other

researchers [23, 35, 45]. When the time was up, participants completed a Post-session Survey.

There was no intervention by the facilitator except in the following areas: For Assumption

Reversals, the technographer deleted duplicated assumptions and items that were not

assumptions (with the agreement of the participants). For Analogies, the facilitator suggested to

the participants the major principle represented by the problem, although they were free to use

other principles if they wished. For example, in the restaurant problem (“How can the restaurant

retain customers?”), a major principle was “retention.” The facilitator also deleted duplicate

analogies, items that were not analogies, and analogies that had no details (with the agreement of

the participants).

The specific instructions given to the participants are detailed in Appendix A.

Operationalization of Dependent Variables

The following sections describe the operationalization of the dependent variables, namely

idea quantity and idea creativity. Ease of use was also assessed for all three methods because of

the hypothesized relationship between cognitive load and quantity.

Quantity of Ideas

Idea quantity was determined by counting the number of ideas generated by each

participant. This was obtained from the computer transcripts. Ideas generated by participants

were identifiable by the subjects PIN. Non-ideas, including extraneous comments and

incompletely expressed thoughts, were excluded. If a participant produced one statement

containing a list of ideas, this was disaggregated and the ideas counted individually.

25

Creativity of Ideas

An evaluation scheme for measuring creativity was developed from the creativity literature.

Nagasundaram and Bostrom [36] advocate evaluating creative ideas according to both creativity

level and creativity style. Creativity level was measured in terms of the originality of the idea.

Creativity style was measured in terms of paradigm relatedness, which represents the degree to

which an idea reflects the currently prevailing paradigm. Several other authors have recognized

these two subdimensions. According to Besemer and Treffinger [3], a novel product must be

both original (the product is unusual or is infrequently seen) and transformational (the product is

so revolutionary that it forces a shift in the way that reality is perceived). Similarly, Jackson and

Messick [26] argue that the criteria for creative products include unusualness and transformation

power. Besemer and O'Quin [2] break novelty down into the subdimensions of original (defined

as novel, unusual, unique, original, ingenious) and germinal (defined as trend-setting, influential,

revolutionary, radical). Idea creativity was therefore decomposed into the following

subdimensions:

• Originality: An idea is most original if no one has expressed it before. Originality was

judged from the perspective of the idea rater; that is, it was assessed according to the

ideas in the mindset of the rater, not according to the sample of the ideas being rated.

• Paradigm relatedness: The degree to which an idea preserves or modifies a paradigm.

In order to determine the creativity of ideas, each idea was scored independently by two

raters using Likert scales on the two subdimensions. Scoring definitions were initially developed

from the literature. Two raters scored a sample of ideas on each of the subdimensions.

Correlations were obtained using Pearson’s correlation coefficient. Differences were discussed,

and the definitions refined for each subdimension. The subdimension score was derived by

26

taking an average of the two raters’ scores. Scores for creativity were calculated by aggregating

the subdimensions. The process was repeated until the interrater reliability was great than 0.7.

The scores were standardized on a scale of 1-7.

Total creativity scores for each participant were calculated by summing the creativity scores

for each idea generated by that person. Since total creativity is correlated with quantity, and the

hypotheses assume that quantity and creativity are not correlated, quantity itself also being a

dependent variable, average creativity, which Lamm and Trommsdorff [28] describe as a "purer"

measure, was used as the creativity measure. Average creativity scores for each participant were

calculated by dividing the total creativity score by the number of ideas generated by that person.

Ease of Use

Perceived Ease of Use was measured as a way to assess the cognitive load of each method.

Perceived Ease of Use was measured using an adaptation of an instrument developed by

Sambamurthy and Chin [49]. Perceived Ease of Use was measured via five 7-point Likert-scale

questions, where 1 = strongly disagree and 7 = strongly agree. The responses to the questions

were summed to calculate the perceived Ease of Use score. The reliability of the questionnaire

was acceptable in the current experiment (Cronbach’s alpha = 0.85). The questionnaire is

presented in Appendix B.

Results

Neither of the dependent variables, Quantity and Creativity, was normally distributed.

Therefore, the Kruskal-Wallis test was used to test for significant differences among the three

treatments. Where this test indicated significant differences, the Mann-Whitney (MW) Mean

27

Rank test was used to test hypotheses H1a, H1b, H1c, H2a, H2b and H2c. Table 2 presents the

results of the treatment comparisons.

[Table 2 here]

The results of the tests of the hypotheses are summarized in Table 3.

[Table 3 here]

Creativity of Ideas

Analogies produced ideas that were significantly more creative than those produced by

Assumption Reversals. Thus H1a was supported. Brainstorming produced ideas that were

significantly more creative than those produced by Assumption Reversals. Thus H1b was not

supported. Although the ranked mean of Analogies was higher than that of Brainstorming, it was

not significant at the 0.05 level (p= 0.144). Thus H1c received only limited support.

Quantity of Ideas

Assumption Reversals produced 837 ideas, Brainstorming produced 696 ideas, and

Analogies produced 572 ideas, making a total of 2,105 ideas. There was a significant difference

between the three treatments. Both Assumption Reversals and Brainstorming produced

significantly more ideas than Analogies. Thus H1a and H1b were supported. There was no

significant difference between the quantity of ideas produced by Assumption Reversals and

Brainstorming at the 0.05 level. Thus H2c was not supported.

28

Discussion

Creativity of Ideas

Figure 4 shows the creative technique effects, namely differences in type, number and

presentation of stimuli; and the GSS effects, namely the partitioning and manipulation of group

memory, for the three techniques.

[Table 4 here]

As expected, Analogies produced ideas that were significantly more creative than those

produced by Assumption Reversals. Force-fitting unrelated stimuli to the problem task had a

greater creative effect than the related stimuli used by the reversed assumptions.

Brainstorming also produced ideas that were significantly more creative than those produced

by Assumption Reversals. Surprisingly, the GSS effect of Brainstorming, with forced movement

among fewer dialogues, was greater than the creative effect of the Assumption Reversals

technique, namely many related stimuli in the form of reversed assumptions, presented as

dialogue labels. Ideas produced by Analogies were the most creative, but not significantly more

creative than those produced by Brainstorming at the 0.05 level (p = 0.144). It was predicted that

the creative effect of the Analogies technique, namely many unrelated stimuli presented as

dialogue labels, would be greater than the GSS effect of Brainstorming with forced movement

between dialogues.

These results suggest that the creative effects of both Assumption Reversals and Analogies

were constrained by the GSS effects, that is, cognitive inertia caused by lack of forced movement

among the dialogues and the increased distribution of ideas across more dialogues. To confirm

29

that cognitive inertia had occurred, an analysis of the distribution of each participant’s ideas

among the dialogues was conducted. Clusters of four or more ideas produced by the same

participant in one dialogue were counted. The Assumption Reversals treatment produced

nineteen clusters of four or more ideas, including one of twenty-three ideas. Nine clusters were

produced by the Analogies treatment.

In addition, the fragmentation of group memory for both Assumption Reversals and

Analogies resulted in fewer ideas per dialogue and hence fewer stimuli in each dialogue view. It

was not known a priori how many dialogues would emerge with Assumption Reversals and

Analogies. Assumption Reversals and Analogies produced an average of 3.5 and 5 times more

dialogues respectively than EBS, which had seven.

Quantity of Ideas

As expected, participants using Analogies generated significantly fewer ideas than those

using Assumption Reversals and Brainstorming. We attribute this result to the cognitive load that

was increased by force-fitting unrelated stimuli back to the problem. Perceived Ease of Use was

measured in this study to assess cognitive load. As expected, Analogies was perceived to be

harder to use than both Brainstorming and Assumption Reversals, significantly more so than

Brainstorming (p = 0.002). The fewer ideas produced by Analogies supports the hypothesized

inverse relationship between cognitive load and quantity.

Assumption Reversals produced more ideas than Brainstorming, but only at the (p = 0.107)

level. It appears that there was little difference in the cognitive load, since participants found

Assumption Reversals and Brainstorming equally easy to use (p = 0.209).

30

Implications

The findings of this research suggest that methods that use remote stimuli to produce creative

ideas do so at the cost of a higher cognitive load, which reduces the number of ideas produced.

Methods using stimuli that activate remote aspects of a thinker’s cognitive web require more

effort to generate ideas than methods such as EBS that use free association and frequent collision

of ideas. Unusual stimuli may be used to generate creative ideas, but methods that use them

should not be used to generate great numbers of ideas. This research shows that methods that list

and reverse assumptions as a way of exploring associations and stimulating new frames for the

problem can produce ideas, but on average, the ideas are less creative than ideas produced with

traditional EBS. Why is this so? Perhaps presenting reversed assumptions does move thinkers to

new starting places, but only to starting places that trigger automatic spreading activations. For

example, the reversed assumption, “restaurants do not serve food,” may succeed in leading the

thinker to explore ideas about use of entertainment and social gatherings in restaurants, but these

ideas may still stem from narrow, traditional activations about how other restaurants use these

devices. Thus, this method, rather than triggering frames that elicited uncommon associations,

triggered traditional frames.

Comparison between this study and other studies also produces important implications.

Analogies produced 22% fewer ideas than EBS. Assumption Reversals produced 20% more

ideas than EBS. Problem Decomposition [10], which used three questions with few non-rotating

dialogues, produced 35% more than an intact question with a single dialogue. And Directed

Brainstorming [50], which used 36 questions and few (N +1) rotating dialogues, produced 95%

more ideas than EBS. The implications are clear that use of multiple questions is a more

powerful intervention, in terms of taking thinkers to many places where they can quickly

31

generate ideas, than the methods evaluated in this study. Based on the findings from the Directed

Brainstorming study, it appears that forced rotation of fewer dialogues magnifies the effect of

many questions, although this could be confirmed by additional research.

This research has additional implications regarding the use of multiple dialogues. Both

Assumption Reversals and Analogies produced large numbers of dialogues--perhaps too many.

Although past research [10, 14] has shown that several dialogues are better than one dialogue,

the findings of this study suggest that many dialogues cause too much fragmentation of group

memory. This spreading of ideas across many dialogues means that fewer ideas are visible in one

dialogue. Consequently, when people visit any one dialogue they see a smaller portion of the

overall idea pool. Plus, without rotation, ideas from the current dialogue do not frequently and

automatically collide with ideas in the new dialogue. Future GSS tools could be designed to

support both dialogue names and rotation, thus reducing the tendency toward cognitive inertia.

However, many named dialogues would still cause idea pool fragmentation.

Limitations and Future Research

There were a number of limitations in this study. The methodology adopted was a laboratory

experiment, which is subject to a set of well-known limitations, including the use of students and

ad hoc groups. Facilitation, a controlled variable in this study, was therefore consistent and more

passive than it might be in traditional organization settings.

Future research should continue to study how variations in both creative techniques and GSS

tool configurations affect creative outcomes. Because of the importance of the idea pool and the

way it is manipulated and presented, future research could quantify the effects of multiple

dialogues with forced movement versus the same number of dialogues with free movement.

Future research should also determine the optimum number of dialogues.

32

Conclusion

This paper has provided both conceptual and practical insights into the use of creative

techniques with GSS. This research supports other researchers [6, 10, 50] who have pointed out

that seemingly small differences in techniques, scripts, and tool configurations can have large

implications for creative outcomes. Given the findings in this study regarding the effects of

different types of stimuli and presentation of those stimuli on creativity, it is clear that that there

is much more to learn. The considerable potential of GSS to support the creative process will

best be realized as research continues to clarify how structural aspects of both creative

techniques and GSS impact creative outcomes.

33

References

1. Aiken, M.; Vanjani, M.; and Paolillo, J. A comparison of two electronic idea generation techniques. Information and Management, 30 (1996), 91-99. 2. Bessemer, S. P., and O'Quin, K. Creative product analysis: testing a model by developing a judging instrument. In S. G. Isaksen (ed.), Frontiers of Creativity: Beyond the Basics. Buffalo, NY: Bearly, 1987, pp. 341-357.

3. Besemer, S. P., and Treffinger, D. J. Analysis of creative products: review and synthesis. Journal of Creative Behaviour, 15, 3 (1981), 158-178.

4. Boden, M. The Creative Mind. London: Abacus, 1990. 5. Bouchard, T. J. A comparison of two group brainstorming procedures. Journal of Applied Psychology, 56 (1972), 418-421. 6. Briggs, R. O.; de Vreede, G-J.; Nunamaker, J. F. Jr; and Tobey, D. ThinkLets: Achieving predictable, repeatable patterns of group interaction with group support systems. Proceedings of the Thirty-fourth Hawaii International Conference on System Sciences, 2001. 7. Collins, A. M. and Loftus, E. F. A spreading activation theory of semantic processing. Psychological Review, 82, 6 (1975), 407-428. 8. Connolly, T.; Routhieaux, R.L.; and Schneider, S.K., On the effectiveness of group brainstorming: test of one underlying cognitive mechanism. Small Group Research, 24, 4 (1993) 490-503.

9. Couger, J. D. Creative Problem Solving and Opportunity Finding. Danvers, Mass: Boyd and Fraser, 1995.

10. Dennis, A. R.; Aronson, J. E.; Heninger, W. G.; and Walker, E. D. II. Structuring time and task in electronic brainstorming. MIS Quarterly, 23, 1 (1999), 95-108.

11. Dennis, A. R., and Gallupe, R. B. A history of group support systems empirical research: lessons learned and future directions. In L. M. Jessup and J. S. Valacich (ed.), Group Support Systems: New Perspectives. New York: MacMillan, 1993, pp. 59-77. 12. Dennis, A. R.; George, J. F.; Jessup, L. M.; Nunamaker, J. F.; and Vogel, D. R. Information technology to support electronic meetings. MIS Quarterly, 12, 4 (1988), 591-624.

13. Dennis, A.; Kelly, G.; Hayes, G.; Daniels, R.; Orwig, R.; and Dean, D. Business process re-engineering with IDEF and electronic meeting systems. A paper delivered at the IDEF Users Group Conference, May 1994. 14. Dennis, A. R.; Valacich, J. S.; Carte, T.; Garfield, M.; Haley, B.; and Aronson, J. E. Research report: the effectiveness of multiple dialogues in electronic brainstorming. Information Systems Research, 8, 2 (1997), 203-211. 15. Dennis, A. R.; Valacich, J. S.; Connolly, T.; and Wynne, B. E. Process structuring in electronic brainstorming. Information Systems Research, 7, 2 (1996), 268-277.

34

16. Diehl, M., and Stroebe, W. Productivity loss in brainstorming groups: toward the solution of a riddle. Journal of Personality and Social Psychology, 53, 3 (1987), 497-509.

17. Dreistadt, R. The use of analogies and incubation in obtaining insights in creative problem solving. The Journal of Psychology, 71 (1969), 159-175.

18. Ellis, H. C. and Hunt, R. R. Fundamentals of Cognitive Psychology. Madison, Wisconsin: WCB Brown & Benchmark, 1993.

19. Gallupe, B. R.; Bastianutti, L. M.; and Cooper, W. H. Unblocking brainstorms. Journal of Applied Psychology, 76, 1 (1991), 137-142.

20. Gallupe, R. B.; Dennis, A. R.; Cooper, W. H.; Valacich, J. S.; Bastianutti, L. M.; and Nunamaker, J. F. Electronic brainstorming and group size. Academy of Management Journal, 35, 2 (1992), 350-369. 21. Gettys, D. F.; Pliske, R. M.; Manning, C.; and Casey, J. T. An evaluation of human act generation performance. Organizational Behaviour and Human Decision Processes, 39 (1987), 23-31.

22. Grossman, S. R. Releasing problem solving energies. Training and Development Journal, 38 (1984), 94-98.

23. Gryskiewicz, S. S. A study of creative problem solving techniques in group settings. Doctoral Dissertation, University of London,1980.

24. Higgs, M. and Dulewicz, V. An investigation into the relationship between divergent thinking and measures of competence. Working Paper No HWP 9709, Henley Management College, Henley-on-Thames, Oxfordshire, 1997. 25. Isaksen, S. G. Frontiers of Creativity Research: Beyond the Basics. Buffalo, NY: Bearly, 1987. 26. Jackson, P. W., and Messick, S. The person, the product, and the response: conceptual problems in the assessment of creativity. Journal of Personality, 33 (1965), 309-329. 27. Koestler, A. The Act of Creation. London: Penguin, 1964.

28. Lamm, H., and Trommsdorff, G. Group versus individual performance on tasks requiring ideational proficiency (brainstorming): a review. European Journal of Social Psychology, 3 (1973), 361-387. 29. Massetti, B. An empirical examination of the value of creativity support systems on idea generation. MIS Quarterly, (March, 1996), 83-97. 30. McGrath, J. E., and Hollingshead, A. B. Groups Interacting with Technology: Ideas, Evidence, Issues, and an Agenda. Thousand Oaks, CA: Sage, 1994. 31. McLeod, P. L. An assessment of the experimental literature on electronic support of group work: results of a meta-analysis. Human-Computer Interaction, 7 (1992), 257-280. 32. Mednick, S. A. The associative basis of the creative process. Psychological Review, 69, 3 (1962), 220-232. 33. Miller, G. A. The magical number seven, plus or minus two: some limits on our capacity for information processing. Psychological Review, 63 (1956), 81-97.

35

34. Mullen, B.; Johnson, C.; and Salas, E. Productivity loss in brainstorming groups: a meta-analytic integration. Basic and Applied Social Psychology, 12, 1 (1991), 3-23.

35. Nagasundaram, M. The structuring of creative processes with group support systems. Doctoral Dissertation, University of Georgia, 1995.

36. Nagasundaram, M., and Bostrom, R. P. The structuring of creative processes using GSS: a framework for research. Journal of Management Information Systems, 11, 3 (1994/1995), 87-114. 37. Nagasundaram, M., and Dennis, A. R. When a group is not a group: the cognitive foundation of group idea generation. Small Group Research, 24, 4 (1993), 463-489. 38. Neely, J. H. Semantic priming and retrieval from lexical memory: roles of inhibitionless spreading activation and limited-capacity attention. Journal of Experimental Psychology: General, 106, 3 (1977), 226-254.

39. Nunamaker, J. F.; Briggs, R. O.; Mittleman, D. D.; Vogel, D. R.; and Balthazard, P. A. Lessons from a dozen years of group support systems research: a discussion of lab and field findings. Journal of Management Information Systems, 13, 3 (1996-1997), 163-207. 40. Nunamaker, J. F.; Dennis, A. R.; Valacich, J. S.; Vogel, D. R.; and George, J. F. Electronic meeting systems to support group work. Communication of the ACM, 34, 7 (1991), 40-61.

41. Osborn, A. F. Applied Imagination, 3rd ed. New York: Scribner, 1963. 42. Perkins, D. N. Novel remote analogies seldom contribute to discovery. Journal of Creative Behaviour, 17 (1983), 223-239. 43. Petrovic, O., and Krickl. O. Traditionally-moderated versus computer supported brainstorming: a comparative study. Information and Management, 27 (1994), 233-243. 44. Pinsonneault, A., and Kraemer, K. L. The effects of electronic meetings on group processes and outcomes: an assessment of the empirical research. European Journal of Operational Research, 46, 2 (1990), 143-161.

45. Potter, R. E. Brainstorming with a GSS: exploring over time the effects of causal thinking on idea generation and synergy. Proceedings of the Thirtieth Hawaii International Conference on System Sciences, 1997. 46. Prince, G. M. The operational mechanism of synectics. Journal of Creative Behaviour, 2, 1 (1967), 1-13. 47. Rickards, T. Creativity and Problem Solving at Work. Aldershot: Gower, 1990.

48. Rosch, E. Cognitive representations of semantic categories. Journal of Experimental Psychology: General, 104, 3 (1975), 192-233.

49. Sambamurthy, V., and Chin, W. W. The effects of group attitudes toward alternative GDSS designs on the decision-making performance of computer-supported groups. Decision Sciences, 25, 2 (1994), 215-241. 50. Santanen, E. L.; Briggs, R. O.; and de Vreede, G-J. The cognitive network theory: a new causal model of creativity and a new brainstorming technique. Proceedings of the Thirty-third Hawaii International Conference on System Sciences, 2000.

36

51. Satzinger, J. W.; Garfield, M. J.; and Nagasundaram, M. The creative process: the effects of group memory on individual idea generation. Journal of Management Information Systems, 15, 4 (1999), 143-160. 52. Tulving, E. Episodic and semantic memory. In Tulving, E. and Donaldson, W. (ed.), Organization of Memory. New York: Academic Press, 1972. 53. Tversky, A. and Kahneman, D. Judgment under uncertainty: heuristics and biases. Science, 185 (1974), 1124-1131. 54. VanGundy, A. B. Techniques of Structured Problem Solving, 2nd ed. New York: Van Nostrand Reinhold, 1988. 55. VanGundy, A. B. Idea Power: Techniques and Resources to Unleash the Creativity in Your Organization. New York: AMACOM, 1992. 56. Wierenga, B. and van Bruggen, G. H. The dependent variable in research into the effects of creativity support systems: quality and quantity of ideas. MIS Quarterly, (March, 1998), 81-87.

57. Winograd, T. and Flores, F. Understanding Computers and Cognition: A New Foundation for Design. Norwood, NJ: Ablex Publishing Corp, 1986.

37

Appendix A – Facilitator Scripts For each of the creative techniques Brainstorming, Assumption Reversals and Analogies, participants were given a warm-up exercise to familiarize themselves with both the technique and the GSS. Brainstorming participants were given the Brainstorming Rules, and Assumption Reversals and Analogies participants were given the instructions for, and an example of, their respective technique.

Facilitator scripts were identical for the three techniques, except for the following:

Brainstorming. To enter your ideas: Click on ‘Go’.

Type one idea at a time into the window. Submit your idea for display by clicking the left mouse button on the Submit button.

You’ll then get another page of ideas entered by other participants. If you are stuck for ideas, use the ideas that have already been generated, by scrolling through the list, to stimulate your own thinking. If you want to see more ideas without entering one, just press Submit.

Assumption Reversals. First, list all the basic assumptions about the problem. To list your assumptions, click the left mouse button on the + sign button, type your assumptions one at a time into the Add Idea window, and submit each one by clicking on the Submit button. Be concise. You will be able to see assumptions entered by other participants by scrolling up and down. You have 5 minutes to list your assumptions.

To make it easier we’ll merge any assumptions that are duplicated and delete any that are more solutions than assumptions.

Now use these reversed assumptions to generate ideas on how to solve the problem. To do this: Select one of the reversed assumptions by double-clicking on it.


You will be able to see ideas entered by other participants. If you are stuck for ideas use the ideas that have already been generated, by scrolling through the list, to stimulate your own thinking.

38

To change to another reversed assumption, click on the Close button, and double-click on the new reversed assumption. Don’t feel that you have to enter ideas under all the assumptions. If you have an idea that doesn’t fit under an assumption, enter it anyway.

Analogies. First, generate analogies by completing the sentence ‘……… is like……….’ To list your analogies, click the left mouse button on the + sign button, type your analogies one at a time into the Add Idea window, and submit each analogy by clicking on the Submit button. Be concise. You will be able to see the analogies entered by other participants by scrolling up and down. You have 5 minutes to list your analogies.

To make it easier we’ll merge any analogies that are duplicated and delete any that are more solutions than analogies.

The Technographer moves the analogies into category buckets.

Now, forget about the problem. The next task is to elaborate on the analogies by listing their details, such as parts, functions or uses. To do this, select one of the analogies by double-clicking on the bucket.

Generate details about the analogy, not the problem, by clicking on the + sign button, typing the detail in the Add Idea window, and submitting it by clicking on the Submit button. Be concise. To select another analogy: click on the Close button and double-click on another analogy bucket. Don’t feel you have to enter details for all the analogies; just choose the ones that look interesting. You have 5 minutes to generate details about the analogies.

Now use these details you generated about the analogies as stimuli to generate ideas on how to solve the problem.

To do this, select one of the details by double-clicking on one of the analogy buckets and double-clicking on one of the details.


You will be able to see ideas entered by other participants. If you are stuck for ideas, use the ideas that have already been generated, by scrolling through the list, to stimulate your own thinking.

39

To change to another analogy detail, click on the Close button and double-click on the new detail. If you want to change to another analogy, just double-click on the appropriate bucket.

Don’t feel that you have to enter ideas under all the details. If you have an idea that doesn’t fit under a detail, enter it anyway.

Appendix B – Ease of Use Questionnaire

I enjoyed using this idea-generation technique

While using this idea-generation technique I felt comfortable

Using this idea-generation technique was fun

I felt frustrated by this idea-generation technique (reverse scaled)

On the whole, I felt very comfortable with this idea-generation technique, and would be

willing to use it again

40

Technique

Remoteness of stimuli

Cognitive load

Creativity of ideas

Quantity of ideas

Brainstorming Low Low Low High

Assumption Reversals

Medium Medium Medium Medium

Analogies High High High Low

Table 1. Relationship Between Stimuli, Cognitive Load, and Quantity and Creativity of Ideas.

41

Brainstorming

(B) Assumption

Reversals (R) Analogies (A) Test Statistics

Creativity

Mean SD Range

3.28 0.46

2.33-4.38

3.16 0.77

1.95-5.77

3.41 0.44

2.63-4.5

Kruskal-Wallis (Mean Rank)

69.77 53.80 80.43 Chi-Square = 10.575 P = 0.005*

Mann-Whitney (Mean Rank)

51.29 39.71 P = 0.036*


41.48 49.52 P = 0.144


37.09 53.91 P = 0.002*

Quantity

Mean SD Range

15.47 5.57 6-27

18.60 8.18 9-39

12.71 4.64 2-23


69.29 81.63 53.08 Chi-Square = 12.119 P = 0.002*


41.07 49.93 P = 0.107


51.22 39.78 P = 0.037*


54.70 36.30 P = 0.001*

Ease of Use

Mean SD Range

31.66 3.34

17-35

30.24 4.65

16-35

29.09 4.79 9-35


78.98 67.63 54.38 Chi-Square = 9.076 P = 0.011*


48.45 41.62 P = 0.209


53.02 35.98 P = 0.002*


49.01 40.90 P = 0.137

* = Significant at the 0.05 level. Table 2. Comparison of Main Effects across Treatments

42

Hypothesis Number Dependent Variable Hypothesis Supported P H1a Creativity A>R Yes 0.002* H1b Creativity R>B No (B>R) 0.036* H1c Creativity A>B No 0.144 H2a Quantity R>A Yes 0.001* H2b Quantity B>A Yes 0.037* H2c Quantity B>R No (R>B) 0.107

* = Significant at the 0.05 level. B = Brainstorming, R = Assumption Reversals, A = Analogies. Table 3. Summary of Tests of the Hypotheses Brainstorming Assumption Reversals Analogies

Stimuli Type None (apart from the problem statement and other ideas)

Reversed assumptions (related to the problem)

Analogy details (unrelated to the problem)

Number of Stimuli N/A 24 (average) 35 (average)

Presentation of Stimuli N/A Dialogue labels Dialogue labels

Partitioning of Group Memory (Number of Dialogues)

7 24 (average) 35 (average)

Manipulation of Group Memory (Movement Between Dialogues)

Forced Free Free

Table 4. Creative Technique and GSS Effects.

An examination of the impact of stimuli type and GSS structure on creativity: Brainstorming versus non-brainstorming techniques in a GSS environment

Documents