1

The Use of Diary Methods to Study New Questions in Employee Innovation

Kerrie L. Unsworth

Work Effectiveness Research Programme

School of Management

Queensland University of Technology

k.unsworth@qut.edu.au

Chris W. Clegg

Institute of Work Psychology

University of Sheffield

c.clegg@sheffield.ac.uk

Stream U

2

The Use of Diary Methods to Study New Questions in Employee Innovation

To date, employee innovation research has been dominated by two research questions: What factors

increase the novelty and usefulness of innovation outcomes?; and What factors increase the

innovativeness of employees? While the answers to these questions are important, we argue that other

questions remain unanswered and that innovative innovation research is needed. We discuss

traditional methods of studying innovation and argue that a new method, diary studies, may be more

useful at answering these neglected questions. We outline considerations involved in diary studies and

then provide a case study where diary methods were successfully used to study employee innovation.

Keywords: employee innovation, diary study, methods

3

Employee innovation is receiving more and more attention in both organizations and the

organizational sciences. We argue, however, that this attention has focused on two research questions,

to the neglect of alternative questions and widened understanding. We examine some reasons for this

dominant approach, highlight alternative questions and discuss the use of diary studies as a means of

addressing these neglected questions.

Before moving on, however, two boundaries must be set. First, following many authors, we

define employee innovation as the generation, suggestion and implementation of an organisationally­

oriented idea (e.g., Forehand, 1963; Kanter, 1983; Mumford & Gustafson, 1988; Robertson, 1967;

Scott & Bruce, 1994; Van de Ven, 1986). The generation is at the individual level, and has often been

studied as creativity, while implementation may occur at all levels including individual, team and

organisation. Second, although some researchers and theorists suggest that methods are inextricably

linked to epistemological positions (e.g., Guba, 1987), we support the views of Bryman (1988) and

others who conclude that methods are best chosen on the basis of technical reasons (i.e., what best fits

the research question).

The Traditional Dominant Approach to Studying Employee Innovation

There are a number of themes which underlie most employee innovation research. We review

three of these and demonstrate how they have led to a focus on a particular approach to studying

employee innovation. As we shall discuss, this traditional approach focuses on two main questions: 1)

What factors increase the novelty and usefulness of employee innovation outcomes?; and 2) What

factors increase the innovativeness of employees? While the answers to these questions are obviously

important, such a narrow focus has hindered a greater understanding of employee innovation. We will

discuss each of these three themes – product versus process orientation, combining innovation types,

and unitary versus pluralistic stance – identifying the traditional dominant approach, and conversely,

the areas that have been neglected.

First, a product­orientation has dominated employee innovation research. Nearly all research

conducted in this field has examined innovation as an outcome of the innovation process (e.g.,

Damanpour, 1991; Oldham & Cummings, 1996; West, 1987), rather than the process itself. Indeed,

even when Scott and Bruce (1994) measured employee innovativeness across five stages of

4

innovation, they aggregated these concepts into a single scale and examined predictors of this

aggregated product outcome instead of examining each stage separately in a processual approach.

Drazin, Glynn and Kazanjian (1999) proposed that creativity be reconceptualised such that the process

itself is measured; we concur and argue that innovation too should be reconceptualised as a process.

Empirically, three studies have shown support for this proposition. Axtell et al., (2000), Clegg,

Unsworth, Epitropaki, and Parker (2001), and Unsworth (2000) all separated the measurement of ideas

generated or suggested from the implementation of those ideas. Across four industries (manufacturing,

engine design, private, and public sector, respectively) the results were surprisingly consistent, and the

two stages were affected by different factors. Nonetheless, these are only crude markers denoting two

major stages, and more work must be done on the finer elements of the innovation process.

Second, employee innovation research has generally not differentiated between types of

innovation or has focussed exclusively on the innovation of employees whose jobs require innovation

(see Unsworth, 2001). It may be that different types of innovation, differentiated by the reason for

engagement or task type (Unsworth, 2001) are affected by different factors or undergo different

processes. While there does seem to be some consideration of product and process innovation (see

Damanpour, 1991), these two are often combined into a single aggregated innovativeness outcome

(e.g., West, 1987; Axtell et al., 2000). More research on the differences between innovation types is

therefore necessary.

Finally, one of the features that underpins much of the research in organizational behaviour

generally, and in innovation in particular, is a relatively unitary orientation (see Fox, 1974; Burrell &

Morgan, 1986). In this view, people working within organizations are assumed to share the same

goals, values and orientations. Others, however, subscribe to a more pluralistic perspective where

employees may have different goals and perspectives, in part depending on their role and function in

the company (see Lawrence & Lorsch, 1965). This causes problems for the traditional criteria of

employee innovation (novelty and usefulness: e.g., Amabile, 1983; Richards, et al., 1988), as the

adoption of a pluralistic perspective renders the notion of ‘usefulness’ problematic. The question

becomes: Useful to whom? Our view is that ‘usefulness’ should not be a criterion on which to assess

whether or not something is innovative. Rather, we interpret usefulness as an important outcome of the

5

innovation process, and one that needs to be evaluated by considering the views of the various

stakeholders over time. In addition, a pluralistic perspective acknowledges the negotiation and other

political processes that will occur throughout the generation and implementation of the idea, as well as

the people and groups that are indirectly involved in, and affected by, the innovation process.

Thus, it can be seen that, historically, the dominant approach in employee innovation has

focused upon the innovation outcome (whether it be the innovation incident or the employee

innovativeness), against novelty and usefulness criteria, without differentiating between innovation

types, and without considering pluralistic viewpoints.

While we acknowledge the importance of this research and of answering the questions proposed

by these dominant themes, we believe that a number of alternative questions may be asked. Below, we

have listed only some of these potential questions:

1. How does the employee innovation process unfold? Do different factors affect different

stages in the innovation process?

2. Do different factors affect the process of different innovation types?

3. Does employee innovation mean different things to different stakeholder groups? How do

the traditional criteria for innovation (novelty, usefulness, innovativeness) differ across

different stakeholder groups? How do predictors differ across different stakeholder groups?

To access this more complete understanding of employee innovation, we must rethink the

methods that are currently used. To date, the most widespread methods in employee innovation are

survey scales, interviews and observations. In this article, we discuss the use of diary studies, a new

method for studying employee innovation, in light of the issues raised above. We believe that a diary

method may be the best approach for addressing some of these new research areas in employee

innovation. Below, we outline the diary method and compare its ability to address these questions with

more traditional methods.

Diary Method

Diary methods have been used in various fields; for example in the study of mood (Totterdell,

Kellett, Teuchmann, & Briner, 1998) and health (Hahn, 2000). Diary studies engage their participants

in a direct and frequent recording of participants’ experiences and events, either in their own words or

6

by responding to rating scales, on a frequent basis. This time basis can be either a fixed time­slot, such

as every evening or every Friday morning, or event­based, such as every time they feel angry. Bolger,

Davis, and Rafaeli (2003) discuss these different types of diaries and suggest that “event­based design

is usually most appropriate for diary studies of specific classes of phenomena or processes, especially

those that are isolated and/or rare” (p.590). Thus, for innovation diary studies, we will be considering

event­based diaries.

As far as we are aware, diaries have previously been little used in innovation research. Recently,

Amabile et al. (2004) conducted a large study whereby 238 managers completed diaries every day.

However, the diaries were used to gather data on leader behaviours and perceptions of work, rather

than innovation per se. Indeed, the innovation variable (subordinate creativity) was based on a rating

at the end of the month as to how creative they, and their fellow team members, had been during that

month. While not dismissing the importance of Amabile et al.’s (2004) research, to our knowledge a

diary study of innovation dependent variables addressing the alternative research themes identified

earlier has not yet been conducted.

Using Diaries To Address New Areas of Research

Before going further, a valid question to be asked is whether diary methods provide any added

value over more traditional forms of studying innovation. Therefore, we will discuss three more

traditional methods (survey scales, interviews, and observations) and their ability to address the

neglected research questions. We then show how diaries can resolve the problems inherent in the more

traditional methods when addressing these questions.

Surveys are perhaps the traditional method of examining employee innovation, including both

self­report (e.g., Axtell et al., 2000; Janssen, 2000) and supervisor ratings (e.g., Scott & Bruce, 1994;

Oldham & Cummings, 1996). However, despite their widespread use, the traditional innovation scales

provide only limited information for the new research areas we proposed. They generally focus upon

the end­product rather than the process itself, thus not allowing for information to be collected about

predictors of different stages within that process. Because of this focus upon the outcome rather than

the trigger or task type, the data needed to differentiate innovation types is also not collected. Finally,

research involving traditional scales tends to adopt a unitary stance, analysing the data using the

7

assumption that all concepts and relationships are interpreted similarly across different stakeholder

groups.

It can clearly be seen that innovation survey scales have been fashioned in a way to fit the

historically predominant questions in innovation research. As such, other questions regarding the

innovation process, innovation types, and stakeholder perspectives are neglected. Indeed, the data

gathered is limited to the research questions determined prior to the study and there is little

opportunity for other issues to arise.

Alternatively, there are methods already used in innovation research which do allow for

processual information and stakeholder perspectives. Interviews and observations have been used

repeatedly in employee innovation research (e.g., Amabile & Gryskiewicz, 1987; King, 1990) and

provide an extremely flexible approach to gathering rich data. Although they have previously been

used to further the traditional research questions, they can easily gather data on product and process

and innovation types, as well as identifying pluralistic viewpoints and emergent issues.

Nevertheless, interviews are reliant upon retrospective data and the recall of participants. The

data are usually concerned with innovations in the last month or last 6 months. Obviously, there may

be problems in forgetting the actual occurrences, as well as hindsight and other retrospective biases.

Alternatively, observational studies are recorded in real time, thus, minimising problems of bias

that may occur due to retrospective recall. Nevertheless, the timing of innovations cannot be predicted:

thus, while one is not reliant upon self­report data, one is reliant upon serendipity of researcher

presence. King (1992) notes that “Not all the innovation histories were equally detailed. In some cases

the observer was absent when key events occurred or decisions made” (p.91). In addition, it is well

recognised that a lot of the innovation processes are cognitive (e.g., Wallas, 1926), and thus, occur

only within the head of the individual. Observation alone can record only the behavioural

manifestations of these cognitions.

Diaries can be designed to incorporate positive aspects of these methods, while minimising the

negative ones. First, unlike surveys, diaries can be used to gather a wide range of information

including the employee innovation process, innovation types, and pluralistic standpoints. We discuss

later some potential designs that may address these questions.

8

Unlike interviews, the information is recorded in real­time, reducing retrospective biases and

forgetting. Similar to observations, the innovation process can be recorded as it happens, alongside

current perceptions of possible predictors.

Unlike observations, the data is recorded by the idea generator as the idea happens, thus

removing a reliance on serendipity. The researcher does not have to be present at all times as the onus

is on the participant to record the ideas. Similarly, the cognitive processes that would be hidden from

the researcher are recorded as well as the overt behavioural processes.

It is clear then, that diaries hold much potential for answering alternative questions in the field

of employee innovation. However, like any methodology, important decisions need to be made

concerning the design of each particular study. While other sources discuss the issues of diary studies

more generally (e.g., Bolger, et al., 2004; Stone, Kessler & Haythornwaite, 1994), we focus on the

issues most relevant to using diaries in new areas of research into employee innovation.

Design Issues for Diary Studies

Perhaps the first, and most important decision to be made is what questions to ask in the diary.

As noted earlier, diaries have the potential to gather much varied information and researchers must

take care to ask the right questions. If one is interested in understanding the process of employee

innovation, the diary could prompt for ideas that have been generated, how they came about, and what

actions are being taken regarding each one. If one is interested in understanding innovation types, the

diary could prompt for specific information regarding each idea and subsequent innovation outcome in

order for classification to occur. If one is interested in pluralistic viewpoints, the diary could prompt

for reasons why the participant feels each idea is considered to be innovative and analysis could look

for differences across groups. These are only a few ways in which diaries can extend our

understanding of employee innovation.

In a similar manner, researchers must decide exactly what constitutes an ‘event’, that is, when

participants should record their innovation information. Bolger, et al. (2004) discuss a number of

issues concerning this point. They suggest that a clear definition of the event be used upfront to

minimise confusion and ambiguity as to which events should or should not be recorded. In the case of

innovation, we believe researchers should emphasise the importance of every idea, large or small,

9

work­related or non­work­related, successful or not successful. This emphasis on over­inclusiveness

should minimise any loss of data through non­recording of ideas – if extra information is recorded it

can always be discarded at a later date. In addition, from a pluralistic standpoint, a study may ask

participants to record every idea, but then to rate that idea on various dimensions such as novelty,

usefulness, likelihood of success, and so on. This allows information to be gathered on how different

groups of people view innovation.

The third issue for consideration is the choice of medium to be used when collecting data. Paper

diaries, in which participants respond to a set of pre­written questions, are easiest; however, to

minimise the bulk of the diary (a too hefty diary would appear intimidating and possibly reduce

participation), only a limited number of questions may be asked. Other media, such as hand­held

computers, could be used to store information more easily. These computers could be programmed to

ask the participant a number of questions regarding different innovation research aims. Although paper

diaries must prompt for all information for all cases, the computer could be programmed to ask only

for the information that is needed for the specific idea recorded. Hence, it would appear (and in most

cases, would be) less time intensive. Despite the advantages of using hand­held computers, they are,

nevertheless, expensive and resource constraints will obviously play a role in choice of diary medium.

Diaries require much more effort from participants than do other, more traditional methods of

studying innovation. As such, it is more likely that participants do not fully complete each diary or

withdraw from the study. Stone, et al. (1991) suggest that the following techniques may help in

improving retention: 1) providing feedback to participants regarding their progress; 2) providing initial

instructions to respondents that inoculates them from the inevitable setbacks of forgetting or being

unable to complete a section of the diary; and 3) providing encouragement and feedback especially

during the first week of the study. Bolger, et al. (2004) also make a number of suggestions for

improving retention including: 1) making the diaries portable either through binding them as pocket­

sized booklets, or providing them with small hand­held computers; and 2) maintaining contact with

participants throughout the study to encourage them and to check for questions or errors.

Finally, the diaries must be returned to the researcher and checked at some point. Stone et al.

(1991) suggest that diaries be mailed back at frequent intervals rather than all at the end to ensure

10

consistent participation. If a participant forgets to complete a diary, the researcher is soon aware of it

rather than waiting until the end. Such feedback to the researcher also decreases their ignorance of

participants “skipping” sections: a problem common in diary studies (Bolger et al., 2004).

Disadvantages

Despite the many benefits of diary studies in employee innovation research, there are still some

methodological issues that are not overcome by the use of diaries. Diaries are reliant upon self­report

data. Thus, there is the potential for omitting “taken­for­granted” issues and processes that are not

easily verbalised or recorded. As noted previously, many innovation processes are cognitive and, as

such, may not be salient to the participant. Also, although the process is ideally captured in real­time,

the likelihood of all participants consistently recording all ideas immediately is moderate. Participants

may forget to complete the diary which has many consequences ranging from retrospective recording

of data or omission of some details, to feelings of guilt which can lead to withdrawal from the study

(see Stone, et al., 1991). Finally, the diary must be relatively short to reduce the burden on the

participant (Bolger et al., 2004). This has obvious implications for the amount of data gathered, and a

trade­off must be made between the amount gathered and the burden on the participants.

To illustrate the use of diary studies in expanding our knowledge of employee innovation, we

now describe research that has used the diary method. The diary study was part of a larger programme

of research into employee innovation, and interviews and observations were also conducted. This

paper will, however, concentrate predominantly on the diary studies.

Case Study

Participants

Twenty­five engineers from two organizations participated in this diary study. Seventeen were

working on the last stages of a 20­year design project in collaboration with German, Italian and

Spanish companies. Eight were working on modifications to the design of turbines and new Navy

engines. Most participants were design engineers (17), three were stress engineers, four were system

design engineers, and one was a technical clerk.

Design of Diary & Procedure

11

The diaries consisted of five sections. The first page contained an introduction to the study and a

‘thank you for participating’, an explanation of how to complete the diary, a reminder that forgetting

to fill them in once did not mean they had to stop, an explanation of the terms used within the diary, a

reminder about our definition of “ideas”, a confidentiality agreement, contact details, their code

number and date of starting. The second section concerned the actual recording of ideas. For each day

there was space to record three ideas (although it was explained that generating less was normal, but if

they generated more then there were two spare pages at the end). For each idea, participants were

requested to record the approximate time of generation; the idea itself; the trigger; and any actions that

they took. If they took actions on the idea later in the week, they were asked to record these also,

noting the date and time at which they occurred. The third and fourth sections were to be filled in on

the Friday of each diary week. First, they were asked four questions: to consider and write down what

they believed helped and hindered them to generate and implement ideas. Then they were asked to

complete a number of scales regarding their week: task motivation (Warr et al., 1979); well­being

(Warr, 1990); work overload (Borrill, et al., 1998); and percentage of time they spend doing different

tasks (design, administration etc). The final section was left blank for comments. Finally, each week

the participant would receive a list of the ideas they had generated in the previous weeks to add any

further actions they had taken in implementing those ideas.

The diaries underwent a pilot study using postgraduate students from the University of

Sheffield. Four participants used the diaries for a week and then commented upon their design. Two

changes were made: 1) additional terms were explained on the front page; and 2) a verbal

encouragement and explanation that ideas are not finite packages to be easily recorded was added to

the initial introduction the study.

The first author then met with each of the diary study participants and went through a diary

explaining how to complete them. Two weeks of diaries were left with them along with contact

details. The diaries were received each Wednesday of the following week. A document for each

individual was created (or added to the previous document), recording the ideas and actions taken

throughout the week. This list was then printed and sent back to the participant with the following

12

week’s diary. Each participant completed four weeks of diaries and received four action sheets (i.e.

they received an action sheet after each week, including the final week). Action sheets also allowed

any misunderstandings in ideas to be rectified.

Findings

The first set of findings was used to validate the use of diaries as an accurate data collection

source. As the diaries were part of a wider programme of research, much of the data could be

compared to that collected from interviews and observations. Although this larger set of results goes

far beyond the remit of this paper, we found that much of the data identified in the diary studies

mirrored that found in the other studies. As such, we believe that the diaries can be used to access

accurate and valid information.

In addition however, data emerged that were not ascertained from the more traditional methods,

and it is this ‘extra’ analysis that is particularly interesting. First, the time taken for completing

innovations was able to be ascertained from the diaries. The majority of ideas were generated and

implemented within a week, and all but a few of these were implemented on the day of generation.

Fifty­four percent of all ideas were completed within a week; 23% within two weeks. Only eight

percent took three weeks or more to complete. It appeared that ideas in which the participant had

control over the implementation domain were implemented more quickly; when the ideas were

implemented elsewhere, they often took longer to be implemented. This finding has obvious

implications for managers wanting ‘speedy’ innovations.

Second, 15% of ideas did not have any action taken on them at all, apparently due to a form of

idea evaluation – when this evaluation was negative, no further actions were taken. This finding was

striking, as it contrasted with the findings from the interviews where participants suggested they

persisted with ideas in the face of negative evaluations (e.g., “people saying, ‘you can’t possibly do

that’, that seems to be the thing that makes me go away and sit in a corner and come up with various

ways you could do it.”, AN8, 228­230). However, the diaries showed no such challenges to negative

evaluations. Examination of this discrepancy showed that evaluations producing persistence were

13

those conducted by other people; when the idea generator themselves negatively evaluated the idea,

the concept was dropped. Thus, it was not the evaluation per se that caused persistence, but a belief

that the idea generator was more correct than the evaluator. This relates to other findings on the

importance of confidence in employee innovation and has implications for both future research and

management. In addition, it shows the value of real­time recording of processes, rather than post­hoc

analysis by the participant which may be subject to social desirability and hindsight biases.

Finally, task opportunity appeared to be the most influential factor in determining whether

innovation occurred or not. Eleven weeks of diaries did not contain any ideas: all eleven referred to the

fact that their work that week did not provide an opportunity for ideas. On the other hand, five people

reported that necessity caused them to innovate. While this data emerged from the interviews also, the

convincing strength of the influence is identified mainly from the diary study.

Conclusions

When conducting research into employee innovation, we must take care to be innovative

ourselves. The traditional approach to understanding employee innovation has led to a narrow focus

on combined innovation outcomes viewed through a unitary perspective. We argue that innovation is

necessary within our field, and that we approach new areas of research including innovation processes,

innovation types, and pluralistic viewpoints.

We believe that diary methods may help in furthering these new research areas. Diaries provide

rich details about innovation including processes, and innovation types. Although they are reliant upon

self­report data and are vulnerable to “forgetting”, they provide real­time data making them less likely

to be affected by retrospective biases, and are not dependent on serendipity of researcher location.

Furthermore, as identified in the case study, diary studies can not only validate data gathered in more

researcher time­intensive manners such as interviews and observations, but can also provide new and

alternative information.

Although much excellent research has been conducted using the traditional frameworks and

dominant paradigm, the future of innovation research relies upon addressing new areas and as such,

we need to begin considering the methods best suited to this future.

14

References

Amabile, T. M. 1983. The social psychology of creativity: A componential conceptualization. Journal

of Personality and Social Psychology, 45(2), 357­376.

Amabile, T. M., & Gryskiewicz, S.S. 1987. Creativity in the R&D laboratory (Technical Report

Number 30). Greensboro: Center for Creative Leadership.

Amabile et al 2004

Axtell, C.M., Holman, D.J., Unsworth, K.L., Wall, T., Waterson, P., & Harrington, E. 2000. Shopfloor

innovation: Facilitating the suggestion and implementation of ideas. Journal of Occupational and

Organizational Psychology, 73(3), 265­286.

Bryman, A. 1988. Quantity and quality in social research. London: Unwin Hyman.

Clegg, C.W., Unsworth, K.L., Epitropaki, O., & Parker, G. (2001). Implicating Trust in the Innovation

Process. Journal of Organizational & Occupational Psychology, .

Damanpour, R. 1991. Organizational innovation: A meta analysis of determinants and moderators.

Academy of Management Journal, 34(3), 555­590.

Drazin, R., Glynn, M.A., & Kazanjian, R.K. 1999. Multilevel theorizing about creativity in

organizations: A sensemaking perspective. Academy of Management Journal, 24, 286­307.

Forehand, G.A. 1963. Assessments of innovative behavior: Partial criteria for the assessment of

executive performance. Journal of Applied Psychology, 47(3), 206­213.

Guba, E.G. 1985. The context of emergent paradigm research. In Y.S. Lincoln (Ed.), Organizational

Theory and Inquiry: The Paradigm Revolution (pp.79­104). CA: Sage.

Hahn, S. 2000. The effects of locus of control on coping and reactivity to work interpersonal stressors:

A diary study. Personality and Individual Differences, 29, 729­748.

Hards, R. 1999. An investigation of barriers to creativity in engineering design: A progress report .

Plymouth: School of Computing, University of Plymouth.

Janssen, O. 2000. Job demands, perceptions of effort­reward fairness and innovative work behaviour.

Journal of Occupational and Organizational Psychology, 73(3), 287­302.

Kanter, R.M. 1983. The Change Masters. New York: Simon and Schuster.

King, N. 1992. Modelling the innovation process: An empirical comparison of approaches. Journal of

15

Occupational and Organizational Psychology, 65, 89­100.

Mumford, M. D., & Gustafson, S.B. 1988. Creativity syndrome: Integration, application and

innovation. Psychological Bulletin, 103(1), 27­43.

Oldham, G. R., & Cummings, A. 1996. Employee creativity: Personal and contextual factors at work.

Academy of Management Journal, 39(3), 607­634.

Richards, R., Kinney, D.K., Benet, M., & Merzel, A.P.C. 1988. Assessing everyday creativity:

Characteristics of the lifetime creativity scales and validation with three large samples. Journal of

Personality and Social Psychology, 54(3), 476­485.

Robertson, T.S. 1967. The process of innovation and the diffusion of innovation. Journal of

Marketing, 31(1), 14­19.

Scott, S. G., & Bruce, R.A. 1994. Determinants of innovative behavior: A path model of individual

innovation in the workplace. Academy of Management Journal, 37(3), 580­607.

Stone, A.A, Kessler, R.C., & Haythornthwaite, J.A. 1991. Measuring daily events and experiences:

Decisions for the researcher. Journal of Personality, 59(3), 575­607.

Totterdell, P., Kellett, S., Teuchmann, K., & Briner, R.B. 1998. Evidence of mood linkage in work

groups. Journal of Personality and Social Psychology, 74(6), 1504­1515.

Unsworth, K.L. 2000. Employee innovation: The roles of idea generation and idea implementation.

Presented at Annual Society for Industrial Organizational Psychology Conference, New Orleans,

14­17 April.

Unsworth, K.L. 2001. Unpacking creativity. Academy of Management Review, 26, 289­297.

Van de Ven, A. 1986. Central problems in the management of innovation. Management Science, 32,

590­607.

Wallas, G. 1926. The Art of Thought. New York: Harcourt Brace.

West, M.A. 1987. Role innovation in the world of work. British Journal of Social Psychology, 26,

305­315.