Plan of action

1
Plan of action

• Discuss some recent research of mine involving
  creativity.
• Transition into a broader discussion of research.
• Take the discussion wherever you want to go.

2
Measuring and Motivating Quantity, Creativity, or
Both
Presentation at Australian National Univ. June 5,
2007

• Steven J. Kachelmeier
• Bernhard E. Reichert
• Michael G. Williamson

3
Research question

• What are the effects of performance-based
  compensation schemes that are contingent on
  measures of quantity, creativity, or both?

4
What we know from the literature

• Several accounting experiments have studied the
  effects of quantity incentives, generally using
  relatively mundane production tasks.
• Examples
• Chow (TAR 1983) Decoding numbers into letters
  to simulate an assembly line setting.
• Young (JAR 1985) and Young et al. (TAR 1993)
  Building toy castles out of Loc-Blocs.
• Bailey McIntyre (TAR 1992) and Bailey et al.
  (JMAR 1998) Various Erector Set assemblies.
• Fessler (JMAR 2003) Solving water jug problems.
• Fisher et al. (TAR 2002 2005) Decoding numbers
  into letters.

5
What we know from the literature

• Several accounting experiments have studied the
  effects of quantity incentives, generally using
  relatively mundane production tasks.
• Management and psychology researchers have
  considered creativity rewards, but have been
  subject to alternative interpretations.

6
What we know from the literature

• Several accounting experiments have studied the
  effects of quantity incentives, generally using
  relatively mundane production tasks.
• Management and psychology researchers have
  considered creativity rewards, but have been
  subject to alternative interpretations.
• Example (from Amabile 1996) Children in an
  incentive condition were told that if they told
  a story based on an assigned reading, they could
  play with a camera.

7
What we know from the literature

• Several accounting experiments have studied the
  effects of quantity incentives, generally using
  relatively mundane production tasks.
• Management and psychology researchers have
  considered creativity rewards, but have been
  subject to alternative interpretations.
• To our knowledge, no previous research has
  considered compensation that is contingent on
  both quantity and creativity.

8
What we mean by creativity

• Original ideas, innovative, and clever

9
Why is this important to business?

• Firms benefit from creative innovations (Fallon
  and Senn 2006).

One of the most important things for our people
to remember is that we look to them for
creativity and innovation. Ideas and
creativity make Microsoft a world
leader. People are at the heart of every
segment of our business. Their creativity,
innovation, energy and motivation are the driving
forces of our success.
10
Why is this important to us?
The Australian National Universitys primary
research objectives are to achieve national
leadership through world-class performance in
scholarship and creative activity. UT Austin
has world-class creative research efforts in
significant areas of interest.
11
How would an accountant approach the problem?

• Multiple inputs (e.g., both quantity and
  creativity) likely contribute to overall
  performance and profitability (Chang and Birkett
  2004).
• Accounting solution Incorporate creativity
  within a multi-dimensional performance measure
  (Kaplan and Norton 1996).
• But firms are reluctant to do so (Ittner et al.
  1997).

12
Possible reasons why firms are reluctant

• Incremental rewards for creativity may not work.
• Even if creativity rewards did work, creativity
  is difficult to measure.

13
Task DESIGN REBUS PUZZLES

• Why rebus puzzles?
• Meaningful variation in both quantity (range 1
  to 64) and creativity (averages by participant
  range from 2.89 to 7.73).

14
Findings in one slide

• Do you get what you pay for?
• Quantity? YES
• Creativity? YES, in terms of average creativity
  ratings
• Quantity Creativity? NO
• Paying people for creativity-weighted
  productivity results in lower weighted
  productivity scores than we get when we pay for
  quantity only.

15
Method

• Task Design rebus puzzles for 20 minutes. In
  total, 78 participants generated 1,360 puzzles.
• All participants informed that we value both
  quantity and creativity.
• All participants given a definition of creativity
  as puzzles that are original ideas, innovative,
  and clever.

16
Design

• Fixed U.S. 25 fixed pay.
• Quantity 5 to 45 (midpoint 25) based on the
  number of puzzles.
• Creativity 5 to 45 based on average
  creativity ratings.
• Both 5 to 45 based on quantity average
  creativity ratings.
• NOTE

S Creativity ratings

• Quantity

17
Obtaining creativity ratings

• 11 raters 2 used initially for payment
  purposes, with 9 others added later in a rating
  session using radio-frequency response devices.
• Raters see the same instructions as the
  participants, absent the compensation paragraph.
• All raters correlate positively with the group
  significantly so for 10 out of 11.
• Cronbachs alpha 0.86.
• Results are insensitive to dropping the two
  initial raters or dropping the three raters with
  the lowest intergroup correlations.

18
Results Quantity
Quantity Produced
19
Results Average Creativity Ratings
Average Creativity Rating
20
Sidebar What is creativity?

• In general, quantity incentives encouraged and
  creativity incentives discouraged participants
  from extending the patterns illustrated in
  instructional examples.

21
Following instructional examples

• Quantity incentives led to more output that
  extended patterns illustrated in instructional
  examples (e.g., extending existing product lines).

Quantity Incentives Output
Instructional Examples
Man Board
pains
22
Results Creativity-Weighted Productivity
23
Whats going on?
Theory

• Creativity incentives lead people to focus only
  on creative output, foregoing gains from more
  straightforward extensions when creative ideas
  are not available.
• This simplification is a type of noncompensatory
  strategy (Payne et al. 1993).
• However, if Amabile (1996) is correct that one
  cannot be more creative simply by trying harder,
  this strategy is unlikely to succeed.
• Meanwhile, participants with quantity-only
  incentives were producing lots of ideas, some of
  which were very creative.

24
In a picture
Result Creativity incentives reduce overall
output without increasing high-creativity output.
25
Examples
Creativity rating 9.18 creativity-only
condition
26
More examples
Creativity rating 8.18 creativity-only
condition
27
More examples
Creativity rating 7.64 weighted
quantity/creativity condition
28
More examples
Creativity rating 7.82 weighted
quantity/creativity condition
29
More examples
Creativity rating 8.27 quantity-only condition
30
More examples
Creativity rating 8.00 quantity-only condition
31
Lowest rating of all 1,360 puzzles
Creativity rating 1.55 quantity-only condition
32
Same person who did Chewing gum
Creativity rating 6.82 quantity-only condition
33
Last example
Creativity rating 4.27 fixed-pay condition
34
So based on these results, is adding creativity
to multi-dimensional performance measures a good
idea or a bad idea?

• Which vita would impress you more?
• Three top-tier publications, or
• Three top-tier publications and five second- and
  third-tier publications?
• It boils down to whether output of lesser
  creativity is still valuable at the margin.

35
Drawing research strategy implications from a
rebus-puzzle experiment is a big stretch, but
here goes

• Just do research
• If you try enough times, some of it will be
  excellent.
• When you have something good, get credit for it.
• Submit your best work to the best journals.

36
Counterarguments and caveats

• Experience on the Promotions and Tenure
  Committee, UT-Austin School of Business
• Candidate X had four publications - three
  top-tier and one near top-tier.
• Candidate Y had seven publications - three
  top-tier, one near top-tier, and three more at
  lower levels.
• X was promoted and tenured. Y was denied.

37
Clarifications previous example

• Xs research was high-impact, as evidenced by
  citations and by outside letter writers.
• Xs research could be easily identified with X.
• X was also an outstanding teacher.
• Ys research was less focused and included
  multiple articles from the same data.
• Ys research was more identifiable with Ys
  dissertation committee than with Y.
• Ys teaching record was marginal.

38
Controversies

• Is there a North American bias in access to
  scholarly business journals?

39
Non-North American University affiliations
reflected in The Accounting Review, April 2006
July 2007.

• University of Zurich
• University of Hong Kong
• Nanyang Technological University
• Tilborg University
• Hong Kong University of Science and Technology
• Melbourne Business School
• Hebrew University
• Maastricht University
• Monash University
• University of New South Wales

40
Non-North American University affiliations
reflected in The Accounting Review, April 1996
July 1997.

• Hong Kong University of Science and Technology
• National Chengchi University

41
Your turn

• Your biggest gripe about journal editors