Advancing Marketing Knowledge - American Marketing Association

Download Report

Transcript Advancing Marketing Knowledge - American Marketing Association

Advancing Marketing
Knowledge
Kent B. Monroe
AMA Winter Educators’ Conference
February 12, 2005
Thanks and Acknowledgements





American Marketing Association
McGraw-Hill/Irwin
Selection Committee
Madhu Viswanathan
Current and former colleagues at Illinois,
Richmond, Virginia Tech, Massachusetts
 Current and former students
 My family
Advancing Marketing Knowledge
2
Three Recommendations
1. Encourage multiple approaches for
developing knowledge
2. Reduce use of traditional rules of thumb
3. Endorse ways of accumulating knowledge
 Researchers, educators, reviewers and
editors all play a significant role
Advancing Marketing Knowledge
3
Questions of Empirical Research
 Is there an effect? (Conclusion validity)
If so, can we detect it?
Have correct statistical analyses been used?
 Are there rival, plausible hypotheses?
(Internal validity)
 Can we generalize back to theory?
(Construct validity)
 Can we generalize to and across … ?
(External validity)
Advancing Marketing Knowledge
4
Some Ways to Enhance Discovery
 Allow for accidents and unexpected
observations to occur
 Determine reasons for variability or error
 Ease the focus on precision
 Get out in the field (observe, listen)
 Tinker around: let the “data” speak to you
Advancing Marketing Knowledge
5
Researchers’ Dilemma
 Maximizing construct validity may reduce
conclusion validity (finding an effect)
 Maximizing internal validity may reduce
construct validity of setting, sample,
treatments
 Generalizing effects to different constructs
requires multiple, diverse research efforts
Advancing Marketing Knowledge
6
Multiple Approaches
1. Theory + Method  Substantive
2. Substantive + Method  Theory
3. Substantive + Theory  Method
–
(McGrath and Brinberg, (Journal of Consumer Research, June 1983)
 It is myopic to believe there is only one
way to develop knowledge.
Advancing Marketing Knowledge
7
Rules of Thumb
 The fallacy of “statistical significance”
 Are empirical effects only significant if the
statistical test of the null hypothesis
achieves a probability level of .05?
 Is this an appropriate or even the only
criterion for accepting new knowledge?
Advancing Marketing Knowledge
8
Criteria for Selecting α
 Select small α if:
 Greater research
control
 Unsure of direction of
effect
 Large confidence
interval needed
 Large effect expected
 Select larger α if:
 Small sample size
(Hunter, Journal of Consumer
Research, June 2001)
 Early stages of
research
Advancing Marketing Knowledge
9
Other Considerations for Selecting α
 Consequences of type I error vs. type II
error
 Prior evidence against the null hypothesis
The convention of p < .05 is popular but
of dubious merit.
Advancing Marketing Knowledge
10
Accumulating Knowledge
 The problem of n = 1
– (Wells, Journal of Consumer Research, December 2001)
 Searching for new knowledge vs. building a “store”
of knowledge
 The need for
 Replication (Can results be reproduced?)
 Robustness (Do results hold across concepts, methods,
substantive areas?)
 Boundary conditions (When are results not supported?)
Re-search
Advancing Marketing Knowledge
11
Accumulating Knowledge



Knowledge accrual requires convergence of
findings derived from divergent methods.
Construct validity requires diversity of methods.
Robustness or generalizability requires diversity
of populations, occasions, settings, and
methods.
 Neither one study nor a few studies are
sufficient to determine knowledge.
Advancing Marketing Knowledge
12
Accumulating Knowledge

Replication is
–
–
complementary to diversity, generalizability,
robustness.
necessary to reduce the small sample bias of our
research

A replicated result represents information about
the reliability of a finding.
 The belief that replication research does not
make a significant contribution is a
misconception.
Advancing Marketing Knowledge
13
Accumulating Knowledge

We should encourage a diversity of
methods applied to any research domain,
by someone.
 It is not expected that any one researcher
be an omnimethod expert.
 However, we must demand it of the
Marketing discipline as a collective.
Advancing Marketing Knowledge
14
A Fantastic and Rewarding Journey
 Discovering and developing new knowledge
 Writing and communicating this knowledge
 Teaching and mentoring
– Doctoral students (36 dissertations)
– Masters students (MBA and MS)
– Undergraduate students
– New faculty
– Editing and reviewing manuscripts
Advancing Marketing Knowledge
15
Thank You!
 I truly appreciate this honor.
 Best wishes for a enjoyable conference and
a productive and happy 2005.
Enjoy the journey – it is a very rewarding
experience to be an educator.
Advancing Marketing Knowledge
16