Sense and Nonsense:   Limitations and Cautions
in the Use of the  Behavioral Sciences

The problem

Although the behavioral sciences comprise one of the key components of modern management, caution should be taken not to overextend the usefulness of theory and research. Some critics question the soundness of most management theories and prescriptions (Lee, 1980), while others challenge the proliferation of "pop" management consulting (Micklethwait & Woolridge, 1996). Behavioral science has clearly made a contribution to the understanding of management, but there are certain cautions that one should keep in mind:

1. Much of the research and literature base of management is based on work conducted in the 1930's-60s with an emphasis on traditional organizational structures, Northern European and English speaking people, and males. Modern competition is leading to multiple and new organizational structures, gender proportion in the workplace, diverse ethnic groups and global economy that is changing demographic considerations.

Example: Maslow's hierarchy of needs, while popularized worldwide, has limited research support.  In a study of 14 countries, only the US and Great Britain appeared to follow the model. It is very clear that it does not fit the many ethnic groups of the United States, let alone the different cultures of the world that are now involved in a global economy.
2. Much research has been based on simple statistics and cause-effect modeling. The use of research design controls are intended to identify certain factors that will be examined and to exclude other factors from intervening. However, factors in the "real world" do not often function in isolation, and research findings may describe limited relationships, but not how the complex world works. Even the logic principle of "Occam's Razor"--find the simplest explanation--may not be the best representation of an event.
Example: A study of performance evaluation was correlated with "liking for the manager." Although a significant relationship was established, this dimension was intricately connected to a large number of other factors--including the difficulty level of the work performed. This finding in isolation did little to explain the complex nature of performance and supervisory relationship.
3. Statistical significance may not mean practical significance. While a finding may show that some event did not occur by chance, it does not mean that a particular measurement or factor is relevant, useful, or that a difference of a few points is meaningful.
Example: Leadership style and self-esteem may have a "significant" correlation of p>.01  (meaning that a score occurs only 1 out of 100 times due to chance), with a correlation of .25. However, if the correlation is squared (becoming the coefficient of determination), it only shows that an extremely low 6% of the total variance is accounted for--next to useless information.
4. A great deal of management research (especially master's research) is conducted on samples of convenience: pools of participants to which the researcher has convenient access. This often means that it is not randomized or stratified in selection, may not represent a larger population, and may not be free of various kinds of bias.
Example: Lippit and White's classic study of leadership styles (e.g., autocratic, democratic, laissez faire) was based, not on representative groups of managers in corporations, but on a group of 29 young boys at a summer camp. Although the design was considerably flawed, the findings have been extended to all kinds of organizations and levels of management.
5. Research is subject to bias, even unintentionally. The political and social climate of a society often shapes the research question and the perspective for viewing findings. Many researchers, rather than "trying to discover" the answer to a research question, set about to "prove" the results they anticipate--with predictable results.
Example: It is not surprising that much of the research on groups conducted in the US during the WWII and Cold War periods found that democratic procedures were most desirable--even though many organizations continued to function in a centralized, directive and autocratic manner.
6. Laboratory research, in an attempt to counter some of the above criticisms, has brought people into artificial situations that are highly controlled, in order to focus on specific behaviors. Although the concept of control is laudable, it also creates an artificial settings that often changes participant's responses.

7. Pressure for quick fixes and short term results has led some people to "pick and choose" concepts and practices without consideration for their internal consistency. The result of this patchwork quilt of interventions may mean that techniques work at cross purposes.

Example: Although Deming stated that TQM requires inclusion of all 14 of his points, some businesses have selectively taken what they like and disregarded others. In other cases, some have been adopted in clear contrast to organizational values. The US Marines in Germany decided to adopt TQM, and when questioned about the principle of "driving out fear" stated, "we will drive it out, or else!"  In spite of TQM requiring 4-7 years to successfully implement, an Air force surgical unit was directed to implement it in 18 months, "no matter what."
8. Popularity does not mean substance. There are a substantial number of books that tout management principles from Attila the Hun, Geronimo, Jesus, and others that people blithely and uncritically follow. The Tony Robbins' motivational seminars that have a strong following and generate high visibility and large revenues for him, have very little evidence to support his claims.

9. The academic environment, while it can foster critical thinking and innovation, can also foster a "publish or perish" attitude. This often means that researchers and writers will overextend conclusions, write about irrelevant topics, or fraudulently tamper with and misrepresent their data.

Example: In the late 1960's-70's when human relations training was prominent, much effort was put into "empathy training" as a communication skill. The impetus for this was the research conducted by Carkhuff, Berenson and Truax who demonstrated the impact of empathy and training effectiveness--until it was discovered that they contrived some of their data.
10. Overgeneralization accounts for much of the misuse of research. While it may be accurate for Motorola to experiment with an innovative procedure that is highly successful, it may not be appropriate for a company that is structured differently, has a different employee base, or provides an entirely different product or service to emulate Motorola.

So--what to do?

In spite of the above criticisms, the behavioral sciences have done a great deal to help identify important aspects of behavior in management, propose leverage points, and stimulate critical thinking about what works and what doesn't. The following are some suggestions about your use of the behavioral sciences:

1. Be cautious of simplistic explanations that isolate a few variables without considering the interactions among them and the influence of a larger system. New computer statistics and modeling programs are able to incorporate greater complexity, and increasingly should be able to render a better picture of what is happening.

2. When you read a finding, conclusion, or prescription, ask how it was derived. Was there an appropriate design, sampling and sample size, measurement, statistic, etc?

3. Be cautious of looking only for what you want to find. Balance your perspective by giving the other side equal time, and see what information might exist that discounts your view. Look for exceptions to the rule and for alternate explanations.

4. Consider what is necessary to really implement a prescription. Are you giving it enough time, resources, training, etc., to really work.

5. When people make claims, ask "how do you know that?" Ask for definitions, evidence, and explanations about how it works. Don't just accept it.

6.  Don't feel self conscious about using these critical perspectives and making demands on others--that's how science works--by replacing outmoded models with better ones. Neither trust too much nor too little, just be curious.

References

Lee, J. A. (1980). The gold and the garbage in management theories and prescriptions. Athens, OH: Ohio University Press.

Mickelthwait, J., & Woolridge, A. (1996). The witch doctors: Making sense of the management gurus. New York: Random House.

Links


Last revised 8-11-98
David X. Swenson Ph.D.