Back then we worked on a rule of thumb that it took around 20 years to grow a senior executive, so about 5% of young staff were identified and developed to provide the pipeline of leadership talent for the next generation. With such a good track record, were we gifted at spotting and developing talent? Or is there another explanation from the human instinct of classifying that more fully explains what unfolded and that we should incorporate into talent planning in organisations?
Rats in the Lab
Like all good lines of research, investigation into the impact of a leader's expectations on performance started with rats in a laboratory! In an early study back in 1964, one group of unsuspecting laboratory students was told that the rats they were studying were bred for maze-brightness (and thus this group had high expectations). A second group of lab students was led to believe that their rats were bred for maze-dullness (and hence had low expectations). The rats were in fact assigned randomly to the students. Well, in a sobering result for sophisticated talent planning, in the groups where expectations were high rats ran faster than in groups where expectations were low!
Kids in Schools
From the rat laboratory it was time to take the research to school classrooms. Students in 18 elementary (infants) classrooms were tested using a non-verbal IQ test. Twenty percent of the students in each class were then randomly labelled as ‘intellectual bloomers', the workplace equivalent of ‘high performers'. Teachers were told that students with this classification were expected to improve markedly in comparison to other students. Eight months later the tests were re-administered. Those students who were labelled as intellectual bloomers improved significantly more than the students who were not given this label. (Reference: N Kierein and M Gold, "Pygmalion in work organizations: a meta-analysis," in Journal of Organizational Behavior, 21, 913-928 (2000)).
In early high school I had a personal experience of the implications of a positive expectation from a school teacher. At the time I was neither enjoying nor doing well in history. In assessing a piece of homework, my teacher at the time, Mr Fisher, said to me, "Andrew, you're good at history!" Well, from that moment on I both enjoyed and did well at history! Thanks, Mr Fisher.
Adults at Work
The next step on the path of studying the impact of expectations on performance was to study adults in the real world. The setting was an Israeli Defence Force training camp. The course was an intensive one involving an average of 16 hours of instructor-trainee contact daily for 15 weeks.
Before the 115 soldiers entered the camp they were tested by the researchers on a range of capabilities. The soldiers were then randomly assigned to three categories of ‘command potential' - ‘high', ‘medium' and ‘unknown'. The researchers created this third category of ‘not yet classified' to add credibility to the process in the minds of the instructors and to create an impression that there was not yet enough information about these trainees.
Four days before the trainees arrived at the training camp the instructors (leaders) were provided (mis)information about the trainees. The leaders were advised of the trainees' command potential, hence creating a performance expectation in the minds of the instructors. The instructors didn't know that the classification was entirely random. Would the expectation become self-fulfilling?
The potential impact of the expectation on trainee performance was measured in three ways. The first was learning performance including knowledge of combat tactics, topography, standard operating procedures and practical skills like navigation and accuracy of weapon firing. The second was attitudinal: how much the trainee desired to go on the next course, the extent to which the trainee would recommend the course to friends and their overall satisfaction with the course. The third dimension was the leadership perceptions trainees had of their instructors.
What impact did the setting of expectations have over the 15 weeks? The results showed a substantial effect on all three dimensions. The expectancy on trainees explained 73% of the variance in performance, 66% in attitudes and 28% in leadership. Trainees whose instructors were led to expect more did indeed learn more. Trainees of whom more was expected responded with more favourable attitudes toward the course. And trainees expected to do well by the instructors had a more positive impression of the instructors' behaviour. ‘High command potential' trainees did better than the ‘unclassified' group who in turn did better than the ‘medium potential' trainees.
After the course the instructors were debriefed on the study. "The expectancy induction was so effective that it was difficult to convince the instructors that it had been random," noted the researchers and concluded that "managers get the performance they expect." (Reference: D Eden and A Shani, "Pygmalion Goes to Boot Camp: Expectancy, Leadership and Trainee Performance," in Journal of Applied Psychology, 1982, Vol 67, No 2, 194-199).
Implications for Managers
Implications of self-fulfilling expectations exist at both the manager level and the level of our talent planning systems in organisations.
For managers, if you have high expectations of your people:
1. Their performance will be higher (and vice versa if you have low expectations),
2. They will enjoy work more, and
3. They will think more highly of you (independent of your actual leadership ability!).
While in a practical sense you might relate better and have a greater regard for some your of team than for others, this inclination should be contained. If you show a greater regard for some then while that might lift the performance of those few it will likely diminish the performance of others. So for managers the tip is to have and show confidence in each and every team member. Your people will surprise themselves by doing things that they never thought possible.
Implications for Organisations
At the organisational level, the fashionable A, B, C potential grid, literally putting people into boxes, is self-destructive. Given that few people are assessed as ‘As', the system means that we are destined to have most of our people (the ‘Bs') be less effective in their roles, be less engaged and to think less of their managers.
It leaves talent planning as a dilemma. On the one hand organisations need to grow future generations of leaders. But on the other hand, once the ‘high potential leaders' are identified there is a high chance those individuals will become the future leaders for no other reason than they were expected to do so. The solution may be a) that young people all be provided the opportunity to be a high potential if they have a desire to be so, b) to provide all self-nominated people similar development opportunities initially and c) that people progress based on actual demonstrated performance in different roles. Given we have around 20 years to grow a graduate into a senior executive we have time on our hands to err on the side of a wide catchment and hold off funnelling into a select few for some years.
So, the answer to the IBM question of whether or not we were gifted at spotting talent is most likely a case of the power of classifying. Certain individuals, once classified as high potentials, were likely to fulfil that expectation. It also means that there were probably many others not so classified early in their careers who did not progress as far for no other reason than that they were not expected to do so.