Daniel Kahneman's and Amos Tversky's paper On The Reality of Cognitive Illusion. ‡ They suggest, through their research, that intuitive predictions and judgements are often mediated by a small number of distinctive mental operations, called judgement heuristics. These heuristics often lead to characteristic errors and biases.
For example, the effect of aerial perspective on apparent distance is confirmed by the observation that the same mountain appears closer on a clear day rather than a hazy day. The intuitive predication and judgement of probability are often based on the relations of similarity between evidence and possible outcomes. This representativeness is an assessment of the degree of correspondence between a sample and a population.
The next heuristic is the availability bias in which the probability is estimated by assessing availability or associative distance. † Experience shows and experiments confirm that large classes are recalled better and faster than instances of less frequent classes. That likely occurrences are easier to imagine than unlikely ones. And associative connections are strengthened when two events frequently co-occur. That these associative bonds are strengthened by repetition is the basis of memory.
So Here's the Issue
When we hear or read that software projects fail often or Standish report says ..., or a personal anecdote that resonates with our own personal experience, we recall that experience from memory. The actual data from the population of all data are not used for comparison. Rather we assume - by applying the cognitive illusion - that the sample sata represents the large class of population data, since our repeated observations of the sample data class has reinforced our illusion that that sample data IS the population.
This is the core issue with anecdotal information when making decisions in the presence of uncertainty. Or speaking about a condition in the absence of statistically testable hypothesis. Or attempting to convey a message in the absence of external confirmation that the message is on solid footing compared to the population of data.
Why This Is Not Good Management
When we hear we're all bad at making estimates, in the absence of actual population statistics about estimate making, we're using Cognitive Illusions and Availability Heuristics. Because we have personal experience with making bad estimates and the majority of people we associate with have the same experience.
This experience is in no way representative of the population of people tasked with making estimates. This would be irrelevant of course if the conversation were simple chatter at the bar. But once that conversation enters the realm of policy making, method development, or suggestions that the anecdotal observations need to result in changing how business conducts its business - we're bad at making estimates so the solution is to stop making estimates - then both availability bias and Cognitive Illusions have displaced the actual conversation about the very validity of the anecdotal concepts. And it is replaced by strong defense of the cognitively biased dea, no matter the credibility of the concept - which is most often weak at best and simply false at worse.
So next time you hear some statement about something involving observational and anecdotal data, ask a simple question.
What's the process by which these anecdotal observation have been tested in the broader population of conditions?
This is the core issue with the Standish Reports. They are self selected samples of projects that are troubled in the absence of the population of projects that are troubled and not troubled.
Always ask for references, data representative of the references, and an assessment of the statistical confidence that the anecdotal data is in fact correlated with the population data. Otherwise it's just an opinion, and very likely an uniformed opinion.
And if you're paying money to listen to someone tell you ancedotal data and don't speak up and ask those questions, you've participated in the availability heuristic and cognitive illusion along with the speaker.
† Availability: A Heuristic for Judging Frequency and Probability, Amos Tversky and Daniel Kahneman, a chapter appearing in Cognitive Psychology, 1973
‡ On the Reality of Cognitive Illusions, Daniel Kahneman and Amos Tversky, Psychological Review, Vol. 103, No. 3, pp. 582-591
