IBM Israel Research Seminars

"The problem with the future is that more things might happen than will happen". Plato
"It's tough to make predictions, especially about the future". Yogi Berra (1998)

Research on projects' success suggests that poor risk assessment is one of the major reasons for projects failure. Projects' success can be predicted by the validity and accuracy of project risk assessment decisions.

The combination of limited human knowledge and cognitive capacity, coupled with the need for decision making, which is characterized by high levels of complexity, ambiguity, and uncertainty, leads project managers to use heuristics in making decisions. These heuristics are sometimes subject to a set of subconscious, systematic errors known as biases that directly affect decisions. A better understanding of decision making biases has the potential to improve the effectiveness of risk assessment practices, manifested in project management success criteria such as time, cost, technical performance and satisfaction.

An examination of the decision making processes underlying risk assessment, can bring insight to some of the potential biases, deriving from cognitive effects, or other sources of bias (e.g. environmental biases).

In this talk, an on-going research work on risk assessment will be presented. The objectives of this work are: 1. Identify different sources of biases in risk assessment and their effect on risk estimation/planning in projects. 2. Enable the generation of better strategies for coping with project uncertainties.

An understanding of cognitive effects on risk assessment may guide project managers to better design their decision making process, allowing a mitigation of "cognitive risk sources".

As part of this research activity, a survey was carried out at the IBM Research Laboratory in Haifa, with the purpose of exploring the effects of risk psychological distance on the timing of risk mitigation and perception of difficulty in real software projects. The results of this survey and some practical conclusions will also be presented during this talk.