Data Bias

While collecting data, the estimator needs to be aware of any potential biases within the data. In particular, data collected through subject matter expert interviews have the potential to be biased.

Bias can originate from different sources, such as over-optimism, group think, dominating personalities, inexperience, or pressure from management. For example, motivational bias is a source of bias that arises when interviewees feel threatened (whether justifiably or unjustifiably) if they give their true thoughts about a program. This threat is typically from fear of being punished by someone in authority. Confidence bias occurs when the estimator is overly optimistic about the success of the program. Other forms of bias include a tendency to give more weight to recent events than earlier events, to assume patterns where none exist, and to assign connections to coincidences.