Research through the latter part of the 20th century established that real world decision making was not sufficiently described by a rational-agent model. The decision making process is susceptible to many individual cognitive biases and heuristic approaches. Aspects of the decision making process include thinking (rational, cognitive), feeling (affective, motive), and behavior (habitual, customary, traditional).

System engineering decisions must weigh performance, cost and schedule as well as other organizational factors, including political considerations. The knowledge base for the performance aspect may come from the decision makers' own observations, if possible, and from the developer's progress reports. Knowledge may also come from third party "independent" evaluation, to obtain a more objective view, though usually at some additional cost. These costs can vary depending upon how the independence is structured, but some of the ways the evaluation structure is varied may affect the meaning of independence for the decision maker, and hence the weight and value of the independent evaluation.

This dissertation establishes a measure for the connotative meaning of evaluation to decision makers. A complex set of scales is used to rate different evaluation scenarios, and factor analysis is employed to identify the connotative meaning of independence to the research participants across the scenarios. The resulting factor scores are used to compare differences across three roles involved in the evaluation scenario to determine the extent to which there is shared meaning among these stakeholders (Decision makers, Developers, and Evaluators).

The findings indicate that: (1) The relative value of independent evaluation alternatives can be measured. (2) Independence is viewed through the same lens by diverse parties to evaluations. (3) Managerial independence is more important than Technical or Financial independence (using IEEE standard definitions).