SMD analysis

The IPPO representatives and topic/subtopic managers who constituted the decision-making panel for NASA's Science Mission Directorate (SMD) opted largely to retain its process of selection via discussion and voting, and to give very high weight to a policy of equitability, i.e., selecting a robotics technology, then a sensor technology, then a telescope technology (in rough proportion to the number of technologies proposed for each discipline), then repeating.

A program manager at JPL who had participated in the SMD panel discussion graciously agreed to attempt to reverse-engineer the panel's selections by compiling a table that evaluates each candidate technology according to a set of attributes on which he thought the panel had implicitly based its recommendations. However, the START team was able to demonstrate mathematically that this table could not reproduce the panel's results with any possible combination of non-zero attribute weights. This illustrates the difficulty in explaining after the fact the reasons for the choices that have been made, and why the START process begins with attributes and weights and derives the selections from them.

We were also able to show that the START-Lite analysis tool could incorporate a constraint of equitability if desired, allowing the decision-makers to retain that policy while still implementing a quantitative, traceable process for assembling an optimal portfolio.

Back to START Lite main page

HRP study

ETDP study