Experimental Assessment of TRIZ Effectiveness in Idea Generation

Abstract

TRIZ is an idea generation method that identifies conflicting principles in a given problem and then searches for previous solutions that have solved similar conflicts. How effective is TRIZ? This is an important question because it is conducive to a second question: what makes TRIZ (more, or less) effective? Understanding this provides the opportunity to improve the teaching of TRIZ, a more efficient use, and even the possibility of improving the method itself. In this paper we present results from experiments comparing TRIZ against control groups; the experiments were conducted simultaneously at three institutions: University of Texas at El Paso (UTEP), Penn State, and University of Maryland. Besides contrasting results at different institutions, the variety of experiments allows us to contrast participants at the graduate and the undergraduate level (all of them engineering students), working a design problem. In order to have significant results, we followed a rigorous experimental procedure for which a set of guidelines were produced. Such experimental guidelines address (1) Design of Experiment, (2) Execution of the Experiment, and (3) Assessment of Results. With respect to assessment, we used traditional outcome-based metrics (quantity, novelty and variety) and adapted and improved them to solve some of their issues at the conceptual (definitions) and implementation levels. Experimental results indicate that TRIZ improves Variety and Novelty of ideas generated while decreasing the quantity of ideas produced.