CRPC-TR95553-S: An Empirical Evaluation of Dependence Analysis in 
Parallel Program Comprehension

Douglas Monk (PhD. Thesis)

May, 1995

This thesis describes an extensive ground-breaking empirical study of 
the effects of dependence analysis and related factors on error severity 
and time required for successful comprehension of loop transformations 
relevant to both sequential and parallel programming. The results 
indicate that annotating code with graphical information about 
dependences as is done in the ParaScope programming environment is a 
good approach, and that different types of dependences have different 
effects on error and comprehension time. The use of Xbrowser, a 
hypertext system combining high quality text formatting using TeX or 
LaTeX with support for computer-assisted instruction and protocol 
analysis, is also extensively described.