an explanation of evaluation of performance over time with window analysis

window analysis is a known technique to evaluate performance during the time periods based on data envelopment analysis. in application of window analysis is assumed that efficient frontier have not changed during the periods in the study. unlike the window analysis, malmquist index tries to identify difference between efficient frontiers and separated them. but the means of different frontier in the window analysis are the their general changes that can be studied with statistical methods.while from this standpoint of malmquist, limited points of frontiers is considered that using mathematical methods is indivisible. in this study, considering conformity of production possibility set in malmquist index with window analysis with width is equal to or greater, a framework is proposed that enables the user by determining appropriate length of the window that prevents violating basic assumption, and by using appropriate kind of malmquist index, calculate the differences point in frontiers in evaluation of performance.