1999 Volume 14 Issue 2 Pages 2_73-2_88
“Laboratory assignment” is a process to assign students to laboratories for their graduation research. It assumes the following conditions: every student should belong to one and only one laboratory, the maximum number of students for each laboratory is fixed and announced in advance, and the process of assignment depends only on the preferences of the students and teachers involved. Under this assumption, we consider a laboratory assignment institution for each student to be able to belong to some laboratory. An extended version of the matching institution which was proposed by Gale and Shapley (1962) not only satisfies the conditions required but also has some theoretically desirable properties. But this institution assumes, among others, that every student should, by preference, linearly order all the laboratories from which to choose only one for his/her graduation research, which, though theoretically valid, is not realistically plausible. To put it into practice, it should be examined whether assignment can be successful when the number is relatively small of the candidates a student should choose first from all the laboratories. Analysis by Monte Carlo simulation reveals that this institution works well even in such situations, given a certain set of conditions on the maximum number of students for each laboratory and the number of the candidates the students should submit first.