Publications
Publications
- July 1999
- HBS Case Collection
Note on Statistical Tests for a Randomized Matched Pair Experimental Design, A
By: Alvin J. Silk
Abstract
Concerns understanding the conditions under which an experimental design that employs matching and randomization may result in gains in precision as compared to a design that utilizes randomization and independent samples--i.e., no matching. An empirical example is presented to illustrate a variety of statistical analyses and tests bearing on this issue.
Keywords
Citation
Silk, Alvin J. "Note on Statistical Tests for a Randomized Matched Pair Experimental Design, A." Harvard Business School Background Note 500-007, July 1999.