Please reply to [log in to unmask] We have a large database containing around 850 cases to which we want to provide a matched control sample database of 250 cases. Our problem is that the matched sample database is over represented by large firms and we need to make adjustments. Possible options include conducting more interviews with smaler firms. However, is it acceptable to adjust the matched sample to the firms size proportions in the main database? In other words we would like to proportion up our smaller firms in the matched sample from 60% to 80% (and conversely our larger firms down from 40% to 20%) in order to match our main database. We need if possible, some examples that set a precendent for this approach, or a suggestion as to what we ought to do. Thanks Regards Kevin Mole Centre for Interdisciplinary Strategic Management Research (CISMR) Middlesex University Business School The Burroughs, London NW4 4BT Tel +00 44 (0)20 8411 5570 Fax +00 44 (0)20 8411 5822