| Variations in results always come from differences in sampling procedures, data weighting, question design, time of launching the data collection, data collection methodology..... I can go on and on. However, the most important aspects revolve around the sampling frame and data weighting. Data weighting is important in order to draw conclusive results - especially when you want to report at a National level. A badly designed sample of 3000, 5000 or even 10000 does not necessarily give you representative information that you could statistically draw correct conclusions on.... all you can ever do is to make inferences. Revisit Research 101 for back to basics on research.My 2-cents |
LEGAL DISCLAIMER: This Message Board accepts no liability of legal consequences that arise from the Message Boards (e.g. defamation, slander, or other such crimes). All posted messages are the sole property of their respective authors. The maintainer does retain the right to remove any message posts for whatever reasons. People that post messages to this forum are not to libel/slander nor in any other way depict a company, entity, individual(s), or service in a false light; should they do so, the legal consequences are theirs alone. Bizcommunity.com will disclose authors' IP addresses to authorities if compelled to do so by a court of law.
