Actually the full text said:
Lest the OS/2 fans among InfoWorld's readership question our methodology, here's how we conducted the survey: We selected a sample audience on an nth name random basis among InfoWorld's subscribers, a sampling that represents 76 percent of our entire qualified base. Then we hired First Market Research, in Austin, Texas, to conduct 300 telephone interviews between May 5 and May 12, screening out readers employed by any computer-related hardware or software vendors. The sampling variance ranges from plus or minus 2.5 percent to 5.8 percent.I guess I was so blinded with rage by the time I got to the end of the last paragraph in the text that I missed that. Again, I apologize (and do so in the same forum which I made my mistake). I guess I was confused as to how sampling on a nth name random basis covers 76 percent of the entire qualified base. What in the world does that mean (I confess ignorance when it comes to stats)? I guess I was confused as to how the variance could range from 2.5 percent to 5.8 percent. Does that gaurantee at least a 2.5 percent error? I guess I was confused on how those percentages were arrived at considering the questions were open ended.
For questions or comments, email the Forums Editor