We have conducted a telephone interview survey with a sample size nearer to nationally representative, then conducted the interview by calling randomly selected respondents. So the sample was not distributed proportionate to size. After assigning weight to the sample, how we can determine its margin of error and power?