New Search
RIVER ROAD RESIDENTS' VIEWS OF ANNEXATION <br />Sampling Variability <br />Every survey is subject to ranges of variability, which refers to the chance <br />variation that could occur when a sample is employed, rather than a census <br />(the entire population). <br />This variability is known as "standard error", and is the difference between <br />sample findings and those which would accrue from a 100% enumeration of <br />each universe -using the same questionnaire and research procedures. <br />Ranges of sampling variability are presented below for various sample sizes, <br />computed at the 95% confidence level. The confidence level can be arbitrarily <br />set at any figure, but 95% is the convention for most market research. These <br />are maximum ranges, and most findings tend to cluster closer to the actual <br />figures as they exist in the universe. <br />Standard Error Ranges <br />Percentages Plus /Minus Range Variation <br />Close to: Total Sample (402) <br />5% or 95% 2.1 % <br />15% or 85% 3.5% <br />25% or 75% 4.2% <br />35% or 65% 4.7% <br />45% or 55% 4.9% <br />50% 4.9% <br />Example: Twenty-five percent of the sample are non-voters in the River <br />Road area. Chances are 19 out of 20 that this figure (25%) is within 4.2%, plus <br />or minus, of the result which could accrue from a complete census of River <br />Road area residents. <br />A -3 <br />