Laserfiche WebLink
<br /> l3 <br /> . then obtain a listing of households in Arden Hills. We recommend drawing an equal <br /> probability random sample of 400 households from this listing using computer- <br /> generated random digits. <br /> Drawing a random sample will ensure that the data will be generalizable to Arden <br /> Hills residents as a whole. This sampling strategy will also ensure that each household <br /> in Arden Hills has an equal chance of being selected for the survey, resulting in a <br /> statistically valid cross section of Arden Hills residents. Furthermore, a sample of this <br /> size will yield data that are accurate at greater than :t 5 percent at a 95 percent <br /> confidence level. <br /> III. Questionnaire Design <br /> Anderson, Niebuhr & Associates, Inc. will communicate closely with City <br /> . representatives during the design of the questionnaire. Experience has shown that <br /> close contact with clients is the most effective way to ensure that appropriate <br /> questionnaires are designed. <br /> Development of valid, reliable, and appropriate questionnaires depends on the <br /> successful completion of the following four activities: (1) identification of issues and <br /> information that need to be addressed, (2) construction of reliable and valid questions, <br /> (3) construction of a questionnaire format appropriate for the particular study, and (4) <br /> pretesting of the questionnaire. Each of these phases is described below. <br /> A. Identification of Issues and Needed Information <br /> We will meet with City of Arden Hills representatives at the beginning of the <br /> project to identify the topics to be addressed in the questionnaire. We will work closely <br /> with City representatives throughout the questionnaire design phase of the project to <br /> . ensure that the questionnaire meets the City's needs. <br />