Laserfiche WebLink
Roseville Parks and Re(:reation <br /> We understand there may be a preference for a telephone survey, and that can be an appropriate <br /> choice, particularly when there is a tight timeline. Data collection can take several weeks longer <br /> than mailed surveys. <br /> Task 2.* Selecting Survey Recipients <br /> In deciding what survey mode to use, we will want to consider the quality and accuracy of the <br /> "sampling frame," a term used by survey researchers to describe the source from which households <br /> will be selected to receive the survey. In general, a mailed survey allows the most accurate <br /> targeting of households within the specified geographic boundaries for the study. A mailing list can <br /> be geocoded so that every address included is determined to be within the study boundaries. <br /> For a telephone survey, we recommend the use of random digit dialing. This allows the inclusion of <br /> listed and unlisted phone numbers. Phone number prefixes do not match exact geographic <br /> boundaries, so we would likely need to include a screening question to determine whether the <br /> household is within the study boundaries.3 <br /> Alternatively, if RPR wishes to only hear from registered voters, we can use the voters list from the <br /> elections division. This list can be used for either telephone and mailed survey data collection. <br /> Voters are not required to provide a telephone number, but we find that generally about two-thirds <br /> of the voters will have a telephone number as a part of their record. <br /> In the Project Cost Estimates section, we provide cost estimates for each of these options. We will be <br /> happy to discuss further the pros and cons of each approach with RPR. <br /> All households or voters located in the study boundaries would be eligible for the survey. We <br /> would systematically sample from all eligible households or voters to select the group that would <br /> actually be called or mailed the survey. Systematic sampling is a procedure whereby a complete list <br /> of all possible items is culled, selecting every Nth one until the appropriate amount of items is <br /> selected. If desired, the sample could be stratified by geographic location; that is, samples of <br /> desired size could be selected for various areas within the study boundaries. <br /> Selecting a within <br /> In cases where more than one household member is older than 18 years and eligible to participate <br /> in the study, a procedure called"the birthday method"is used to select the survey respondent <br /> without bias. The birthday method requests that the respondent be the adult (18 or older) in the <br /> household who most recently had a birthday (irrespective of the year of birth). This methodology <br /> helps ensure the attitudes expressed by our respondent"sample"closely approximate the attitudes <br /> of all residents or voters living in the District. We use this method for both mail and phone <br /> household surveys; in a mailed survey the instructions are included in the cover letter while in a <br /> phone survey the interviewer will ask to speak to the adult in the household who most recently had <br /> 3 While telephone numbers are originally assigned based on some broad geographic areas,telephone number"portability"means <br /> that some people may now have telephone numbers from a different geographic area. <br /> P v <br />