Description
Surveys are useful tools for original data collection, but it is imperative that you develop clear, unbiased survey questions if you want to collect accurate and useful information. A variety of methods can be used for developing surveys. For instance, many surveys consist of both open-ended and closed questions. Open-ended questions allow respondents to answer in any way they see fit. It is then up to the researcher to find themes and commonalities in these answers. To answer closed questions, respondents must choose from a predetermined set of response categories.
In addition to open-ended and closed questions, many surveys also include filter and contingency questions. Filter questions instruct respondents to continue to the next question based on how they answered the filter question. For example, if you wanted to know how satisfied homeowners are with the services they are receiving from a city, you could include a filter question asking respondents if they are homeowners. Those who answer “yes” would be instructed to proceed to the next question, which would ask them how satisfied they are with city services. If they answered “no,” respondents would be instructed to skip to the next question and proceed with the remainder of the survey. Filter and contingency questions are great ways to create a subsample for further analysis.
The phrasing and sequencing of questions on surveys also have a major impact on the quality of the original data that is generated. Questions must be clear and response categories must match the questions. Questions are usually sequenced so that “easy to answer” questions (such as age, etc.) are at the beginning and more sensitive questions (such as income) are near the end. The more questions answered by a respondent, the more likely that respondent is to complete the survey.
As you can see, survey construction is a difficult task. This week’s Discussion provides the opportunity to develop a survey of your own and receive feedback from your colleagues.
For this Discussion, review this week’s Learning Resources. Consider the program, problem, or policy you are using for your Final Project. Then, develop a 10-question survey that could be used to evaluate the program, problem, or policy in the organization you selected. The survey should include both open-ended and closed questions, and filter and contingency questions. The survey should be appropriate for the sample that you identified in Week 4.
During a polling survey, stratified sampling is the most appropriate when there exists a
well-defined population that can be divided on the context of their characteristics
necessary for the particular research (Kenett, Pfeffermann, & Steinberg 2018). This
sampling design is the most appropriate for the evaluation because it solely focusses on
the elections of one area: New Harbor, Delaware. Also, the evaluation focuses on
Councilmember Alec Coppel regarding his popularity in the area. Using the stratified
sample method, it recreates the statistical features of the population (Ott & Longnecker
2015) in New Harbor on a smaller scale. Therefore, surveyors must first divide the
population into the essential characteristics that magnify the target groups for Mr.
Coppel’s election campaign. Since his main agendas for New Harbor are a safe
community, a healthy environment, and education, surveyors will divide the
population under the following characteristics: Democrats, Independents, Republicans,
gender, the issues they most favor, ethnic background, income, and education.
Statisticians refer to these divisions as subsets or subgroups (Kenett, Pfeffermann, &
Steinberg 2018). Nevertheless, surveyors use stratified sampling to optimize group
comparisons, hence, the most appropriate method for this evaluation. If one focused on
comparing two or more subsets, equally represented in the population, one could use a
single random sample for further analysis.
One significant advantage of stratified sampling is that it does not have the errors of
random sampling because it splits a population into different distinct segments and
selects entities for each (Kenett, Pfeffermann, & Steinberg 2018). It aids in the complete
representation of the population in each sample. Since the aim of the research involves
polling in Councilmember Alec Coppel’s campaign, stratified sampling is helpful
because the analysts can utilize it in two ways: comparing between two or more subsets
of the population and the representativeness of the sample for purposes of analyzing
the population.
However, the main limitation is that stratified sampling is the most complicated
method compared to other sampling methods. Its criteria prove difficult in fulfilling it
and needs many resources for sampling (Ott & Longnecker 2015). Therefore, this
sampling method requires more effort and additional expertise for accurate results.
Consequently, there needs a practical limit to the number of subsets used and a
complete list of the population contained in each subset must be constructed. For
example, if safety is a “hot’ issue in one area in New Harbor, we may obtain
disproportionate replies from the population. Afterward, we would construct a list of
the entire population for each area in New Harbor and randomly sample within each
area proportional to its representation of the total population (Ott & Longnecker 2015).
Overall, stratified sampling is optimal from a methodological point of view as well as a
practical point of view.
References
Kenett, R. S., Pfeffermann, D., & Steinberg, D. M. (2018). Election Polls—A Survey, A
Critique, and Proposals. Annual Review of Statistics and Its Application, 5, 1-24.
Ott, R. L., & Longnecker, M. T. (2015). An introduction to statistical methods and data
analysis. Nelson Education.
Purchase answer to see full
attachment
