Design a survey Homework Help

Topic: design a survey

Order Description

1. Make sure you know how to design a professional survey
2. Choose an appropriate topic
3. Follow the method of how to design a survey

MOCK SURVEY RESEARCH PROPOSAL GUIDELINES
The goal of this project is to help you integrate and apply the survey methods covered throughout the course. The assignment challenges you to design a hypothetical study on a topic of your own choosing and present it in the form of a research proposal. The level of detail of your “mock” proposal will be modest in comparison to an authentic one, but the sections are roughly the same. Most significantly, the Background and Relevance and Research Study Plans sections will be relatively short and your survey questions will be limited to a single topical module.  With the exception of the demographic section that generally appears at the end of a survey, you may select any module you like. This restriction does NOT apply to substantive modules dealing with demographic patterns or processes (e.g., a section on marital or fertility histories). The major sections of the mock proposal are outlined in Figure 1. Figure 2 shows the links between the mock proposal sections and the steps of scientific survey research.
Figure 1. Proposal Sections and Information to Include

.Figure 2. Major Steps in Survey Research Process

StepsSection in Mock Proposal
1. Establish clear research question and specific purpose for the study
*Determine if a survey is the best strategy for answering the research question
A. Introduction
and
B. Background/Relevance
2. Develop a comprehensive study design
?    Choose mode of administrationC. Research Study Plan?    Determine desired level of precision sampling strategy (e.g., probability, non-probability)C. Research Study Plan?    Select sampling strategy and identify sampling frame?    Develop plans for analysis (e.g., descriptive, inferential statistics)* Optional (Section C)3. Formulate conceptual model – identify key constructs and operational definitions C. Research Study Plan4. Based on operational definitions, create survey questions that will yield necessary variables – test for validity and reliability E. Survey Instrument
(Include only ONE module)5. Create well-designed questionnaire or interview script E. Survey Instrument
(Disregard lay-out)6. Develop schemes for data capture, naming, and coding variables C. Research Study Plan
Disregard variable names and codes7. Pre-test and revise the instrument as many times as needed C. Research Study Plan8. Devise effective communication protocol (pre-notification, follow-up, etc.)C. Research Study Plan  (Include only cover letter)9. Deploy survey, carefully track returns, and perform follow-ups as necessaryC. Research Study Plan10. Create data base of responses, clean and edit data, and prepare analysis file C. Research Study Plan11. Analyze and interpret the results D. Deliverables12. Communicate the findings with full disclosure of methodology, response rates

AIDitional Details for Selected Sections of Proposal
A. Introduction
Identifying a Problem or Issue for Investigation
The idea is to identify an issue or problem of concern to the general public, scholars, policy makers, opinion leaders, or other groups. The focus of the problem could be local, global, micro, or macro in nature. It could be a problem of considerable and far-reaching consequence, but for this assignment it would be wise to choose something more narrow.
Example: Children in low-income families face health risks due to poor nutrition.

Moving from Problem to Research Question
The goal is to translate the general concern into a specific research question that is empirically testable with measurable results. Moreover, the underlying concepts must be quantifiable in a fashion that is valid and reliable.
Example 1: How does nutrition education help children?
It is not difficult to imagine sensible ways that “nutrition education” and “helpful” to children could be quantified.
Example 2 (more specific): What are the major health risks related to diet for school-aged children?
Similarly, medical and nutritional science offer very practical indicators for operationalizing “considerable health risks” and “diet.”
Nonetheless, it is often challenging to formulate a “testable” research question. To spur ideas, consider the general categories of research questions presented in Figure 3.

Figure 3. Selected Types of Research Questions

B. Background and Relevance

The intention of this assignment is not for you to write a scholarly research paper, but rather to demonstrate your ability to apply practical tools of survey research to a hypothetical research problem. Thus, in Section B make a logical and convincing case for the importance of your investigation and the suitability of conducting a survey. This can be written in one page or less and requires minimal reference to theory or existing research.

C. Research Study Plan

General Tips on Style of Writing

In homework exercise it can sometimes be strategic to explain your rationale for certain assumptions and choices. For example, consider this hypothetical homework question and a student’s response.

Question: What sample size would the investigator need to achieve his stated research goals?

“Homework” Answer: The first things to consider when determining an appropriate sample size are the purpose of the study, the size of the target population and the resources that are available. After those are taken into account, sample size should be chosen using three criteria: 1) the desired level of precision, 2) the level of confidence that the researcher wants to achieve, and 3) the degree of variability in survey responses. Given that the investigator in this case wants ________, I would recommend a sample size of ____.”

In your mock proposal, avoid being that didactic. Instead, assume that the hypothetical reviewers of your proposal are knowledgeable about survey methods and “best practices” in survey research.

Published Description of Sample Size for a Child Health Survey:
“The goal of the NSCH sample design was to generate samples representative of populations of children within each state. An aIDitional goal of the NSCH was to obtain state-specific sample sizes that were sufficiently large to permit reasonably precise estimates of the health characteristics of children in each state. To achieve these goals, state samples were designed to obtain a minimum of 1,700 completed interviews. The number of children to be selected in each NIS estimation area was determined by allocating the total of 1,700 children in the state to each NIS estimation area within the state in proportion to the total estimated number of households with children in the NIS estimation area.” (Blumberg SJ, Foster EB, Frasier AM, et al. Design and Operation of the National Survey of Children’s Health, 2007. National Center for Health Statistics. Vital Health Stat 1, Forthcoming)

A Study Plan Begins with a Conceptual Framework

In our formulation of research questions, we generally regard one concept as an “outcome” and one or more aIDitional concepts as sources of influence or explanation (Hence the common reference to dependent and explanatory variables). To conduct a quantitative analysis we have to be explicit about the interrelationships we hypothesize.  A traditional way to articulate those assumptions is with a conceptual framework or model. Conceptual models can be expressed in a number of ways including through words, diagrams (See Figure 4), and formulas.

Figure 4 is a simplified representation of the factors associated with marital disruption and child well-being because many other explanatory factors have been identified in the literature as well (e.g., acrimony related to custody arrangements, increased parental depression, etc.) As is generally true in social science, there is an almost an infinite number of factors and processes related to the outcomes we study. Thus our conceptual models are always incomplete. For this assignment, keep things particularly abbreviated.

PLEASE NOTE:  I have included Figure 4 as a heuristic device only. While you may find it helpful to sketch out a model like this while formulating your research plan, you are NOT required to include a conceptual diagram in your proposal.

Be Clear about Your Unit of Analysis

Recall that the unit of analysis is the type of entity being studied. Units of analysis can take many forms: individuals; couples; groups of individuals, such as a classroom of students; neighborhoods; agencies; organizations; countries and so on. As you work out the details of your conceptual framework it is essential to be absolutely clear about which unit of analysis is appropriate for each of your concepts and to ensure that they are compatible.

Selecting Topical Modules for Your Survey

Use the logic of a conceptual framework to guide your choice of topics to include in your survey.  As we’ve discussed, it is advisable to split long surveys into separate sections or modules, each containing groups of substantively related questions (i.e., questions designed to capture the same or related concepts).

To keep this assignment manageable, limit the number of modules you include in your mock survey to no more than 6, including one for a standard set of demographic questions (i.e., 5 topical modules + Demographic module = 6 total).

Many times it makes sense to create separate modules for each of the concepts depicted in your conceptual model, but not always. The primary goal in the grouping and ordering of survey questions is to facilitate the respondent’s successful completion of the survey.

Figure 5 shows several web sites that provide “summaries” of topics/modules for surveys. These may help you determine your own modules.

Figure 5. Descriptions of Topical Modules within Large Surveys

Identifying Indicators

Although you are only required to provide survey questions for one of your modules, identify the indicators that would form the basis of the questions you would include in all of your modules. The list of items for each module should be illustrative, but do not have to be exhaustive.

If you plan to use a well-established test, inventory, etc. within your key module, note the psychometric properties if available. Describe any notable risks of survey error.

Sampling plan
There are many acceptable methods of sampling, each with advantages and disadvantages in specific situations.  Remember that a sample that adequately represents your universe, is a vital requirement of high-quality, scientific research.  Assume that you are submitting your proposal to an organization or agency that is prepared to offer sizable funding. That is, do not worry about the expense of anything you propose to do.
Provide sufficient detail about your sampling plan to allow clear judgments about your sampling strategy and research question. This section should include:
•    Definition of the universe that the survey is intended to represent and any relevant sub-groups

•    Sampling design (e.g., method of selecting sample elements, qualifying/ disqualifying criteria, methods for making selections within households, etc.)  Note any exceptional challenges. DO NOT worry about statistical details, such as how to select strata, over-sample, find terrific sampling frames, and apply weights. It is sufficient to say, something along the lines of this example:

“I will use stratified random sampling procedures to allow for State-level estimates of immunization rates among school-aged children.”

•    Targeted sample size.

•    Foreseeable practical challenges specific to your sampling strategy and your plans to aIDress them.

•    Notable sources of potential survey error

RESOURCE:
Clearly presented Power Point slides on Sampling and Sampling Strategies: http://www.scribd.com/doc/2034905/Research-Methodology-Part-5-Sampling-Sampling-Strategy-or-Plan

Procedures for Administration of Survey

This section should detail the major tasks involved in the actual administration of the survey. Among the details you should include are:
•    Protocol for recruiting respondents

•    Plans to contract with any third-parties to perform specific functions for you?

•    Potential risks to participants in your survey and the steps you will take to minimize or eliminate them. Discuss how you will aIDress issues of confidentiality and anonymity.

•    Any issues related to data entry.  NOTE: For the purposes of this assignment, assume that paper questionnaires will be scanned into an electronic data file and other modes will use computer-aided technology for capturing data.

•    Mechanisms for logging in returned questionnaires

•    Steps in preparing analysis file, etc.

•    Description of any incentives you plan to use.

•    Your cover letter

D. Deliverables

Let us assume that among other requirements, the organization or agency from which you are seeking funding expects grantees to:

1.    Submit a Summary Report at the end of the conclusion of the study

2.    Submit survey data files and documentation to be archived and made available for public-use  (See Student Proposal from 2008)

OPTIONAL: Feel free to come up with aIDitional ideas for deliverables that would be suitable for your study (e.g., presentation of results at professional meetings, scholarly articles, press release, Town Hall meeting to share results with interested constituents)

E. Survey Instrument

For one module of your choosing, provide 12-15 questions. At least 5 questions must be your own.
When possible, draw upon existing, credible surveys for appropriate questions and inventories (See Figure 6). Judicious use of items drawn from prior scientific surveys generally enhances, rather than undermines the credibility of your study.  Many items used in large-scale, government sponsored surveys have attractive psychometric properties, documented across multiple studies (which you should report if available) that often make them superior to an “un-tested,” ad hoc questions developed by individual investigators. Provide complete information on each secondary data source you use.

Figure 6: Survey Question Banks

F. Bibliography

List of full sources for secondary material cited in the text.  Use the professional citation format of your choice.