Homework answers / question archive / Survey Proposal The assignment is worth 35% of the final course mark

Survey Proposal The assignment is worth 35% of the final course mark

Statistics

Share With

---

Survey Proposal

• The assignment is worth 35% of the final course mark.
• The project will take you about 20 hours, most of which should be spent developing the necessary parts of your proposal.

Objectives of this assignment

• To put into practice the theory of survey methodology;  
• To experience the challenge of developing objectives into a research proposal that works in practice;
• To appreciate the limits of your knowledge of Survey Design.

 

The Task

• Write a proposal for a Survey.  
• The proposal should be practical and realistic - one which could feasibly be funded and carried out.
• It should incorporate a probabilistic sample design based on a well defined sample frame so that weights and valid sampling errors could be calculated.
• The proposal should be clearly written and be self-contained and should be set out in a logical order.
• The proposal should be 8 pages at the most, in a font size of 11pt or larger
• You may not submit a report containing work you have submitted for any other course. If you are repeating the course you may not resubmit a survey proposal from a previous year, even if it is edited. Please see Richard Arnold if you are unsure about what you intend to submit.

 

• The proposal must include the survey cover sheet (at the end of this document) as its FIRST page [BEFORE any title page]
• All assignments must be submitted electronically via Blackboard, and will be scanned with the Turnitin software.  Instructions about how to submit will be given in lectures.
   

 

Marking Scheme

 

The marking scheme outlined below is a guide. Clearly in some proposals more thought and effort will be needed in some aspects of the proposal (to satisfactorily address the issues) than is indicated in this average marking scheme. Also the list here is not meant to be exhaustive, or relevant to every proposal. It is given to jog your memory.

 

At the outset you should clearly identify:

• the sponsor of the survey (who is paying?)
• the organisation who will carry out the survey;
• the principal stakeholders of the survey – people or organisations who would be interested in the results, and who would be consulted about the survey objectives and development
• the audience for your survey proposal: the proposal would normally be written by the organisation carrying out the survey for the sponsor;
• the users of the results – this includes the principal stakeholders, and other interested parties.

[Note - It could be that the sponsor is also the organisation which is carrying out the survey, and is also the only stakeholder]

 

You are not expected to construct a frame, but you should indicate what your frame is and what steps you would need to construct it, or get access to it if it exists. (NB: don't simply assume that organisations will provide lists of population units - you may have to imagine that you ARE a certain organisation in order to get access to the data required to construct a frame.)

 

You are expected to show how the sample design will be implementable from your choice of frame and to set accuracy requirements and determine the sample size required to meet these requirements.

 

You are not expected to develop a questionnaire in full, but you should indicate what questions and/or topic areas are required to get the information needed to satisfy your objectives and any problems you see with these questions.

 

You should include a section describing the contents of a report which would come out of your survey. It should state the main headings in the report, the intended audience, and may mention decisions that may flow from the survey.

Write the proposal as a document that is written by the collector and addressed to the sponsor.  Do not write it as a standard statistics assignment where you answer a set of questions.  Only include information that is relevant to your particular survey.

 

 

PLEASE NOTE THE FOLLOWING:

• You must have a practicable sample frame identified;  
• You must develop a survey with a probability sample;
• Your sample cannot be a census or just a simple random sample (i.e. propose a stratified, cluster or some other complex design);
• You must include a sample size estimate;
• You CANNOT propose a survey about University students!
• If your proposal is for a survey that is similar to an existing survey, you MUST acknowledge the fact, refer to a description of that existing survey, and show that you are doing things differently.   Talk to Richard if you’re unsure:  your work must be original, but some surveys necessarily cover some ground that other surveys have covered.
• Take care that your proposal is not too general - i.e. don't talk in generalities about sample design, instead focus everything around the proposal you are writing. In particular DO NOT copy out chunks of the lecture notes without connecting the material to your survey.
• Make sure you write in formal language (i.e. not in a conversational style)
• Make sure you define all symbols used in any equations you include.
• Your Survey Proposal should follow the broad structure of sections 1 to 5 given below.

 

 

1. Introduction, Background, Objectives [5 marks]

• State the reasons/motivation for and background of the survey
• Identify the sponsor, surveyor, stakeholders and users of the results
• State the objectives/research questions of the survey
• State any key definitions of concepts specific to the survey
• Make sure you reference appropriate background information
• Identify overall limitations on the survey
  • Are there any alternatives to doing a survey?
  • Are the results needed by a certain date? Why?
  • Any limitations on resources; time; ethical issues; confidentiality?

 

2. Sample Design [10 marks]

• Propose a sample frame; identify the survey population
• Comment on scope and coverage, and the quality of the sample frame
• Identify the source of the frame – where will the information for the frame come from?
• Identify any auxiliary data on the frame;
• How will the sample be selected? 
• State the design variable and design estimate that you will use for your sample size calculation, and any prior estimates you will use;
• State the required accuracy, and estimate the sample size required – including adjustments for nonresponse;
• Show how sample weights will be computed;
• Discuss any limitations of the chosen sample design and estimation method.

 

 

3. Questionnaire/Survey Instrument Design [5 marks]

• Identify the survey instrument and mode of data collection
• Identify the principal topic areas and give a brief summary of the data that is to be collected, including e.g. the order of topic areas on the questionnaire;
• Discuss the nature of the data to be collected (measurements? Responses to questions put by an interviewer? Open/closed questions? Probes allowed? Proxy responses allowed?)
• Discuss questionnaire development and testing

 

4. Fieldwork [5 marks]

• Describe the full sequence of events that take place during data collection
• How will the selected units be contacted/sampled?
• What are the noncontact/nonresponse follow-up procedures?
• How will you allocate the sampling units to interviewers/observers?
• When will the data collection take place? What is the reference period?
• How will data capture be done?
• What edit/consistency checks and other quality control measures will be in place?
• What interviewer training is required?
• How will you control the flow of documents to and from the interviewers or respondents? What about long term storage and destruction of documents?
• Comment on any sources of survey error in the data collection process

 

5. Estimation, Analysis and Reporting [10 marks]

• What post-survey adjustments will be made?
• If post-stratification is to be done, which benchmarks will be used?  Are these data available?
• If imputation is to be done - then on which items? how? why? what assumptions and what risks are associated with your proposed approach?

• What tables, summary statistics will be produced? (you might include some example tables and graphs: empty of data, but showing the types of outputs to be expected)
• What will the final report contain? (May give a table of contents)
• What subgroup analyses will be done?
• How will levels of uncertainty be shown in estimates?
• Show how the outputs will be used to meet the survey objectives.
• What biases, if any, do you expect and why?
• State the consequences of non-sampling errors – don’t merely say that these errors exist.
• If more complicated analysis is being done, how have you incorporated the survey design into your analysis?
• What limitations are there on the interpretation of the estimates?
• How will the results be disseminated?

 

In addition 5 marks will be awarded for overall presentation, clarity and overall quality.  In particular marks will be awarded for

• A clear logical layout
• Keeping to the page limit (8 pages)
• Key facts being easily located
• Clarity of expression [Note: for non-native speakers of English: your English does not need to be perfect, it is the logic and correctness of your presentation that is most important.  Nevertheless you are advised to get someone to proof-read your proposal first.]
• Clear explanation of the principles of survey design and analysis – with content being specific to your proposal (i.e. not just a set of generalities)
• Referencing should be correctly done: a complete list of references must be included.  You can use any referencing style you wish: APA is fine if that’s what you like.
• Make sure each time you use/refer to someone else’s work you cite the source in the text, and include the reference in the list at the end.

 

Total marks: 40
Citation of sources

 

In order to avoid plagiarism you must take care to cite all sources correctly.
This includes material from the internet just as much as printed books and articles.   Make sure you cite sources at all appropriate places in the text, with  a full reference to every source in a bibliography at the end.  You can use any referencing style you wish (APA is fine).

 

 

If you take an image or a diagram that you did not create yourself, the source needs to be acknowledged.

 

You are permitted to take material from the lecture notes without citation, however DO NOT simply copy out material from the notes – rewrite it in your own words, to fit the situation you are describing in your assignment.

 

It should always be totally clear what is your own work, and what you have sourced from elsewhere.

 

Note that your proposal will be submitted via Turnitin, a computer program which checks for and detects copied material – including material copied from all assignments ever submitted to STAT392/439.

 

You can obtain advice on writing from Student Learning Support, or from Richard.

In the StudyHub section of their website Student Learning Support have a set of useful documents about writing assignments:
https://www.victoria.ac.nz/student-learning/studyhub/writing-assignments

 

and a set of generally helpful handouts

https://www.victoria.ac.nz/student-learning/study-resources

(there’s one about avoiding plagiarism – make sure you understand what plagiarism is)

 

 

Working Together

 

It is quite acceptable to work together to develop your ideas about your proposal.   You are encouraged to get someone else to read your proposal before you submit it, to make sure it is clear.

 

However your proposal must ultimately be your own work, which you go away and write yourself – on your own.

 

Make sure you choose a topic which is quite different from everyone else’s. 

 

 

Student Name
Student ID
Survey Title
Country
Sponsor
Collector
Stakeholders
Purpose
Target Population
Sampling Frame
Sample Design
Coverage
Sample Size
Use of Interviewer
Mode of Administration
Target Response Rate

 

 

 

 

Your proposal should be of greater scope and detail, demonstrating a deeper and more extensive knowledge of the subject.

 

Students taking STAT 439 should note:

• In the marking scheme an additional 5 marks will be allocated to each of the following sections
  • Sample Design
  • Analysis and Reporting
• The discussions in these two sections should include a more detailed treatment of the sample design and the analysis possibilities. You should consider a wider range of sampling options as part of the justification for your chosen strategy.
• You might consider simulating some data, or working with existing data from another survey to show the implications of those different sample designs.
• Likewise your analysis proposal should address specifically any statistical techniques you intend to use.

 

 

All students taking this paper at advanced level should discuss their proposals and what possible additional work to include with Richard well before the end of term.