1 / 42

Off-Campus Housing Instrument (OCHI)

Off-Campus Housing Instrument (OCHI). Jake Kazensky , Allison Moore, Kaitlin Onthank , Kevin Reynolds, Rebecca Rosenberg, Andrew Wanamaker. STAT 4524 Survey Sampling Methods December 12, 2011 Lemons. Instrument. Name: The name of our survey is the “ Off-Campus Housing Instrument”

lotta
Download Presentation

Off-Campus Housing Instrument (OCHI)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Off-Campus Housing Instrument (OCHI) Jake Kazensky, Allison Moore, Kaitlin Onthank, Kevin Reynolds, Rebecca Rosenberg, Andrew Wanamaker STAT 4524 Survey Sampling Methods December 12, 2011 Lemons

  2. Instrument • Name: The name of our survey is the “Off-Campus Housing Instrument” • Customers: VT Department of Student Life, respective apartment complexes, freshman and transfer students • Acronym: “OCHI” (pronounced ‘OX’) • Type: Terminal • Administration: October 2011 to December 2011

  3. Background • Administered by the Students for Off-Campus Research (SOCR). • Customers of our research include the following: • VT Student Life • Respective apartment complexes • Freshmen and Transfer students

  4. Mission Statement • The SOCR aims to quantify and summarize attributes that contribute to the off-campus housing experience at Virginia Tech. In conducting our qualitative and quantitative studies , utmost care is placed on the confidentiality and privacy of our subjects as well as the data and results produced.

  5. Purpose • The purpose of this survey isto quantitatively and qualitatively measure the motivating factors behind a person’s decision to live in an apartment complex. Specifically, this will be measured by the satisfaction levels of residents (as they compare to other apartment complexes) as well as how transportation factors affect a resident’s decision. To supplement further, the order of the survey will be observed.

  6. Survey Specifics • Population: All Virginia Tech students eligible to live off-campus. • Sampling Units: Apartment complexes • Elements: An individual who is a Virginia Tech student living within the sampling frame. • Sampling Frame: The apartment complexes that were chosen.

  7. Research Questions • Does transportation have a statistically significant impact on a residents’ decision to live there? • On average, how likely are our residents of interest to renew their leases, given that they can for the following year?

  8. Sampling Design • To achieve a reasonable representation of the population, a stratified random sample will be used. • The five (5) largest apartment complexes will be chosen. • Within them, the management will be contacted, and a maximum number of units will be attained. • From this, a minimum sample size will be calculated. These will be proportionate to one another to avoid over and underrepresentation.

  9. Problems with Other Random Sampling Designs

  10. Handling Bias

  11. Qualification Questions • The qualification survey is being used to: • Prevent time from being wasted from both parties involved. • Avoid response bias that would result from surveying subjects outside the sampling frame.

  12. Qualification Questions cont. • These questions consist of: • Please confirm that you are a resident of this apartment. • Please confirm that you are currently a Virginia Tech undergraduate student. • Please confirm that you are NOT a freshman. • The first subject to answer the door is considered the representative of the sampled unit provided they respond in the affirmative.

  13. Flow Chart of the Qualifiers Resident of the apartment? (Y/N) The interview is concluded Should these be negative … Undergraduate student at VT? (Y/N) Should these be all in the affirmative, the survey continues. Not a freshman? (Y/N)

  14. The Survey Questions • With one being extremely unsatisfied, two being unsatisfied, three being satisfied, and four being extremely satisfied, what is your overall satisfaction level with your respective apartment complex. • Specifically, what are some major factors that play a role in your response? • With one being extremely unsatisfied, two being unsatisfied, three being satisfied, and four being extremely satisfied, how much did transportation affect your decision to live here?

  15. The Survey cont. • Do you have a vehicle that is available for daily use at school? (If no, proceed to the sixth question) • How do you feel about the parking availability (for residents & visitors) that is available in this complex? • What is your main method of transporting yourself to campus during the business day (8 AM – 5 PM) (Car, Bus, Bike, Walk)? • With one being extremely unlikely, two being unlikely, three being likely, and four being extremely likely, how likely are you to renew your lease for next year, given that you have the opportunity?

  16. Survey Flow Chart (Version 1) • dsdf Overall satisfaction level (1-4)? Vehicle present(Yes, No)? Main method of transportation to campus (Walk, Bike, Bus, Car)? Major factors that contributed to the response (Open)? Renew the lease given it’s an option (Yes, No)? If ‘no’ skip following question How much did transportation affect your decision (1-4)? Thoughts on parking availability (Open)? Conclusion of interview.

  17. Reorganization of the Survey • Before the survey is administered, the research team will assign every sampling unit either an even or odd number so that there is an equal amount of each in the sample. • Those sampling units that are even numbers will be asked the original form of the survey. • Those with odd numbers will instead be asked the questions in a different order so that questions #1 and #2 become #5 and #6 respectively.

  18. Survey Flow Chart (Version 2) • dsdf How much did transportation affect your decision (1-4)? Thoughts on parking availability (Open)? Main method of transportation to campus (Walk, Bike, Bus, Car)? Vehicle present(Yes, No)? Renew the lease given it’s an option (Yes, No)? Overall satisfaction level (1-4)? Conclusion of interview. Major factors that contributed to the response (Open)? If ‘no’ skip following question

  19. Assumptions of the Survey • We are assuming equal variance among the apartment complexes. • We are assuming that the units sampled are being leased, and not owned, by individuals living there. • We are assuming that students must commute to campus in order to attend classes.

  20. Population Sizes

  21. Calculating Minimum Sample Size For a Mean

  22. Calculation of Sample Size For a Proportion Used because we are maximizing the equation; π is unknown. Therefore, a minimum sample size of 97 is needed.

  23. Actual Sample Size • The minimum sample size calculated was 97, but the research team wishes to prevent undercoverage.

  24. Selection • During the administration of the survey, a random number generator was used to determine which specific apartments would be surveyed. • Demonstrated here: Odd numbers received Version 1 of the survey, while even numbers received Version 2.

  25. Selection Example • For example, the first number randomly selected was ‘378’. • This corresponded to Apt 6800F in Foxridge and thus received Version 2 (even number). • This process was repeated until the minimum sample was achieved.

  26. Answering the First Research Question • Does transportation have a statistically significant impact on a residents’ decision to live there? • To answer this, the research team has chosen to estimate using a population proportion. • Categorize the data with values ‘1’ and ‘2’ being ‘no impact’ & ‘3’ and ‘4’ being ‘impact’. • Only interested in whether it did or did not affect the decision to live there, regardless of how much.

  27. Estimating Population Proportion • We can use this unbiased estimator: • The standard deviation and bound on this estimate are:

  28. Interpretation of Results • With the following parameters: • The following confidence interval is found: • The research team is approximately 95% confident that between 48.98% and 69.18% of residents believe that transportation did not affect their decision to live in their respective complex.

  29. Answering the Second Research Question • On average, how likely are our residents of interest to renew their leases, given that they can for the following year? • To answer this, the research team has chosen to estimate using a population mean. • Each of the specified values has an actual interpretation. • Respondents seemed to want to gravitate towards the middle, and therefore a mean makes more sense to interpret.

  30. Estimating Population Mean • We can use this unbiased estimator: • With the following estimators for variance, standard deviation, and bound:

  31. Interpretation of Results • With the following parameters: • The following confidence interval is found: • The research team is 95% confident that the true value of residents likeliness to renew lies between 2.336 and 2.8364 on the scale of 1-4.

  32. Additional Interpretation • The research team is interested in whether or not the order of the questions made a statistically significant difference. • To calculate this, the research team will perform a two sample z-test to determine if a difference exists between the two population proportions. • Those answering with either a ‘3’ or a ‘4’ for satisfaction level are categorized into the positive satisfaction responses, and the proportions for Version 1 and Version 2 will be compared.

  33. where and

  34. Output from Minitab 16 Interpretations: The research team is 95% confident that the true difference between satisfaction levels of Version 1 and Version 2 lie in our interval of -.0037 to .3738. Because 0 is in this interval, we fail to reject the null hypothesis that there is a significant difference .

  35. Limitations • Due to the lack of pre-existing data, we had to use values that would maximize our sample size equations. • The research team did not have enough resources to be able to generalize for the entire population. • A subject may have answered the final question while considering other factors that are not taken into account by this survey.

  36. References • "Incoming Freshman Housing." Housing & Residence Life. Division of Student Affairs, 17 Oct. 2011. Web. 15 Oct. 2011. <http://www.housing.vt.edu/contracts/apply/freshmen.php>.

More Related