Team Sampling Data Collection Plan

Sampling & Data Collection Plan
Matthew Bell, David Cintron, Christopher Grunenberg, Shilo Morin
QNT/561
May 18, 2015
Russell Heigl
Sampling & Data Collection Plan
The Grub n’ Go sponsored research study is aimed at answering, “How does outside temperature (IV) impact the frequency of cold, non-alcoholic beverage sales (DV)? To answer this question top notch research team, Learning Team A, will have to go beyond this small restaurant brand.
Population
While the research question came from Grub n’ Go’s management looking toward the future, the best way to answer it with any certainty is to look into the past. The population for this study will be the flagship Grub n’ Go restaurant and their historical beverage sales records over the past six years, 2,191 unique data points! By looking at where the brand had been, the research team is able to identify beverage sales trends rather than focusing on a single, isolated period of time.
Sampling
Given of the objects to be observed, the team has decided that a data mining approach following a stratified sampling methodology is most appropriate. This approach allows for the division of the large number of data points into manageable groups while still being able to include the full six year history. The team will take six years’ worth of beverage sales data, 2,191 data points, from the flagship restaurant, satisfying the minimum sample size requirement of 385 using a 95% confidence level with a 5% margin of error (see Appendix 1.0, Sample Size Calculation). This data will then be matched with the publically accessible temperature highs by applicable region available from United States National Weather Service.
This data will be stored on the team’s cloud with an electronic copy given to each member of the Grub n’ Go leadership team. Additionally, the data and final report will be archived into the research team’s business database for no less than 10 years.
Validity
The reliability and overall validity of this research study will be strengthened through the interpretation of a large number of data points collected over the past six years. The large amount of data points greatly lessens the likelihood that “one offs” or outliers will throw off the overall results. The inclusion of data collected over the past six years serves as a check for consistency of any identified trends and helps to remove outliers generated from unseasonably warm or cool periods.
The findings of this large scale research study can then be applied to the original Grub n’ Go business problem. With the findings of this study expected to deliver the foreseeable ebbs and flows of beverage sales with 95% confidence, Grub n’ Go management should be able to predict and adjust inventory plans. Plans are expected to account for changes associated with regional, seasonal weather patterns and customer demands to appropriately predict beverage sales for years to come.

Appendix
1.0 Sample Size Calculation
Assuming a 95% confidence level, 5% margin of error, and a STDEV value of .5:
(Z-score)² * StdDev*(1-StdDev) / (margin of error)² [Source: Qualtrics.com, 2013] (1.96)2 x .5(1-.5)/(.05)2
= 384.16
= 385 minimum restaurants in sample

Reference
Qualtics (2013). Determining sample size: how to ensure you get the correct sample size. Retrieved from https://www.qualtrics.com/blog/determining-sample-size/.