| Course | GB794 Advanced Topics in Marketing |
|---|---|
| Unit | Unit 8 |
| Paper type | Field study proposal |
| Length | About 990 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | MBA |
| Updated | October 2026 |
Free sample paper for GB794 Unit 8
Nothing Has to Change Price: A Unit 8 Field Study Proposal for Testing a Decoy on a Grocery Site
[Student Name]
Purdue University Global
GB794: Advanced Topics in Marketing
Unit 8 Field Study Proposal
[Instructor Name]
[Date]
Thornbury Market, its roaster and every figure are composite and illustrative; no review board has acted, and the methods sources cited are real.
The Question and Hypothesis
The study asks whether adding a dominated 26-ounce bag of house coffee to Thornbury's online coffee page increases the share of coffee buyers who choose the 32-ounce bag. The hypothesis comes from the Unit 7 memo and is deliberately modest: any rise in the share of buyers taking the largest bag will fall short of what laboratory work on the attraction effect would suggest, and may be zero. The laboratory literature began with Huber et al. (1982); later tests with realistic stimuli found the effect weaker and conditional (Frederick et al., 2014). A field test on a live grocery site would add evidence from a setting neither literature has examined closely: real purchases, made by habitual buyers, on a page with legally required unit prices.
Design
Unit of randomization. Loyalty accounts, not browsing sessions, will be randomized. A shopper who visits several times during the study will always see the same version of the page. Randomizing sessions would let the same shopper see both versions, which would contaminate the comparison and could confuse customers.
Conditions. Control accounts see today's page, with two sizes, 12 and 32 ounces, at their current prices. The treatment condition adds the 26-ounce bag at 24.49. No existing price changes in either condition. Placement, images and descriptions are otherwise identical.
Assignment. Each logged-in account will be assigned by a hash of its account number, producing two groups of roughly equal size. The assignment rule will be recorded before launch.
Window. Six weeks, chosen because most regular coffee buyers purchase at least once in that time.
Outcomes. The main measure is the share of coffee-buying accounts whose coffee purchases during the window were of the 32-ounce bag. Secondary outcomes are total house-coffee ounces purchased per account and total house-coffee revenue per account, since a decoy that shifts some buyers to 32 ounces but diverts others to the 26-ounce bag could leave revenue unchanged.
Power
Last year, about 6,200 logged-in accounts bought house coffee in a typical six-week period, giving roughly 3,100 per condition. About 35 percent of those buyers chose the 32-ounce bag. With 3,100 accounts per condition, the study has an 80 percent chance of detecting a gap near 3.5 points in that share, testing at the usual two-sided 5 percent level. Differences smaller than that, which are plausible given the Unit 7 prediction, could go undetected. The proposal states this limit openly: a null result would rule out large effects but not small ones.
What Thornbury Provides and Receives
Thornbury provides the page variants, the assignment through its site platform, and, once the six weeks end, a de-identified extract of coffee purchases by account. Thornbury further provides access to the roaster, who must agree to pack the 26-ounce bag for six weeks. The roaster has indicated it can do so from existing stock in a new bag size, at a cost Thornbury will cover.
In return, Thornbury receives a results memo before any academic paper is submitted, and a recommendation on whether to keep, drop or modify the listing. Thornbury may comment on the memo but may not change the findings or prevent publication, a term the hosting letter states explicitly. Thornbury will not be named in any publication without its consent.
Data Handling
The extract will contain, for each account, a random study identifier, the assigned condition and the coffee items purchased with dates and prices. It will not contain names, addresses, payment details or any purchase outside the coffee category. The extract will be held on an encrypted drive and deleted three years after publication.
Ethics
The study involves changing which products some customers see, not deceiving them: every product shown is real and available at the listed price. No customer is charged more. Because the study uses customer data for research, the candidate will seek review from the university's research ethics board before launch, following the reporting standards that Gerber and Green (2012) set out for field experiments. No review has yet taken place, and the study will not begin without it. The board may determine whether customer notice is required beyond Thornbury's existing privacy policy, which permits use of purchase data for analytics.
Risks
The main risk is that the decoy works too well in reverse: shoppers buy the 26-ounce bag rather than the 32-ounce bag, lowering revenue. The secondary outcomes will show this, and the hosting letter allows Thornbury to end the study early if revenue per account in the treatment group falls by more than 5 percent after three weeks.
A decoy test that changes no price and shows every shopper real products at real prices asks the host to risk six weeks of a coffee shelf, not its customers' trust.
What Each Result Would Mean
If the share choosing 32 ounces rises by three and a half points or more, with total coffee revenue per account steady or higher, the listing can be kept and the field evidence reported as support for the attraction effect in a real grocery setting. If the share does not change detectably, the study rules out a large effect and suggests the laboratory result does not travel to habitual online grocery buying, which is useful evidence for the literature and a reason for Thornbury not to keep an extra item. If revenue falls, the listing should be dropped regardless of what happens to the 32-ounce share.
Timeline
Two weeks for review board submission and approval, one week to build the page variants and test assignment, six weeks in the field, two weeks for the analysis and results memo.
References
Frederick, S., Lee, L., & Baskin, E. (2014). The limits of attraction. Journal of Marketing Research, 51(4), 487-507. https://doi.org/10.1509/jmr.12.0061
Gerber, A. S., & Green, D. P. (2012). Field experiments: Design, analysis, and interpretation. W. W. Norton.
Huber, J., Payne, J. W., & Puto, C. (1982). Adding asymmetrically dominated alternatives: Violations of regularity and the similarity hypothesis. Journal of Consumer Research, 9(1), 90-98. https://doi.org/10.1086/208899
GB794 Unit 8 instructions, in plain terms
Unit 8 in GB794 asks you to write a field study proposal that a real organization could host. State the question and hypothesis, connecting them to the literature and to your earlier work. Describe the design: the unit of randomization and why, the conditions, the assignment method and the window. Name the primary and secondary outcomes. Report power from the organization's own volumes, and state the smallest effect the study can detect. Set out what the host provides and receives, including any rights to review results. Describe exactly which data will be shared and how they will be protected. Address ethics honestly, stating that review has not yet occurred if it has not. Name risks and an early stopping rule. Say how you would read each possible outcome, and finish with a timeline. Ten pages with a draft hosting letter is common.
How the GB 794 Unit 8 example is put together
The proposal opens with the question and the modest hypothesis from Unit 7, citing the 1982 finding and the 2014 realistic tests. The design randomizes loyalty accounts by hash so each shopper sees one page, and keeps every price unchanged. Outcomes are the share buying 32 ounces, ounces per account and revenue per account. Power is worked from 6,200 buyers in six weeks, with the highlighted limit of a 3.5 point detectable gap. Hosting terms give Thornbury a results memo and no veto over findings. The data extract and its deletion are specified, and ethics review is requested, not claimed. A revenue stopping rule, the meaning of each result and a timeline close the proposal.
Reading the GB794 Unit 8 grading rubric
The GB794 field study proposal earns points when a host could say yes to it. Instructors give credit for a hypothesis grounded in the literature and stated modestly enough to be tested. The unit of randomization should be justified, and randomizing accounts rather than sessions shows care about contamination. Power from the host's own volumes, with the detectable effect stated, is weighted heavily. Hosting terms earn marks when they protect both the organization and the integrity of findings. Data handling must be specific about what is shared, how it is stored and when it is deleted. Ethics should be described honestly, without claiming approvals that do not exist. A stopping rule and an account of what each result would mean show the proposal has been thought through.
Common GB794 Unit 8 mistakes, and how to avoid them
Many GB794 field proposals describe an experiment that no organization would host, because they ignore the host's costs and concerns. Others randomize at the wrong level, letting the same customer see both conditions. Power is often omitted or based on guesses. Data requests are vague, which alarms a host's legal team. Ethics sections sometimes claim approval that has not been sought. Few proposals say what a null result would mean. Ground the hypothesis, randomize the right unit, compute power from real volumes, spell out what the host gives and gets, specify the data, state the review status honestly, add a stopping rule and interpret every outcome. Tell us your host and question, and a writer can draft the proposal for you.
Get a GB794 Unit 8 example written to your instructions
Tell us which organization might host your study, what it would need to agree to and the idea you want tested, then attach the Unit 8 prompt and rubric. Drafted from that, the proposal puts the host's concerns first, sizes the study honestly and leaves any review approval as a stated open item. Free for a first sample, back within 24-48h. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.
More GB794 papers and MBA samples
- GB794 Unit 1: Discussion Board Post
- GB794 Unit 2: Consumer Study Critique
- GB794 Unit 3: Experiment Design Memo
- GB794 Unit 4: Attribution Analysis
- GB794 Unit 5: Measurement Validity Review
- GB794 Unit 6: Seminar Reflection
- GB794 Unit 7: External Validity Memo
- GB794 Unit 9: Evidence Summary
- GB794 Unit 10: Marketing Research Paper
- GB793 Unit 8: Replication Assessment
- GB792 Unit 8: Valuation Model Critique
- GB791 Unit 8: Comment Letter Analysis
- GB740 Unit 8: Stakeholder Cost Analysis
GB794 Unit 8 questions, answered
Does a field study proposal need a signed letter from the host?
Sections differ, and your instructions decide. Many accept a draft hosting letter describing terms, since approval normally follows the proposal rather than preceding it. What reads as feasible is specificity: who at the organization would sign, what they provide, what they receive and how either party can withdraw. Vague statements of interest usually draw feasibility comments.
Is showing customers different pages ethical without telling them?
It can be, depending on risk and review. Here both groups pay identical prices and nobody is contacted, which keeps the risk minimal, but the decision belongs to the university's review process, not the author. The proposal describes the argument for a waiver of consent and leaves the determination open, which is the honest position before review.
What if the organization wants to see results before publication?
That is a common and reasonable condition. The proposal can grant an advance review period and a veto over naming the company, while reserving the right to report findings accurately. Stating the arrangement in writing at the start prevents disputes later, and instructors typically read such terms as evidence that the host relationship is real.
Where can I find a free GB794 Unit 8 sample paper?
A complete GB794 Unit 8 field study proposal is on this page, testing a decoy coffee bag on a grocery website with account-level randomization and written hosting terms.
Why randomize loyalty accounts instead of website sessions?
A shopper who visits several times would see both page versions if sessions were randomized, which contaminates the comparison. Randomizing accounts keeps each shopper in one condition throughout.