| Course | GB531 Advertising |
|---|---|
| Unit | Unit 9 |
| Paper type | Campaign measurement plan |
| Length | About 920 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | MBA |
| Updated | October 2026 |
Free sample paper for GB531 Unit 9
Two Survey Waves and an Unadvertised Market: A Unit 9 Campaign Measurement Plan Set Before Any Ad Runs
[Student Name]
Purdue University Global
GB531: Advertising
Unit 9 Campaign Measurement Plan
[Instructor Name]
[Date]
The battery brand, the samples and every figure are composite and illustrative; the methods and sources cited are real.
Table 1: Measures Mapped to Objectives
The objectives memo in Unit 2 set four objectives. Each has a measure.
Awareness: share of the target audience recognizing the brand name when shown it; survey.
Comprehension: share who connect Quietwall with running without fuel in a blackout; survey; the campaign's primary measure.
Conviction: share saying they are at least somewhat likely to request a quote within a year; survey.
Action: monthly quote requests, scaled to the number of audience households and drawn from the electricians' booking records by ZIP code.
Colley (1961) argued that advertising can only be judged against objectives defined and measured before the campaign begins, and the plan holds every question, sample and comparison fixed from April.
The Design
The plan uses a difference-in-differences comparison. In each of the three advertised markets, about 400 target homeowners are interviewed in April and another 400 in mid-July. In Baton Rouge, which receives no campaign advertising, 400 are interviewed at each wave.
The campaign's effect is estimated as the change in the advertised markets minus the change in Baton Rouge. If association with needing no fuel rises 15 points in the advertised markets and 2 points in Baton Rouge, the estimated effect is 13 points. Angrist and Pischke (2009) describe this design as a way to remove changes common to both groups, such as hurricane season news coverage that makes every homeowner think about backup power, provided the two groups would have moved in parallel without the campaign.
Without Baton Rouge, any rise in the advertised markets could be the season talking rather than the campaign.
The Smallest Detectable Change
With about 1,200 interviews across the advertised markets and 400 in Baton Rouge at each wave, the standard error of the difference-in-differences estimate for a measure near 20 to 30 percent is about 3.3 percentage points. To detect an effect with the usual confidence and power, the change must be at least about 2.8 times that, roughly 9 points.
Comprehension: the target is a rise from 17 to 35 percent, about 18 points, well above the detectable change.
Awareness: the target is a rise of 17 points, also detectable.
Conviction: the target is a rise from 22 to 28 percent, 6 points, below the detectable change of about 9.6 points. The survey could not distinguish a real 6-point rise from chance.
Fixing the Conviction Problem
Two fixes are possible. Raising the Baton Rouge sample to 800 at each wave would lower the detectable change to about 7.3 points, still above the 6-point target. Raising every sample would be costly. The plan therefore reports conviction as a directional measure, alongside quote requests, which are counted rather than sampled and can reveal smaller changes. The memo's conviction target should be relabeled as an indicator, not a pass-or-fail objective.
Quote Requests
Quote requests are counted from the electricians' booking system for every ZIP code in the four markets, by week, from March through August. The same difference-in-differences logic applies: the change in requests per 10,000 target households in advertised ZIP codes minus the change in Baton Rouge ZIP codes. Because requests are counts from the whole population rather than a sample, smaller changes can be detected, though weather and electricians' availability add noise.
Checking the Parallel Assumption
The design works only if Baton Rouge would have moved like the advertised markets without the campaign. Two checks help. First, the April wave should show similar starting levels on each measure in Baton Rouge and the advertised markets; a large gap would warn that the markets differ in ways that could affect trends. Second, quote request data from the previous two years can show whether Baton Rouge and the advertised markets followed similar seasonal patterns before this campaign. If they did, the assumption is reasonable; if not, the plan should adjust the comparison or add a second control market.
The Survey Questions
The questions are written once and not changed. Comprehension is measured with an open question before any prompt, "What, if anything, do you associate with the Quietwall brand?", followed by a list on which needing no fuel appears among six other attributes in rotating order. That order matters: asking about fuel first would teach respondents the answer.
Threats to the Design
Spillover. Baton Rouge is near New Orleans, and some New Orleans television signals and travelers will reach it. Connected TV and streaming audio can be restricted by location, but broadcast television cannot. The plan records which Baton Rouge ZIP codes receive New Orleans stations and checks whether results differ between them.
Storms. A hurricane striking one market during the campaign would change homeowners' attitudes there for reasons unrelated to advertising. If that happens, the plan reports results with and without the affected market.
Survey timing. Both waves must be fielded in the same weeks in all four markets, and the July wave must close before the August pulse begins.
Reporting
Results will be reported market by market and pooled, with the difference-in-differences estimate, its margin of error and whether it exceeds the detectable change. Belch and Belch (2021) emphasize that measurement should help managers learn what worked, so the report will also compare markets by spending per target adult.
References
Angrist, J. D., & Pischke, J.-S. (2009). Mostly harmless econometrics: An empiricist's companion. Princeton University Press.
Belch, G. E., & Belch, M. A. (2021). Advertising and promotion: An integrated marketing communications perspective (12th ed.). McGraw Hill.
Colley, R. H. (1961). Defining advertising goals for measured advertising results. Association of National Advertisers.
The Purdue Global GB531 Unit 9 instructions
In GB531 Unit 9 you design a plan to measure a campaign's effects. Map each measure to an objective, stating its source. Choose a design that separates the campaign's effect from other influences, such as a comparison market, and explain its key assumption. Work out the smallest change your samples can detect, and compare each objective's target with it. Where a target is too close to detect, say so and propose how to handle it. Explain how counted behavior, such as quote requests, will be analyzed, how the design's assumptions will be checked and how survey questions will be written. List threats to the design and how results will be reported. Expect about six pages, citing measurement sources in APA 7.
Inside the GB531 Unit 9 example
Table 1 maps awareness, comprehension, conviction and action to survey questions and booking data, citing Colley on fixing objectives before launch. The design compares advertised markets with Baton Rouge, citing Angrist and Pischke, and the highlighted sentence explains why the comparison market is needed. The detectable change is worked out at about 9 points, so the conviction target is reported as directional. Quote requests are analyzed the same way by ZIP code. A section checks the parallel-trends assumption with April levels and two years of request data, and another fixes question order. Threats from spillover, storms and survey timing follow, with reporting by market and pooled, citing Belch and Belch.
Where the points sit in the GB531 Unit 9 rubric
The GB531 measurement plan is assessed on a design fixed in advance and on honesty about what the samples can detect. Mapping each measure to an objective earns early credit. A comparison design with its key assumption stated shows rigor. Calculating the smallest detectable change carries significant weight, because it shows which targets the plan can judge. Admitting that one target falls below that limit, and handling it, shows integrity. Checks on the parallel-trends assumption and on question order show care. Threats with responses, and a reporting plan, complete it. Colley's DAGMAR, Angrist and Pischke on difference-in-differences and the Belch and Belch text form the plan's APA 7 sources.
GB531 Unit 9 help: mistakes that cost points
Many GB531 Unit 9 plans measure before and after the campaign in the advertised markets only, so any change could be caused by the season rather than the advertising. Others set survey samples without asking what change they can detect. A frequent gap is the target that cannot be judged: if a goal is smaller than the detectable change, the result will be noise either way. Some writers let question order teach respondents the answer. Map measures to objectives, choose a comparison design, compute the detectable change, flag targets below it, analyze counted behavior, check assumptions, fix the questions and plan reporting. Recording which comparison ZIP codes receive advertised signals handles the most likely spillover.
Get a GB531 Unit 9 example written to your instructions
Copy your campaign's objectives word for word, list the markets involved, and add the Unit 9 prompt and rubric. The custom measurement plan maps every objective to a test it could fail and states detection limits, returning free as a first sample in 24-48h. Survey costs without a quote are labeled estimates. 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 GB531 papers and MBA samples
- GB531 Unit 1: Discussion Board Post
- GB531 Unit 2: Advertising Objectives Memo
- GB531 Unit 3: Creative Brief
- GB531 Unit 4: Audience Profile
- GB531 Unit 5: Message Strategy Analysis
- GB531 Unit 6: Media Plan
- GB531 Unit 7: Budget Allocation Model
- GB531 Unit 8: Seminar Reflection
- GB531 Unit 10: Integrated Campaign Report
- GB530 Unit 9: Integrated Promotion Plan
- GB528 Unit 9: Implementation Plan
- GB527 Unit 9: Demand Forecast
- GB526 Unit 9: Change Leadership Plan
GB531 Unit 9 questions, answered
Do I need a control market in a course measurement plan?
Not always, but some form of comparison strengthens any plan: an unadvertised market, a holdout group of households, or the same weeks in a prior year. Without one, the plan can show change but not cause. If your section's scenario offers no comparison, say so and name what else could have produced the change.
How do I work out the smallest change a survey can detect?
For comparing two proportions, the standard error depends on the proportion and the sample sizes; multiplying it by about 2.8 gives the change detectable with 80 percent power at the usual 5 percent significance level. Show the formula once with your figures. If a target falls below the limit, enlarge the sample or revise the target.
Should clicks and impressions appear in the plan?
They can, as delivery measures showing the campaign ran as bought. They should not stand in for objectives unless an objective is itself about response online. Separate the two in the plan: what confirms the media ran, and what judges whether the advertising worked. Graders in many sections look for that distinction.
Where can I find a free GB531 Unit 9 sample paper?
This page carries a full GB531 Unit 9 campaign measurement plan with two survey waves, an unadvertised comparison market, a difference-in-differences design, a detectable-change calculation and APA 7 references.
Why is Baton Rouge used as a comparison market?
It is similar to the advertised markets but receives no campaign advertising. Subtracting its change removes effects common to both, such as hurricane-season news, leaving an estimate of what the campaign itself caused.