| Course | GB532 Marketing Research |
|---|---|
| Unit | Unit 8 |
| Paper type | Data analysis report |
| Length | About 1,100 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | MBA |
| Updated | October 2026 |
Free sample paper for GB532 Unit 8
Significant and Small: A Unit 8 Data Analysis Report on Propane Customers With Heat Pumps
[Student Name]
Purdue University Global
GB532: Marketing Research
Unit 8 Data Analysis Report
[Instructor Name]
[Date]
The propane distributor, the survey results and every figure are composite and illustrative; the methods and sources cited are real.
Response
The survey went to 2,177 automatic-delivery households in two strata. In stratum A, the accounts flagged for falling propane use, 301 of 960 invited households responded, about 31 percent. In stratum B, the other automatic accounts, 352 of 1,217 responded, about 29 percent. Respondents in both strata matched the frame closely on last winter's change in propane use per degree day, the measure that defined the strata. Customers of more than ten years were slightly over-represented, so weights were adjusted for tenure as well as stratum.
How Results Are Judged
Each comparison is reported with a test of statistical significance and a measure of size. Wasserstein and Lazar (2016) caution against reading a p-value as a measure of an effect's size or its value to a decision, and note that crossing 0.05 does not by itself make a finding worth acting on. Sullivan and Feinn (2012) recommend reporting effect sizes alongside p-values for that reason. For differences in means, the report uses Cohen's d, read with the usual rough guide that values near 0.2 are small and values near 0.8 are large.
Table 1: Heat Pump Presence
Stratum A: about 71 percent of respondents have a heat pump in use.
Stratum B: about 16 percent.
Weighted across all automatic accounts: about 24 percent, with a margin of error of about 3.4 percentage points.
About 1,930 of the 8,160 automatic-delivery homes are therefore estimated to be hybrid homes. The share is well above the 15 percent in the first hypothesis, which is supported.
What the Flag Missed
The results also test the secondary review's flag. Stratum A, the flagged accounts, is about 71 percent hybrid, so the flag is a good indicator when it fires. But stratum B is so much larger that its 16 percent hybrid share amounts to more homes: about 1,120 hybrid homes in stratum B against about 810 in stratum A. More than half of all hybrid homes were not flagged. Most of them, according to the age question, installed heat pumps more than two years ago, before the two winters the flag compared. The flag finds recent adopters, not all adopters, and any plan or delivery change should not rely on it alone to identify hybrid customers.
Table 2: How Hybrid Homes Are Heated
Among the roughly 270 hybrid respondents, about 58 percent said their home is heated mostly by the heat pump in a typical winter month, about 27 percent about equally by both, and about 15 percent mostly by propane.
Table 3: The Coldest Days
Asked which source carries the load in the deepest cold, about 64 percent of hybrid respondents said propane, with a margin of about 6 points. The third hypothesis, that hybrid homes lean on propane when the weather is harshest, is supported. This matters for delivery: hybrid homes use little propane most of the winter and then a great deal in cold snaps, exactly when delivery capacity is tightest.
Table 4: Interest in a Backup-Heat Plan
Mean interest, five-point scale: hybrid households 3.42; other households 3.21.
Difference: 0.21 points; t about 2.5, p about 0.01.
Effect size: d about 0.20, small.
Share very or extremely interested: hybrid households about 49 percent; other households about 40 percent.
The plan appeals to hybrid homes a little more than to others, but it appeals to many customers of both kinds, which is the more useful finding for the company.
The second hypothesis, that hybrid households rate the plan more favorably, is technically supported but the difference is small. The more important result is that about four in ten other customers are also very interested, which suggests the plan should not be marketed only to hybrid homes.
Table 5: The Feature That Matters Most
Among hybrid respondents, guaranteed delivery during cold snaps was chosen as the most important feature by about 46 percent, a lower fixed annual fee by about 31 percent, pricing tied to actual use by about 16 percent and a tank monitor by about 7 percent.
Table 6: Heat Pump Services
Asked whether they would have Otter Creek install or service heat pumps, about 9 percent of all respondents said yes, 24 percent maybe and 67 percent no. Among hybrid households, the figures were about 11, 22 and 67 percent. Interest is limited, and it is no higher among the customers who already own the equipment.
Delivery Satisfaction
Hybrid households rated delivery timing lower than other households, about 3.6 against 4.0 on a five-point scale, a difference of 0.4 with p well below 0.01 and d about 0.42, between small and medium. Several hybrid respondents wrote that deliveries came too often in mild weather and too late in cold snaps. The company's scheduling, based on degree days and past use, appears not to fit homes whose propane use now depends on the weather in a different way.
What Customers Wrote
About 40 percent of respondents added comments. Among hybrid households, the most common themes were delivery timing, mentioned by about one in four commenters; uncertainty about whether to keep the propane tank at all, about one in six; and questions about minimum delivery charges, about one in eight. Several wrote that they kept propane mainly for reassurance during long cold spells and power outages. Those comments support the plan's emphasis on guaranteed delivery and suggest that reliability, not price, is what keeps hybrid customers with propane.
Charts
The first chart shows heat pump presence by stratum and the weighted total, with error bars. The second shows interest in the plan for hybrid and other households as distributions rather than means, which makes clear how much the two groups overlap.
Summary of Findings
About a quarter of automatic customers are hybrid, and most still rely on propane in the coldest weather. Interest in a backup-heat plan is substantial across customers and only slightly higher among hybrids. Hybrid customers are less satisfied with delivery timing, by a moderate margin. Interest in buying heat pump services from the company is low.
Limits
Interest in a hypothetical plan is not a reliable guide to behavior. Respondents may differ from nonrespondents in ways not captured by fuel use and tenure. Will-call customers were outside the frame, so these results apply only to automatic-delivery households (Lohr, 2021).
References
Lohr, S. L. (2021). Sampling: Design and analysis (3rd ed.). CRC Press.
Sullivan, G. M., & Feinn, R. (2012). Using effect size, or why the p value is not enough. Journal of Graduate Medical Education, 4(3), 279-282. https://doi.org/10.4300/JGME-D-12-00156.1
Wasserstein, R. L., & Lazar, N. A. (2016). The ASA statement on p-values: Context, process, and purpose. The American Statistician, 70(2), 129-133. https://doi.org/10.1080/00031305.2016.1154108
Reading the GB 532 Unit 8 instructions
GB532 Unit 8 asks you to analyze your survey data and report the results. Begin with response rates and how respondents compare with the frame, and describe any weighting. State how results will be judged, reporting both statistical significance and a measure of effect size. Present each table with its finding and its test. Use the data to check any assumption built into the design, such as a proxy used to define strata. Summarize open-ended comments by theme with rough frequencies, describe what each chart shows and summarize the findings in plain terms. Close with limits. Expect about seven pages with six tables and two charts, citing statistics and sampling sources in APA 7.
Inside the GB532 Unit 8 example
The response section reports 301 of 960 and 352 of 1,217 and adjusts weights for tenure. A section on judging results cites Wasserstein and Lazar on p-values and Sullivan and Feinn on effect sizes. Table 1 gives 71 and 16 percent by stratum and about 24 percent overall. A section shows that more than half of hybrid homes were outside the flagged stratum. Tables 2 and 3 describe heating patterns and coldest-day reliance. Table 4 reports plan interest with d of about 0.20, and the highlighted sentence reframes the small difference. Tables 5 and 6 cover features and services. Delivery timing shows d of about 0.42. Comments, charts, a summary and limits citing Lohr close it.
GB532 Unit 8 rubric: what earns full points
The GB532 analysis report is credited for pairing every test with a measure of how much the difference matters. Reporting response and weighting first earns early credit. Stating the two-part standard before the tables shows rigor. Each table with its finding and test carries significant weight. Using the data to check the design's own proxy shows critical thinking. Distinguishing a small significant difference from a moderate one, and saying which matters more, shows judgment. Coding comments into themes and describing charts by what they show add depth. Limits complete it. Wasserstein and Lazar, Sullivan and Feinn and Lohr are the report's APA 7 references.
Common GB532 Unit 8 mistakes, and how to avoid them
Many GB532 Unit 8 reports list p-values and declare every significant result important. Others report means without saying how many respondents stand behind them or whether the data were weighted. A frequent gap is the design check: survey results can show whether the proxy used to build strata worked, and here it missed most hybrid homes. Some writers leave comments uncoded. Report response and weights, set the two-part standard, give each table a finding and a test, check the design's assumptions, code comments, describe charts and state limits. Showing distributions rather than only means helps readers see how much two groups overlap, which is often the most honest way to present a small difference.
Get a GB532 Unit 8 example written to your instructions
Bring whatever your section handed over, raw data, software output or summary tables, together with the Unit 8 instructions and the GB532 rubric. The custom report runs each comparison your prompt calls for and pairs every p value with an effect size and a practical reading. It is free as a first sample and returned 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 GB532 papers and MBA samples
- GB532 Unit 1: Discussion Board Post
- GB532 Unit 2: Problem Definition Memo
- GB532 Unit 3: Secondary Data Review
- GB532 Unit 4: Research Proposal
- GB532 Unit 5: Sampling Plan
- GB532 Unit 6: Questionnaire Draft
- GB532 Unit 7: Seminar Reflection
- GB532 Unit 9: Findings Interpretation
- GB532 Unit 10: Research Report
- GB531 Unit 8: Seminar Reflection
- GB530 Unit 8: Channel Strategy
- GB528 Unit 8: Value Chain Audit
- GB527 Unit 8: Network Design Analysis
GB532 Unit 8 questions, answered
What is the difference between statistical and practical significance?
Statistical significance says a difference is unlikely to be chance alone; practical significance says it is large enough to change a decision. With several hundred respondents, tiny differences pass the first test easily. An effect size, such as Cohen's d or phi, plus a sentence on what the difference means for the manager, shows you have addressed both.
Which tests does a GB532 analysis report usually need?
Whatever fits the variables. Two categorical variables call for a cross-tabulation and chi-square; a numeric outcome across two groups calls for a t test; more groups, an analysis of variance; a numeric outcome predicted by several variables, regression. State why each test fits the data. Your prompt may name the tests expected, and those come first.
Do I need to weight survey data in a course report?
Only if the design sampled groups at unequal rates or the returns differ sharply from the population on something measurable. Where your sampling plan oversampled a stratum, weighting is part of doing the analysis correctly, and reporting unweighted overall figures overstates that group. Show both figures once, explain the difference, then use the weighted one.
Where can I find a free GB532 Unit 8 sample paper?
This page carries a full GB532 Unit 8 data analysis report on propane customers with heat pumps, with response and weighting, six tables judged by significance and size, comments and APA 7 references.
Why is a significant difference called small?
With several hundred respondents, even a 0.21-point gap on a five-point scale passes the significance test. Its effect size of about 0.20 shows the two groups' ratings overlap heavily, so the difference should carry little weight.