| Course | BU 351 Introduction to Business Analytics |
|---|---|
| Unit | Unit 4 |
| Paper type | Exploratory analysis |
| Length | About 990 words, 6 pages |
| Format | APA 7 student paper |
| School | Herzing University |
| Program | BS in Business Analytics |
| Updated | September 2026 |
Free sample paper for BU 351 Unit 4
Five Passes Through a Membership File: An Exploratory Analysis of Early Cancellations at a Climbing Gym Chain, Including the Pass That Found Nothing
Student Name
Business, Herzing University
BU 351: Introduction to Business Analytics
Instructor Name
March 4, 2026
Five Passes Through a Membership File: An Exploratory Analysis of Early Cancellations at a Climbing Gym Chain, Including the Pass That Found Nothing
What Was Being Looked For
This analysis explores the cleaned membership file for Basecamp Climbing, the composite seven-gym chain in this course, looking for structure behind the pattern described in Unit 3: steady growth in members, a sharp rise in cancellations and a spike in April and May 2025. The search is not neutral. It was guided by the diagnostic question framed in Unit 1, which named five characteristics to examine, and by the Unit 3 findings, which pointed at the spring spike and at the Aurora gym. Tukey (1977) described exploratory analysis as detective work that looks for what the data suggest before any confirmation is attempted, and this paper records each pass, including one that turned up nothing, and ends with questions rather than conclusions.
Pass 1: When Members Leave
The first pass looked at a single field: how long members had belonged when they canceled in 2025. Of 1,193 cancellations that year, 36.5 percent came within 90 days of joining, 30.3 percent between 91 days and a year, and 33.3 percent after more than a year. A member's first three months account for a larger share of cancellations than any later quarter of membership, which is why the rest of the exploration concentrates on early cancellation. That analysis draws on everyone who signed up between the start of June 2024 and the end of September 2025, 1,901 people, all of whom had at least 90 days in which to cancel before the end of the file. Of those, 29.4 percent canceled within 90 days.
Pass 2: Pairs of Fields
The second pass crossed early cancellation with each characteristic in the framing. The results are shown in the table. The largest differences involve autopay failures and the introductory class. Members with no payment failures canceled early at 23.5 percent, those with one failure at 43.5 percent and those with two or more at 68.8 percent. Members who took the introductory class canceled early at 24.8 percent, against 34.9 percent for those who did not. Student plans had the highest early rate and family plans the lowest, and referred members canceled somewhat less often than others.
| Characteristic | Group | Members | Canceled within 90 days |
|---|---|---|---|
| Introductory class | Took class | 1,042 | 24.8% |
| Did not | 859 | 34.9% | |
| Autopay failures | None | 1,465 | 23.5% |
| One | 340 | 43.5% | |
| Two or more | 96 | 68.8% | |
| Plan | Family | 289 | 24.2% |
| Monthly | 1,181 | 28.4% | |
| Student | 431 | 35.5% | |
| Referral | Referred | 311 | 25.7% |
| Not referred | 1,590 | 30.1% |
Pass 3: Does the Pattern Hold Everywhere?
A relationship that appears in the whole file may come from one part of it. The third pass split the introductory class comparison by gym. In six of the seven gyms, members who took the class canceled early less often, with gaps ranging from about 8 to 20 percentage points; at Loveland, 21.9 percent against 41.7 percent. At Lakewood the gap nearly disappeared: 33.6 percent for class takers and 34.3 percent for others. The class appears to matter almost everywhere except Lakewood, which is itself something to explain.
The pass also crossed the introductory class with autopay failures. Among members with no payment failures, those who took the class canceled early at 17.5 percent against 30.4 percent for those who did not. Among members with at least one failure, the gap shrank to 47.1 percent against 52.0 percent. Whatever the class does, it seems to do little for members whose payments are failing.
Pass 4: The Aurora Gap
Unit 3 found Aurora's cancellation rate nearly a point above the chain's. The fourth pass asked whether Aurora's members simply differ in makeup. Aurora's recent joiners include more students, 26.7 percent against 22.7 percent chainwide, and students cancel early more often. But among non-student members, Aurora's early cancellation rate was still 30.8 percent against 27.1 percent at the other gyms. Plan mix explains part of the gap but not all of it. Aurora's share of members taking the introductory class, 51.9 percent, and its share with payment failures, 21.8 percent, are close to the chain's, so neither explains the rest.
One more possibility was noted but not tested: Aurora opened most recently of the seven gyms, and newer gyms may attract more first-time climbers who try the sport and decide it is not for them. The file does not record whether a member had climbed before, so this idea is listed among the candidate questions rather than claimed.
Pass 5: The Spring Spike, and a Pass That Found Nothing
The fifth pass asked whether the April and May 2025 spike came from a particular kind of member, such as long-standing members reacting to something, or new members leaving quickly. It compared the tenure of members who canceled in those two months with members who canceled in the rest of 2025. The two groups looked almost the same: 34.6 percent of spring cancellations had belonged for more than a year, against 32.9 percent in other months, and 37.0 percent had belonged for 90 days or less, against 36.3 percent. The spike seems to have affected members of every tenure roughly equally. The pass also examined age bands, and early cancellation varied only between 26.6 and 30.9 percent across them, with no clear pattern. Both results are recorded because a search that reports only its hits overstates what the data show (Provost & Fawcett, 2013).
Candidate Questions
The exploration earns questions, not conclusions. First, does the introductory class reduce early cancellation, or do members who choose it differ in ways that also keep them? Second, why does the class seem to make little difference at Lakewood and for members with payment failures? Third, can payment-failure outreach reduce cancellations among members whose failures may be accidental, such as expired cards? Fourth, what happened in April and May 2025 that affected members of every tenure, and does it coincide with the membership price increase the chain introduced in March? Fifth, what beyond plan mix sets Aurora apart? The predictive unit will use the relationships found here as inputs, and the prescriptive unit will return to the first and third questions. Every member and gym here is fictional (Few, 2012).
References
Few, S. (2012). Show me the numbers: Designing tables and graphs to enlighten (2nd ed.). Analytics Press.
Provost, F., & Fawcett, T. (2013). Data science for business: What you need to know about data mining and data-analytic thinking. O'Reilly Media.
Tukey, J. W. (1977). Exploratory data analysis. Addison-Wesley.
The Herzing BU 351 Unit 4 instructions
The Unit 4 assignment asks students to explore a dataset for patterns and relationships without reaching firm conclusions. Most versions want several passes: single fields examined for shape, pairs of fields compared, and segments checked to see whether a relationship holds everywhere or only in one group. Many rubrics ask you to record passes that found nothing and to end with questions for further analysis rather than claims. Instructors look for plain descriptions of what was seen and honesty about the limits of exploration. Correlation and group comparisons are appropriate. Expect a few pages with one or two tables, citing sources on exploratory analysis in APA 7.
Inside the BU 351 Unit 4 example
The sample states what it was looking for and admits the search was guided by earlier findings. Pass 1 examines how long members had belonged when they canceled and defines the cohort used for the rest of the analysis. Pass 2 presents a table of early cancellation rates across four characteristics. Pass 3 checks the introductory class pattern gym by gym and within payment-failure groups. Pass 4 tests whether Aurora's higher rate comes from its plan mix. Pass 5 examines the spring spike and age bands and reports that neither showed a pattern. The paper closes with five candidate questions for the predictive and prescriptive units.
Where the points sit in the BU 351 Unit 4 rubric
The rubric for exploratory analysis usually assesses the range of passes, the clarity of what was observed, the checking of patterns across segments and the quality of the resulting questions. Range earns full points when single fields, pairs and segments are all examined. Clarity is judged by plain, numeric descriptions of each pass. Segment checks score highest when they test whether a pattern holds everywhere, as the sample's gym-by-gym comparison does. Instructors reward honest reporting of passes that found nothing. The questions criterion looks for specific, testable questions rather than conclusions. Clear organization and appropriate citations complete the grade. A short statement of how the search was guided also earns credit, since exploration is never fully neutral.
Common BU 351 Unit 4 mistakes, and how to avoid them
The most frequent Unit 4 mistake is presenting exploration as proof, for example claiming the class reduces cancellations when the data show only an association. Another is reporting only the passes that found something, which overstates the evidence. Students often skip segment checks and miss that a pattern disappears in one group. Some papers repeat the descriptive analysis instead of looking for new structure. Others end with recommendations the exploration has not earned. Finally, record exactly what you compared and how; a reader should be able to repeat each pass and see the same result. Keep each pass to one comparison so its result is easy to read. End by ranking your questions by how much they would matter to the decision.
Get a BU 351 Unit 4 example written to your instructions
Bring the Unit 4 instructions, the BU 351 rubric and the dataset your section works from, together with the descriptive results the assignment builds on. We write a custom example that logs every pass including the empty ones, moves from single fields to crossings, checks patterns across segments and closes on testable questions. First custom sample free, returned in 24 to 48 hours.
More BU 351 papers and business analytics samples
- BU 351 Unit 1: Analytics types and framing
- BU 351 Unit 2: Data quality assessment
- BU 351 Unit 3: Descriptive analysis
- BU 351 Unit 5: Predictive exercise
- BU 351 Unit 6: Prescriptive or optimization exercise
- BU 351 Unit 7: Reporting to a business audience
- BU 351 Unit 8: End-to-end analytics case
- BU 455 Unit 4: Batch and streaming design
- BU 407 Unit 4: Optimization model summary
- BU 451 Unit 4: Control flow and functions
- BU 453 Unit 4: Schema implementation
BU 351 Unit 4 questions, answered
How is this different from the descriptive unit?
By what it is trying to do. The descriptive stage answers questions already framed and reports the answers; this stage goes looking for questions nobody has asked yet. A useful test is whether you knew what you were going to show before you opened the file. If you did, you are still summarizing, and the exploration has not started.
Can I use correlation in this unit?
As a description of how two fields move together, yes, and as evidence that one causes the other, no. Report what the measure says, note how many records it was computed over, and add a line on what else could produce the same pattern. Sections vary on whether they want the coefficient at all, so check the instructions before including one.
What if the exploration turns up nothing interesting?
Then write that up, because it is a real result and a rare chance to show judgment. Say how many comparisons were made, what would have counted as a pattern, and why the absence itself might matter for the decision behind the case. A document reporting a genuine null honestly scores better than one dressing up a difference too small to matter.
Where can I find a free BU 351 Unit 4 sample paper?
A free copy of the BU 351 Unit 4 exploratory analysis, all five passes included, is shown above. To see the same approach applied to your dataset, share the prompt and a description of your fields, and a free custom exploration will be written, typically within 48 hours.
What if my BU 351 Unit 4 exploration finds nothing?
Report it. The sample records a pass on the spring spike and another on age bands that showed no pattern. Null results are part of honest exploration, and instructors usually credit them as evidence of careful work.