BUS 225 Module 3 Data Analysis Assignment Example

Callback Data Analysis, September 2024 to August 2025

[Student Name]

Southern New Hampshire University

BUS 225: Critical Business Skills for Success

Module Three Assignment

[Instructor Name]

[Date]

The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.

Callback Data Analysis, September 2024 to August 2025

Introduction

Module Two identified six possible causes for the rise in callbacks: a cheaper perimeter product introduced in March 2025, more stops per route, newer technicians with shorter training, a wet summer, a discontinued post-treatment text to customers and lower effort by technicians. This paper tests these explanations against the company's service records for September 2024 through August 2025. Each explanation predicts a different pattern, so the analysis is built around comparisons that would come out differently depending on which is true.

The Data and Its Preparation

The records cover 31,400 visits, each with a date, customer, technician, service type, treatment product, route stops that day and, for callbacks, the technician's note on what was found. Three problems had to be fixed before analysis. Callback visits were coded four different ways, which were combined into one code. About 900 rows were duplicates created when visits were rescheduled and were removed. And the technician's findings on callbacks were free text, which was sorted by reading each note into three categories: pests found and treated, no pests found, and other. Tukey (1977) urged analysts to look at data in simple ways before drawing conclusions, and the cleaning alone revealed that callbacks were not spread evenly through the year.

Before and After the Product Change

Callback rate by service type, before and after March 2025

Perimeter treatments, the only service that used the new product, saw callbacks more than double after March, while interior treatments and inspections barely changed. If technicians had become careless across the board, all service types should have risen. If the wet summer were the main cause, interior callbacks should have risen too, since moisture drives pests indoors. The pattern points strongly toward the product.

A month-by-month view makes the timing clearer. Perimeter callbacks hovered between 4.6 and 5.6 percent from September through February, jumped to 9.8 percent in March, the first month of the new product, and stayed between 12 and 15 percent through August. A gradual decline in technician effort would have produced a slow drift, not a step change in the month the product changed. Interior callbacks show no step at all.

Technician Experience and Route Density

Technicians hired within the past year had a callback rate of 12 percent from March to August, compared with 10 percent for experienced technicians. That difference is real but small compared with the jump in perimeter callbacks, about two points against eight, and experienced technicians' callbacks also rose sharply after March, which the effort explanation would not predict for a product problem. Callbacks rose modestly with route density: technicians averaging seventeen or more stops a day had a rate of 11.8 percent, against 10.1 percent for those under fifteen. Knaflic (2015) advises analysts to tell readers which differences matter, and these two are worth noting but are not the main story. They suggest that restoring the third week of training and easing route loads would help at the margin, but neither would reverse most of the rise.

Callback rate by technician group, March to August 2025

What Technicians Found on Return

The company stopped sending its post-treatment text, which warned customers that bugs driven out by the treatment could show up for a couple of weeks, in September 2024 when it changed messaging software. Among callbacks before that change, technicians found no pests on 15 percent of visits. From October 2024 onward, the share was 38 percent. That rise suggests many callbacks are customers reacting to normal activity they were no longer warned about, a cause that has nothing to do with the treatment itself. The office manager confirmed that customers calling in the first week after treatment often say they did not expect to see more bugs. If the share of no-pests-found callbacks returned to its earlier level, total callbacks would fall by roughly a quarter on that change alone.

Weather and Effort

Interior callbacks would have risen in a wet year, and they did not. Two neighboring companies told the owner their callbacks were steady. The effort explanation would predict a rise across services and technicians unrelated to the product change, and the data show instead a rise concentrated in the treatment that changed. Neither explanation can be ruled out entirely, but the evidence for both is weak.

Putting the Findings Together

Taken together, the comparisons rank the six explanations by the strength of evidence. The product change has the strongest support: a sharp step in the month it began, confined to the service that used it. The missing customer text comes second, with a large rise in callbacks where nothing was found. New technicians and busy routes each add a small, consistent effect. The weather and carelessness explanations have little support. This ranking matters because it tells the owner where a fix will pay off: the top two causes are both decisions the company made and can undo quickly, while the smaller effects call for slower changes in training and scheduling.

Limits

The product changed in March, the same month the new neighborhood was added and route density rose, so part of the March jump could reflect busier routes. However, route density explains only a small difference, and interior treatments on the same busy routes did not rise. The data show associations, not proof; a controlled test, such as using the old product on half of the routes for two months, would confirm the product's role. Paul and Elder (2019) stress that a well-reasoned conclusion states the limits of its evidence, and these limits shape the recommendation in Project One.

Conclusion

The service records point to two main causes: the cheaper perimeter product, which accounts for most of the rise, and the discontinued customer text, which explains why many callbacks find nothing wrong. New technicians and busier routes add smaller effects. The owner's explanation, that technicians became careless, is not supported. Project One will turn these findings into a recommendation.

References

Knaflic, C. N. (2015). Storytelling with data: A data visualization guide for business professionals. Wiley.

Paul, R., & Elder, L. (2019). The miniature guide to critical thinking concepts and tools (8th ed.). Rowman & Littlefield.

Tukey, J. W. (1977). Exploratory data analysis. Addison-Wesley.