BUS 799 Module 3 assignment: capstone methodology and data collection plan, a full sample

Following Every Invoice: The Methodology for Measuring Where Seventy-One Days of Receivables Come From

Student Name

Master of Business Administration Program, Aspen University

BUS 799: Graduate Capstone

Instructor Name

Month Day, Year

Following Every Invoice: The Methodology for Measuring Where Seventy-One Days of Receivables Come From

The methodology determines whether a capstone's findings can be trusted. This section describes how the project at Keystone Electric, a composite electrical subcontractor with days sales outstanding of 71 against a target of 50, will measure the contribution of each stage of billing and collection, identify causes of delay, and gather the perspectives of the people involved. It covers the design, data sources, variables, sampling, coding, interviews, analysis, and steps to protect validity and confidentiality.

Design

The project uses a mixed-methods process study within the measure and analyze phases of the define, measure, analyze, improve, and control sequence (Linderman et al., 2003). The quantitative strand traces payment applications through each stage to show where days accumulate. The qualitative strand explains why they accumulate, through coding the documented reasons for late invoices and interviewing the people who prepare, review, and pay them. Neither strand alone would be sufficient: the numbers show where the time goes, and the interviews show what could change it.

Data Sources and Variables

The primary source is Keystone's accounting and project management systems for the 24 months ending last quarter: 1,146 payment applications across 64 projects. For each application, the project will extract the end date of the billing period, the date submitted, the date approved by the general contractor, the date paid, whether it was rejected and returned, the stated reason for any rejection, the amount billed, and the retainage withheld. Change order logs will provide the date work on each change began and the date the change was signed. Receivables aging reports will show retainage balances and disputed amounts at each month end.

Measuring the Stages

Days sales outstanding will be decomposed into six components: billing lag, from the end of the period to submission; review time, from submission to approval; payment time, from approval to cash; rejection delay, the added days for applications returned for correction; change order delay, the days that billed but unsigned changes remain unpaid; and retainage, the share of receivables that is retainage expressed in days of revenue. Each component will be calculated as a revenue-weighted average so that large invoices count more than small ones, and the components will be checked to confirm that they sum to the measured days sales outstanding within two days. Results will also be segmented by general contractor and by contract type, since a few customers may account for much of the delay.

Sampling and Coding Late Invoices

To understand causes, the project will draw a stratified random sample of 80 payment applications that were paid more than 60 days after the end of their billing period, stratified by general contractor so that each major customer is represented. For each, the author and the controller will independently code the primary cause of delay using a scheme with eight categories developed from the literature and a pilot of ten invoices: late submission, missing lien waiver or backup, arithmetic or format error, disputed quantities, unsigned change order, retainage held, general contractor awaiting owner payment, and other. Agreement will be measured with the kappa coefficient, which corrects the raw percentage of agreement for the agreement two coders would reach by chance (Cohen, 1960). If kappa falls below 0.70, the categories will be clarified and the sample recoded before analysis.

Interviews

Twelve semi-structured interviews of about 30 minutes will be conducted: six Keystone project managers, the controller, two billing staff, one owner, and accounts payable staff at two general contractors. Questions will ask how each person prepares or reviews payment applications, where delays occur, and what would make the process faster. Notes will be analyzed using the six phases of thematic analysis described by Braun and Clarke (2006): familiarization with the data, generating initial codes, searching for themes, reviewing them, defining and naming them, and producing the report. The interviews are not meant to prove anything; they are meant to explain what the numbers show and to surface fixes that the people doing the work already know.

Analysis

The decomposition will produce a table of days contributed by each stage, overall and by general contractor. Coded causes will be displayed as a Pareto chart to show which few causes account for most delays. Interview themes will be matched to each stage. The analysis will then classify each cause as controllable by Keystone, partly controllable, or outside its control, which will guide the recommendations.

Timeline and Responsibilities

Data extraction will take place in the first two weeks, carried out by the author with help from the company's accounting software administrator, who will write the report that pulls dates for all 1,146 applications. The pilot coding of ten invoices and refinement of the categories will follow in week three, and the independent coding of the 80-invoice sample in weeks four and five, with the kappa calculation at the end of week five. Interviews will be scheduled across weeks three to six to fit project managers' field schedules. Analysis and the draft current-state report will be completed in weeks seven and eight. The controller has agreed to spend about eight hours on coding and review, and the owners have approved the interview time for project managers.

Validity, Reliability, and Ethics

Several steps protect the findings. Using the full population of payment applications for the decomposition avoids sampling error in the main measurement. Independent coding with a kappa check improves reliability of cause attribution. Having the controller, not only the author, code invoices reduces the risk of bias toward the author's department. Customer names will be replaced with codes, interview participants will be told their comments will not be attributed by name, and data will be stored on Keystone's secure server. Limitations will be reported, including that interviews with only two general contractors cannot represent all customers.

Conclusion

The methodology traces 1,146 payment applications through six measurable stages, codes the causes of 80 late invoices with a reliability check, and interviews twelve people involved in billing and payment. Together these methods will show where Keystone's 71 days come from, why, and which causes the company can address, providing the basis for the current-state analysis in the next module.

References

Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. https://doi.org/10.1191/1478088706qp063oa

Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20(1), 37-46. https://doi.org/10.1177/001316446002000104

Linderman, K., Schroeder, R. G., Zaheer, S., & Choo, A. S. (2003). Six Sigma: A goal-theoretic perspective. Journal of Operations Management, 21(2), 193-203. https://doi.org/10.1016/S0272-6963(02)00087-6