DOC 715 Week 4 Defending the Methodology Example

Why Numbers First? Defending a Quantitative Methodology for an Onboarding Study

[Student Name]

University of Phoenix

DOC/715: Doctoral Seminar I

Week 4 Assignment

[Instructor Name]

[Date]

The learner, the company, its customers and all figures are composites written for a model paper.

Weeks 1 to 3 of the course strengthened the problem, refined the purpose and questions and repaired alignment for this onboarding dissertation. The committee will ask why she chose a quantitative methodology and what she gave up by choosing it. This paper prepares a full answer she can give in her own words.

Philosophical Stance

The learner takes a pragmatic stance. Morgan (2007) argued that the social sciences could move beyond the metaphysical paradigm debates by adopting a pragmatic approach that focuses on methodology as the link between abstract epistemology and practical methods, emphasizing abductive reasoning, intersubjectivity and transferability, and judging research choices by their consequences for answering questions. For this study, pragmatism means choosing the methodology that best answers questions about whether a practice is associated with outcomes across a large customer base, while recognizing that other questions about the same practice would call for other methods.

The Research Questions Decide

The three questions ask whether proactive onboarding is associated with twelve-month retention, whether it is associated with feature adoption and whether the retention association differs by business size. All three ask about the size and direction of relationships across a population of about 12,000 small service businesses. They are questions of how much and for whom, not how or why.

Comparing Methodologies

Quantitative: estimates associations across all customers in the study window using existing records, accounts for differences between groups statistically and tests whether effects differ by size. It cannot explain why onboarding helps or how customers experience it.

Qualitative: interviews with 20 to 30 small-business owners could reveal how they experience onboarding and why some stay. That would be valuable, but it could not estimate whether onboarding is associated with retention across the customer base or by how much, and it could not test whether the association differs by size.

Mixed methods: a quantitative strand plus interviews could answer both kinds of questions. Bryman (2006) analyzed hundreds of published mixed methods articles and found that many did not integrate their quantitative and qualitative components in the way their rationales promised, often treating them as parallel studies. Integration requires time and skill, and adding a qualitative strand would extend a working professional's timeline by six to nine months.

The Choice

The quantitative methodology is chosen because the questions require estimates across thousands of customers, because the data already exist in company records and because a well-designed quantitative study can answer the questions the company and the literature most need answered first: whether proactive onboarding matters, and for whom. If it does, explaining why becomes the natural next study.

A study of how owners experience onboarding would be richer, but it could not tell the company whether to hire three more customer success managers.

Objection 1: "Numbers Will Miss the Human Side"

The objection is fair. The study will not capture customers' experiences. Response: the problem as stated concerns the effect of a practice on retention, which numbers can estimate. The human side matters for designing better onboarding, which is why a qualitative follow-up is planned.

Objection 2: "You Can't Show Causation Without an Experiment"

Response: correct, and the purpose avoids causal claims. The quasi-experimental design defended in Week 5 will reduce bias from differences between groups, and the study will report associations with clear caveats. A randomized experiment was considered and rejected because the company would not withhold onboarding from customers who request it.

Objection 3: "Why Not Mixed Methods?"

Response: the questions as written do not require qualitative data to answer them, and adding a strand that cannot be meaningfully integrated within the timeline would weaken rather than strengthen the study, the pattern Bryman's analysis documented. Johnson and Onwuegbuzie (2004) argued that mixed methods are most useful when research questions call for the strengths of both approaches; here they do not.

Objection 4: "Company Records Are Messy"

Response: true of most organizational data. The learner has examined the fields needed, found them complete for 96 percent of customers from 2023 onward and will report missing data and how it was handled. Records also avoid the response bias that a customer survey would introduce, since customers who cancel rarely answer surveys.

Objection 5: "Your Role Biases the Study"

A committee member may note that the learner leads the team whose work is being evaluated and may hope for a positive result. Response: the risk is real. The analysis plan, including the matching method, covariates and tests, will be written and shared with the committee before the learner sees outcome data; an analyst outside her team will extract the records; and results will be reported whatever their direction. Choosing existing records over a survey she designed also limits her influence on the data.

Fit With the Learner's Skills and Resources

Methodology must also fit the researcher. The learner has used the company's analytics tools for years and has completed doctoral statistics courses, including regression and logistic regression. She has less training in qualitative analysis. Skills alone do not justify a choice, but they matter for feasibility: a methodology she can execute well within the timeline serves the study better than one she would be learning while doing it. Her committee includes a member experienced in causal inference from observational data, who can guide the design.

Contribution of the Chosen Methodology

A rigorous quantitative estimate fills the gap identified in Week 1 more directly than any other approach. Research on customer success has been largely descriptive or conceptual; an estimate of whether proactive onboarding relates to retention among small-business customers, with attention to bias, adds evidence where little exists.

What the Methodology Cannot Do

It cannot explain why onboarding helps, describe what happens in sessions or show how customers feel. It cannot rule out every unmeasured difference between groups. And it describes one company's customers, so findings must be judged for transfer to others.

A Planned Follow-Up

If the study finds an association, a qualitative follow-up will interview owners in onboarded and non-onboarded groups who stayed and who left, to understand mechanisms. That sequence, quantitative first, then qualitative to explain, matches the order in which the company needs answers.

How the Defense Will Be Presented

At the proposal defense, the learner will present the methodology in three slides: the questions and what kind of evidence each requires, the comparison of methodologies against those questions and the limits with the planned follow-up. She will keep the objections and responses in reserve for discussion rather than presenting them all, since committees prefer to raise their own concerns.

Conclusion

From a pragmatic stance, the research questions, which ask how much and for whom across thousands of customers, call for a quantitative methodology. Qualitative and mixed approaches were considered and set aside for this study, objections are answered and limitations are stated, with a qualitative follow-up planned. Week 5 will defend the specific design.

References

Bryman, A. (2006). Integrating quantitative and qualitative research: How is it done? Qualitative Research, 6(1), 97-113. https://doi.org/10.1177/1468794106058877

Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher, 33(7), 14-26. https://doi.org/10.3102/0013189X033007014

Morgan, D. L. (2007). Paradigms lost and pragmatism regained: Methodological implications of combining qualitative and quantitative methods. Journal of Mixed Methods Research, 1(1), 48-76. https://doi.org/10.1177/2345678906292462