DOC 715 Week 3 Alignment Among Components Example

Does Every Piece Fit? Evaluating Alignment in a Dissertation on Proactive Onboarding

[Student Name]

University of Phoenix

DOC/715: Doctoral Seminar I

Week 3 Assignment

[Instructor Name]

[Date]

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

Weeks 1 and 2 strengthened the problem and refined the purpose and questions for the onboarding study at the invented Texas software firm. Earlier courses produced a framework, a design and a draft analysis plan. This paper checks whether all of them still fit together.

Why Alignment Matters

Creswell and Creswell (2018) emphasize that the components of a research design should be connected: the problem leads to the purpose, the purpose to the questions, the questions to the methods and the methods to the analysis, with consistent terms throughout. Locke et al. (2014) note that proposal reviewers look first for this coherence, and that misaligned proposals are among the most common reasons for revision. A misalignment caught now saves months of revision later.

The Alignment Table

Each row of the table is one research question; columns are problem, purpose, question, hypothesis, framework, variables, population and sample, data source and analysis.

RQ1 row: early loss of small-business customers; examine association of proactive onboarding with retention; RQ1 and H1; customer success theory on helping customers realize value; onboarding, retention, covariates; small service businesses, 2023 to 2025; company records; matched comparison and logistic regression.

RQ2 row: low feature use among early leavers; association with adoption; RQ2 and H2; same framework plus a pathway from guidance to use; onboarding, adoption; same sample; product logs; matched comparison and count regression.

RQ3 row: need to target scarce managers; differences by size; RQ3 and H3; framework silent; onboarding, retention, size; same sample; records; interaction term.

Misalignment 1: Population

The purpose says "small service businesses." The sample definition in the earlier design draft read: "customers with fewer than 20 employees," which includes retail shops that use the software for appointments. Retail customers behave differently: they use fewer features and churn less. Revision: the sample definition now reads "customers with fewer than 20 employees whose primary industry at signup is a service category," listing the eleven categories.

Misalignment 2: Framework and H3

The framework, drawn from customer success research (Hochstein et al., 2020), explains why proactive onboarding might help customers realize value and stay. But it says nothing about why the effect should differ by size, so H3 had no theoretical basis. Two options: drop H3 or extend the framework. The learner extended it with the logic of organizational resources: businesses with one to four employees typically have no one whose job includes learning software, so external guidance substitutes for internal capacity. Thong (1999) found that small businesses' adoption of information systems depended on characteristics of the business, such as size and employees' knowledge of information systems, as well as on the owner's attitudes. That reasoning now appears in the framework section, and the alignment table's RQ3 framework cell is no longer empty.

An empty cell in the framework column was the clearest sign that H3 was a hunch, not a hypothesis.

Misalignment 3: Analysis and Data Structure

The earlier analysis plan for RQ1 treated each customer as an independent observation. But onboarding is delivered by 23 customer success managers, each of whom onboarded between 40 and 300 customers. Customers of the same manager may share outcomes because of that manager's skill or style, which violates the independence assumption and understates standard errors. Revision: the analysis now uses standard errors clustered by manager and includes a sensitivity analysis with manager as a random effect. The data source column now notes the manager identifier field.

Misalignment 4: Variable Definition

Week 2 counted a feature as adopted once a customer had used it three or more times in the first ninety days. An older measures section defined it as "features enabled," which counts settings switched on, not use. The definitions measure different things. Revision: every section now uses the Week 2 definition, and the measures section explains why use is preferred to enablement.

Reading Down the Columns

Beyond rows, the learner read each column top to bottom. The design column shows the same design, quasi-experimental with matching, for all three questions, which is consistent. The data source column shows that all variables come from three company systems, which is feasible. The analysis column revealed that RQ3's interaction test would rely on fewer customers in the one-to-four-employee group who received onboarding; Week 6's sampling plan will check whether that subgroup is large enough.

What Stayed Aligned

Not everything needed repair. The problem's focus on first-year loss carries through to the twelve-month retention outcome. The quasi-experimental design fits questions framed as associations after accounting for group differences. And the data sources, all existing company records, can supply every variable. Reporting what already fits shows the committee that the evaluation was balanced rather than a search for faults.

Peer Review

A classmate reviewed the table without reading the proposal and asked one question the learner had missed: whether "retention" meant the customer stayed on any plan or on the same plan. The definition now specifies any paid plan, since downgrades are not losses in the problem's terms.

Why Small Misalignments Matter

Each of these problems could look minor. But together they would have weakened the study in ways a committee would notice. Including retail shops would have mixed two populations with different churn patterns and muddied the estimate. An unsupported H3 would invite the question of why size should matter at all. Ignoring clustering by manager could have made a modest association look statistically significant when it was not. And two definitions of feature adoption would have left readers unsure what RQ2 measured. Alignment is not about tidy tables; it protects the validity of the conclusions.

Checking Terms Across the Document

After the repairs, the learner searched the full proposal draft for each key term, such as "proactive onboarding," "retention," "feature adoption" and "small service business," to confirm that each appears with the same meaning everywhere. Four stray uses of "engagement" as a synonym for adoption were replaced, since engagement suggests a broader construct that the study does not measure.

Effect on the Timeline

The repairs add work. The framework section needs a new subsection on small-business resources, and the analysis plan needs clustered and multilevel models, which the learner will practice on a sample of anonymized records before the proposal defense. She estimates three additional weeks, a cost worth paying to avoid a revision request after the defense.

Version Log

Each revision is recorded with the date, the sections changed, the old and new wording and the reason. The log will accompany the proposal to the committee, so members can see how the design changed and why each change was made.

Conclusion

An alignment table read across rows and down columns revealed four misalignments, in population wording, theoretical support for H3, the analysis's handling of clustered data and a variable definition, plus one ambiguity found by a peer. Each has been repaired and logged. Week 4 will defend the choice of methodology.

References

Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). Sage.

Hochstein, B., Rangarajan, D., Mehta, N., & Kocher, D. (2020). An industry/academic perspective on customer success management. Journal of Service Research, 23(1), 3-7. https://doi.org/10.1177/1094670519896422

Locke, L. F., Spirduso, W. W., & Silverman, S. J. (2014). Proposals that work: A guide for planning dissertations and grant proposals (6th ed.). Sage.

Thong, J. Y. L. (1999). An integrated model of information systems adoption in small businesses. Journal of Management Information Systems, 15(4), 187-214. https://doi.org/10.1080/07421222.1999.11518227