RES 6302 Module 3 Anticipated Committee Questions Example

Ten Questions I Expect From My Committee: Prepared Answers for the Defense of a Survey Study of Supervisors, Safety and Employee Voice

Student Name

American College of Education

RES6302: Defending the Dissertation

Module 3 Assignment

Instructor Name

October 26, 2026

Introduction

The question period is the heart of a defense. It is where committee members test whether the candidate understands the study's choices and limits as well as its results. Most questions can be predicted from the design, and a candidate who has thought them through can answer calmly instead of improvising. This paper lists the ten questions I consider most likely, drawn from my chair's comments, my own limitations section and two practice defenses with doctoral colleagues, and drafts an answer for each. The list is not meant to be memorized word for word; the aim is to know the shape of each answer well enough to say it naturally under pressure.

How Each Answer Is Built

Every answer follows the same four steps. First, acknowledge the substance of the question in a sentence, which shows I have heard it and gives me a moment to think. Second, answer it directly, in plain terms, before adding detail. Third, support the answer with evidence: a result, a citation or a backup slide. Fourth, return to what the study can claim, so the answer ends on firm ground rather than trailing off. Each answer is drafted to take about a minute aloud, because long answers invite follow-up questions about the parts the committee did not need.

Questions About Design

1. You use the word mediation, but your data come from one survey. Can you claim mediation? That is a fair challenge. The study estimated an indirect effect consistent with mediation, but one survey cannot establish that leadership comes first, safety second and voice third, and cross-sectional estimates of indirect effects can be biased (Maxwell & Cole, 2007). I have described the result throughout as consistent with the proposed pathway. A three-wave panel is my first recommendation for research.

2. Why not a longitudinal design from the start? Within the program's timeline and with access to three systems for one survey window, a single wave was what could be completed with an adequate sample. I chose to do one wave well, with preset screening rules and robustness checks, rather than two waves with heavy attrition.

Questions About Measurement

3. How do you know employees actually speak up as much as they say? I do not know it with certainty. Voice was self-reported, and self-reports can overstate behavior and share method variance with the other measures (Podsakoff et al., 2003). The survey separated the measures, varied response formats and included a marker item, and the adjusted correlations remained meaningful. Supervisor ratings or logged suggestions would be stronger, and I recommend them.

4. Why measure transformational leadership as a whole rather than specific behaviors such as openness? The global scale was short, validated and suited to a survey that had to stay under ten minutes. The cost is that I cannot say which behaviors mattered. Earlier work found openness to input especially important for voice (Detert & Burris, 2007), and separating behaviors is a recommendation in Chapter 5.

5. Your voice scale mostly measures suggestions. What about raising concerns? Correct. The measure captures promotive voice more than prohibitive voice, and meta-analytic work shows the two have partly different correlates (Chamberlin et al., 2017). In revenue cycle work, concerns about errors may matter most, so this is a genuine limit on what the results say about compliance and risk.

Questions About Analysis

6. Employees share supervisors. Why didn't you use multilevel models? To protect anonymity, I did not collect supervisor or team identifiers, so clustering could not be modeled. Ignoring it can make standard errors too small. The main effects were several standard errors from zero, so I do not expect them to change, but the borderline direct path could, and I treat it cautiously for that reason.

7. The direct path is not significant. Are you saying leadership does not affect voice directly? No. Its 95% interval stretches up to 0.38 and only just crosses zero, so a modest direct effect remains plausible. What the results show is that most of the observable association runs through safety (backup slide 9 shows the interval).

Questions About Meaning and Use

8. Is the effect large enough to matter in practice? Leadership and the controls explained about 13% of the variance in voice, a medium effect by convention, and leadership explained about a third of the variance in safety. In practical terms, employees whose supervisors rated a point higher on the five-point scale reported almost half a point more voice. Whether that matters depends on what employees say when they speak up; in revenue cycle work, one early report of a payer rule change can prevent weeks of denials.

9. Would these findings hold in other hospitals? They describe three systems in two states and roles that match those systems' workforces closely. Transfer to small rural hospitals or outsourced revenue cycle vendors is uncertain. The pattern agrees with research in very different settings, which gives some confidence that the basic pathway is not unique to these three systems.

10. What will you do with these findings in your own work? I plan to pilot two recommendations in my own department, supervisor practice in responding to audit errors and a visible suggestion log, in one team for a quarter, tracking denial rework and staff comments. My department was not in the study, so the pilot will also test whether the findings travel.

Questions I Cannot Fully Answer

Some questions will have no complete answer, and preparing for that is part of the defense. If asked why safety differed across systems, I can offer possible reasons, such as recent layoffs at one employer, but the study was not designed to explain system differences. The honest response is to say what the data show, offer a possible explanation clearly labeled as such and note it as a question for future research. Committees generally respect a candidate who can say that a question falls outside what the study can answer.

Rehearsing the Answers

Each answer has been spoken aloud and timed. In two practice defenses, colleagues asked the questions in a different order and with different wording, which revealed that answers 3 and 6 ran long and that answer 1 depended too heavily on a citation. Both were shortened, and answer 1 now leads with plain language before the reference. The backup slides referred to in the answers have been numbered to match. A final run-through will take place two days before the defense, with my chair playing the most skeptical committee member, so that the hardest questions are fresh in mind.

Conclusion

Ten likely questions, grouped by design, measurement, analysis and meaning, now have drafted answers that follow a four-step structure, cite evidence where it helps and end on what the study can claim. Preparation for questions without full answers has been planned as well. Module 4 builds the limitations and recommendations portion of the presentation.

References

Chamberlin, M., Newton, D. W., & LePine, J. A. (2017). A meta-analysis of voice and its promotive and prohibitive forms: Identification of key associations, distinctions, and future research directions. Personnel Psychology, 70(1), 11-71. https://doi.org/10.1111/peps.12185

Detert, J. R., & Burris, E. R. (2007). Leadership behavior and employee voice: Is the door really open? Academy of Management Journal, 50(4), 869-884. https://doi.org/10.5465/amj.2007.26279183

Maxwell, S. E., & Cole, D. A. (2007). Bias in cross-sectional analyses of longitudinal mediation. Psychological Methods, 12(1), 23-44. https://doi.org/10.1037/1082-989X.12.1.23

Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879-903. https://doi.org/10.1037/0021-9010.88.5.879