GB893 Unit 7 Defense Preparation Outline Example

Course	GB893 Doctoral Project IV: Project Dissemination
Unit	Unit 7
Paper type	Defense preparation outline
Length	About 870 words, 6 pages double-spaced
Format	APA 7 student paper
School	Purdue University Global
Program	MBA
Updated	October 2026

Free sample paper for GB893 Unit 7

Seven Questions Ranked by Damage: A Unit 7 Defense Preparation Outline for a Dock Appointment Study

[Student Name]

Purdue University Global

GB893: Doctoral Project IV: Project Dissemination

Unit 7 Defense Preparation Outline

[Instructor Name]

[Date]

The study, Haverly Logistics and every figure are composite and illustrative; the methods sources cited are real.

What this part is doingLeading with the hardest question shows the outline is ranked by risk, not by the order of the chapters.

Preface: The Weak Points, Ranked

The study's weak points, as the candidate sees them, are ranked by the damage a poor answer would do. Selection of the portal centers is first, because if the comparison fails, every estimate fails with it. The outcome switch is second, because it can look like a choice made to improve the result. The number of clusters is third. The candidate's own role is fourth, the interview shortfall fifth, the single firm sixth and the missing detention dollars seventh. Each question below has five parts: the question as a member might phrase it, the honest concession, the evidence, the appendix page that holds it, and a line on what not to say.

Question 1: Selection of the Portal Centers

The question. The six busiest centers got the portal. How can the six quieter ones tell you what would have happened without it?

Concession. The centers were not assigned at random, and portal centers had longer waits before adoption.

Evidence. The design compares changes, not levels, and center fixed effects absorb stable differences. Pre-period slopes differed by 0.003 log points a month, interval minus 0.005 to 0.011. Dropping DC-09, whose own second receiving shift began in month five, moves the estimate to about 5 percent. If asked about the recent literature, the candidate can note that adoption happened in one month for all six centers, so the problems with staggered timing that Roth et al. (2023) summarize do not arise, though their point that parallel trends cannot be proven still applies.

Appendix. C-2 and C-4.

Do not say. That the trends prove the assumption.

What this part is doingRecording what not to say for each question guards against overclaiming under pressure, which is when it is most likely.

Question 2: The Switch to Log Minutes

The question. You changed the outcome after seeing the data. Why should we believe the result?

Concession. The change was made after the data were seen.

Evidence. The change was dated in week three of analysis, before any estimate existed, because skewness in raw minutes was 1.77. The planned model is reported, at 6.2 minutes with interval 2.3 to 10.0, and points the same direction. The departure log records both.

Appendix. D-1.

Do not say. That log scales are standard, as if that settled it.

Question 3: Twelve Clusters

The question. With twelve centers, are your intervals believable?

Concession. Conventional clustered errors would not be.

Evidence. All intervals come from a wild cluster bootstrap, which Cameron et al. (2008) showed performs well with few clusters. The conventional interval would have been narrower, about 2 to 6 percent, and is not used.

Appendix. C-5.

Do not say. That 196,518 visits make the sample large. The effective sample for inference is the centers.

Question 4: The Candidate's Role

The question. You helped choose the portal. How do we know you were not invested in its success?

Concession. The candidate recommended the vendor and has an interest in the result.

Evidence. The analysis plan was filed before adoption, the departure log is complete, and the reported effect is modest. Brannick and Coghlan (2007) defend research by members of the organization studied, so long as the researcher states that position openly and examines what it may distort, and the methods chapter does both.

Appendix. A-1.

Do not say. That the candidate was objective.

Question 5: Six Dispatchers

The question. Only six dispatchers. Is the early-arrival theme trustworthy?

Concession. Dispatchers are underrepresented against the planned twelve.

Evidence. The yard data confirm the theme independently: early trucks waited about twice as long for a door. The chapter flags the theme as resting on fewer accounts.

Appendix. E-3.

Do not say. That saturation was reached.

Question 6: One Firm

The question. What does this add beyond Haverly?

Concession. The estimate describes one firm's centers.

Evidence. The mechanism generalizes even if the size does not: appointments change arrivals, not the work at the door. Any operator can test that pattern with its own segment timestamps. The difference-in-differences logic Angrist and Pischke (2009) describe makes the comparison transferable as a method.

Appendix. None; answered from the discussion chapter.

Do not say. That the result applies to the industry.

Every answer starts with a concession, because a committee trusts a candidate who names the weakness before being asked to.

How the Outline Will Be Rehearsed

Two mock defenses are planned, one with a classmate trained in econometrics who will ask Questions 1 to 3 without warning, and one with a Haverly general manager who will ask Questions 6 and 7 as a practitioner would. After each, the candidate will record which answers ran long and which concessions came late, and revise the outline before the real defense.

What this part is doingPlanning rehearsals with members who mirror the committee's training tests the outline against the questions most likely to be asked.

Question 7: Detention Dollars

The question. Your sponsor wanted savings. Why are there no dollars?

Concession. The study cannot measure them.

Evidence. Detention is invoiced monthly by carrier without visit numbers. The two-hour share is reported as exposure to detention, not its cost.

Appendix. D-3.

Do not say. Any estimated dollar figure.

References

Angrist, J. D., & Pischke, J.-S. (2009). Mostly harmless econometrics: An empiricist's companion. Princeton University Press.

Brannick, T., & Coghlan, D. (2007). In defense of being "native": The case for insider academic research. Organizational Research Methods, 10(1), 59-74. https://doi.org/10.1177/1094428106289253

Cameron, A. C., Gelbach, J. B., & Miller, D. L. (2008). Bootstrap-based improvements for inference with clustered errors. The Review of Economics and Statistics, 90(3), 414-427. https://doi.org/10.1162/rest.90.3.414

Roth, J., Sant'Anna, P. H. C., Bilinski, A., & Poe, J. (2023). What's trending in difference-in-differences? A synthesis of the recent econometrics literature. Journal of Econometrics, 235(2), 2218-2244. https://doi.org/10.1016/j.jeconom.2023.03.008

What the GB893 Unit 7 assignment asks for

The seventh GB893 assignment asks you to prepare for your defense by anticipating the hardest questions. Start with a preface ranking your study's weak points by the damage a poor answer would do, and order the outline by that ranking. For each question, write it as a committee member might phrase it, give an honest concession, set out the evidence that bounds the weakness, point to where it appears in your appendices and note what you should not say. Cover design, departures from the plan, inference, your own position, sample limits, generalization and anything your sponsor expected but did not receive. Cite methods sources that support your answers. Describe how you will rehearse. Write in outline form, with consistent parts under each question. Nine pages is common for this outline, with each answer short enough to say aloud.

How the GB 893 Unit 7 example is put together

The sample's preface ranks seven weak points. Question one, selection of the busiest centers, concedes non-random adoption and answers with fixed effects, a pre-period slope difference of 0.003 and the DC-09 model, noting through Roth and colleagues that staggered-timing problems do not arise. A margin note explains the do-not-say lines. Questions follow on the log outcome, twelve clusters with Cameron and colleagues, the candidate's role with Brannick and Coghlan, six dispatchers and one firm with Angrist and Pischke. The highlighted sentence explains why every answer begins with a concession. A rehearsal plan with two mock defenses precedes the final question, on detention dollars, which concedes they cannot be measured.

Reading the GB893 Unit 7 grading rubric

The GB893 defense outline is graded on anticipation, honesty and preparation. Instructors give credit for ranking weak points by risk. Each question should be phrased as a member would ask it. Concessions earn credit when they come first and are precise. Evidence should bound each weakness with figures and appendix references. Lines on what not to say show awareness of how candidates overclaim under pressure. Coverage of design, departures, inference, the researcher's position, sample, generalization and unmet expectations is expected. Methods sources should support the answers. A rehearsal plan with suitable mock examiners shows readiness. The outline should be short enough to use during preparation and should match the figures reported in the results chapter.

Common GB893 Unit 7 mistakes, and how to avoid them

Many GB893 defense outlines list easy questions with long, defensive answers. Others prepare for questions in chapter order rather than by risk. Students often skip their own role in the organization studied, though committees often ask about it. Concessions are avoided, which invites follow-up. Answers cite no figures or appendix pages. Nothing warns against the overclaims candidates make under pressure. Rehearsal is left to chance. Rank your weak points, phrase questions as members would, concede first, bound each weakness with evidence and pages, note what not to say, address your own position and plan rehearsals with people who mirror the committee. Bring your results chapter, and our team can help draft the outline.

Get a GB893 Unit 7 example written to your instructions

List the weaknesses you already worry about in your project, even ones you hope nobody raises, and include the Unit 7 prompt and rubric with your defense format. The outline we draft ranks the likely questions, pairs each with an honest concession and evidence, and flags the deflections to avoid. Free as a first custom sample; 24-48h. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.

More GB893 papers and MBA samples

GB893 Unit 7 questions, answered

How can a candidate predict defense questions?

By starting from wherever the study is thinnest, which its author usually knows better than anyone. Design choices, departures from the plan, sample limits and the candidate's position in the organization cover most of what committees ask. Reviewing written comments from earlier committee meetings helps too, since members often return to concerns they raised before.

Should a defense answer concede a limitation?

Usually first, and precisely. Stating what a limitation costs, then the evidence that bounds it, reads as command of the project. Deflecting or minimizing invites a follow-up that is harder to answer. Concession is not collapse: an accurate concession often ends the line of questioning faster than a defense would.

How should an insider researcher address a conflict of interest?

Directly and early. Name the role, then point to the safeguards: an analysis plan fixed before data were seen, a departure log, code others can rerun, and any independent checks. Committees rarely expect the interest to vanish; they expect evidence that it was managed and that the results were allowed to disappoint.

Where can I find a free GB893 Unit 7 sample paper?

This page carries a full GB893 Unit 7 defense preparation outline for a dock appointment study, ranking seven committee questions by the damage a weak answer would do.

Should a defense answer start with a concession?

Usually. Naming the weakness precisely, then giving the evidence that limits it, shows command of the study and removes the reason for a follow-up question.