GLAS7999 Chap.10 Assemble and Defend the Prospectus
Assemble and Defend the Prospectus
Make the design a reversible chain
The prospectus connects question, field position, sources, method, ethics, feasibility, limits and contribution. This chapter drafts in dependency order, audits nouns and verbs across sections, and separates expected contribution from predicted findings.
It then turns the oral presentation into a defence of choices, alternatives, boundaries and revision triggers, so peer questions improve the design rather than merely adding clarification.
Draft in dependency order
Begin with a compact logic paragraph before writing full sections.
State the question, the field relation it changes, the materials needed, the operation applied to them, and the boundary of the resulting claim. Read those sentences as a chain. If the literature describes a different object from the question, or the method cannot operate on the available sources, repair the link while revision is inexpensive.
The final prospectus may use another heading order, but its reasoning still depends on earlier choices. Background earns space only when it explains why the research problem exists or why a design decision is necessary.
Separate intervention from prediction
A proposed contribution should survive more than one possible finding.
A project can connect two debates, test a period boundary, recover a neglected mechanism, or show how a method changes the visible object without promising what the evidence will reveal. This distinction protects openness. Write the expected intervention as an operation on existing knowledge, then list at least two findings compatible with it.
If only the preferred result appears worthwhile, the question may already contain its answer. A defensible prospectus promises a disciplined inquiry and a consequential relation to scholarship, not a guaranteed discovery.
Audit coherence in both directions
Forward reasoning asks what evidence and method the question requires. Backward reasoning asks what scale and verb the actual evidence can support.
Run both tests. A question about audience reception cannot be answered by publicity material alone, while a discourse analysis of publicity can answer how an imagined audience is constructed. Repair either the question or the evidence plan. Then track key nouns and verbs across sections. Small shifts from platform to industry, or from represents to causes, often disclose that adjacent sections belong to different projects.
Consistent vocabulary is not cosmetic; it records a stable object and burden of proof.
Make feasibility intellectually active
Language, access, ethics review, time and preservation conditions should alter the design rather than sit in a closing disclaimer. Place each constraint beside the decision it affects.
A delayed archive may change the case; unavailable participant access may change a reception question into an institutional one; uneven language capacity may require a narrower comparison. For each condition, record a revision trigger and the earliest link it would change. A realistic timeline follows dependencies: permission before collection, collection before analysis, and analysis before claims.
A schedule that ignores those dependencies is evidence of incoherence, however polished its dates appear.
Defend choices, not headings
Prepare oral answers around need, option and revision trigger. Explain what the project requires, why the selected approach fits better than the strongest alternative, and what evidence would make the choice change.
When challenged about a small corpus, identify the inference it supports before conceding the stronger inference it cannot support. When asked why a period begins or ends, point to a mechanism or institutional change rather than a convenient decade. Questions at the seams between field, sources, method and contribution are especially valuable because they test design.
A precise concession strengthens the defence by showing that the boundary was anticipated.
Finish with a decision record
Preserve peer comments, rejected alternatives and the reasons for revision. The prospectus is not a contract that forbids later learning; it is a baseline from which changes can be explained.
Before submission, write one sentence beneath every section stating what it needs from the previous section and what it enables next. Remove generic material that performs neither function. The finished document should make the next research action obvious, identify the assumptions that evidence will test, and remain revisable when the archive produces surprise.
That combination of coherence and openness is the central standard.
What this chapter covers
- 01
Question-to-method chain
- 02
Field intervention
- 03
Evidence and method fit
- 04
Feasibility timeline
- 05
Contribution without prediction
- 06
Oral defence
Worked example · free
Defend a single-case design
- 2State the claim type.
- 3Justify case selection and access.
- 3Concede the inference boundary.
Key terms
- Prospectus
- A justified plan linking an answerable question to literature, evidence, method, ethics and contribution.
- Coherence
- Consistency of object, claim and logic across every section of the research design.
- Revision Trigger
- Evidence or a constraint that would require a stated design choice to change.
Assemble and Defend the Prospectus FAQ
What should the prospectus establish?
It should establish that the project is coherent and feasible, not that the final conclusion is already known. Every section should answer the same research problem and acknowledge its inference limits.
How should I state the contribution?
Name the operation the project performs in a conversation: connects, tests, revises, compares, historicises or reconstructs. Phrase it so contrary findings can still produce a worthwhile scholarly result.
How do I prepare for oral questions?
For each choice, state what the project needs, why the selected option fits better than the main alternative and what evidence or constraint would trigger revision. Practise precise concessions about scope.
Assessment move
Write one sentence each for question, field relation, sources, method, ethics, limit and contribution. Read them as one paragraph, repair every mismatch, then rehearse the strongest objection to each choice using decision, boundary and revision trigger.