Quality and access worksheet
Establish status without assumptions
A PDF carrying a university logo is not necessarily the accepted dissertation. It may be a proposal, a deposited author manuscript or a book revised after examination. Check the title page, repository record and library catalogue together. Degree wording, faculty, examination date and publication date help identify the object, although conventions differ by institution and country. State only what the document or an authoritative catalogue establishes.
A doctorate is examined, but its procedure should not be described as though it were automatically identical to a journal's anonymous peer review. Examiner roles, public defence and publication conditions vary. Assess the source for the job at hand. A dissertation may provide the fullest account of an instrument, while a subsequent article may supply the more current analysis.
Read the research design, not just the conclusion
Locate the operational research question, source or participant selection, data origin, measures, analytical steps and acknowledged uncertainty. In qualitative work, inspect sampling, collection, coding and reflexivity. In quantitative work, examine the sample, variables, model choices, missing observations and robustness checks. A long methods chapter creates space for transparency, but length itself is not evidence of sound decisions.
Test whether the conclusion is proportionate to the results. Separate original findings from literature review, theoretical framing and speculation. Follow decisive secondary claims back to their cited originals where possible. A dissertation's bibliography is often an excellent discovery trail, yet it does not independently verify every summary in the thesis.
Map later outputs to the thesis
Doctoral projects commonly produce journal articles, monographs, conference papers and datasets. Do not assume that a later title is a duplicate or definitive replacement. Compare chapters, samples, analyses and results. The article may report a narrow part, revise the model or introduce co-authored work. Cite the dissertation when you rely on its distinctive detail; cite the paper when your claim comes from its updated analysis. If both matter, make their roles explicit so one study is not counted twice.
Look for corrections and subsequent scholarly discussion, but avoid using citation counts as a proxy for truth. For an older dissertation, ask whether terminology, measurements or background data have been superseded. The work can remain important historical evidence while being a poor source for a present-tense generalisation.
Choose an identifiable, durable copy
Prefer the institution's repository, a national library catalogue or a clearly designated publisher edition. A personal upload may lack annexes, use different pagination or pre-date examination. Capture the persistent identifier, institution, year and access date. If confidential appendices or embargoed parts are unavailable, disclose that limit rather than implying that the complete evidence was inspected.
Open the file from the saved reference and test every pinpoint citation. Repository and printed editions can paginate differently. Do not take the quote from one and the page number from another. Preserve a version date or filename in your research log so another reader can reconstruct the choice.
Write a proportionate source judgement
A useful evaluation states both contribution and limit: for example, the thesis offers the only full account of the sampling procedure but predates a later validation study. That is more informative than calling all dissertations “authoritative” or “grey literature”. Tie the judgement to the passage being cited. Rich methodological detail may be excellent support for how the study was conducted while the same thesis remains insufficient for a broad claim about current prevalence.
Build the reference from the consulted title page and official metadata. Include the document type and institution as required by the selected style, plus DOI, handle or repository URL where supplied. Keep the citation stable throughout the manuscript. If you switch to a published book version, repeat the content and locator checks instead of merely replacing the bibliography entry.
Cross-check a pivotal claim
For a claim central to your argument, build a second evidence trail. Look for the underlying dataset, a paper produced from the doctorate or independent work that tests the same method. Differences do not necessarily indicate error: sample, analytical cut-off or definitions may have changed. Record the explanation you can verify. Where later papers report only selected positive results, the thesis may remain the fuller account of the project, including null findings and abandoned analyses.