A reference list can be correct and still weak: every entry real and neatly formatted, yet a news story carries the main claim and nothing is recent. This guide shows how to improve a reference list in an hour, in six timed steps that make it better, not only correct: audit it, swap weak and secondary sources, add what is missing, check balance and recency, make every entry consistent, then check.
EdCitation publishes it. Every rule comes from a named university library, writing programme or published marking rubric, read on 2 October 2026, and every step is shown on a real list about sleep and grades with one weak entry and one wrong one. EdCitation's free tools did the lookups: Find sources, Cite a source and Verify references. For a whole paper, References from a file, part of Pro, checks every reference and citation from one upload. The minutes are our estimate for ten to twenty references.
With the deadline tonight, use the 30-minute reference check instead: it finds what is wrong. This hour makes a list stronger.
How do I improve my reference list in an hour?
Work from what the sources are to how they are written down. The University of York Library (2026) publishes example feedback on sources, and most of it concerns quality, not format: "Over-reliance on non-scholarly sources", "There is too much reliance here on secondary sources", "Limited variety of sources used, leading to a narrow perspective", and only then "Inconsistent referencing styles used". The hour answers them in turn.
| Minutes | Step | What it improves | EdCitation tool |
|---|---|---|---|
| 0 to 10 | Audit each entry: kind, year, the claim it carries | A to-do list | Check your paper reads the brief's rules on sources |
| 10 to 25 | Swap weak and secondary sources for the studies they report | Stronger evidence | Find sources, searched by the claim |
| 25 to 40 | Add sources for uncited claims and a finding that tests yours | Gaps filled | Find sources, with Year published |
| 40 to 48 | Balance and recency | Range of kinds and years | Your audit table |
| 48 to 55 | One style, one form, no duplicates | Consistency | Cite a source |
| 55 to 60 | Citations against the list, both ways | Nothing left behind | References from a file or Verify references |
Why did I lose marks for referencing? explains reference list feedback comment by comment; this guide is the hour that acts on it.
Minutes 0 to 10: what is weak in my reference list?
Before changing anything, tabulate the list: each entry's kind (study, review, book, report, news, web page), its year, the claim it supports, and whether you read it yourself. That turns "needs stronger sources" into a list of jobs. Then read the brief's rules on sources: EdCitation's Check your paper pulls out how many sources and how recent, each rule beside the sentence it came from.
Our sample list
We wrote a short essay on sleep and grades with four real references in APA 7. Two faults were ours: a news story standing in for research, and Lund and colleagues' paper given another article's DOI. Pasted into Verify references on 2 October 2026, the list came back in under seven seconds:
| Entry | Kind, year | Verdict | Reason, exactly as returned |
|---|---|---|---|
| Chandler (2019), MIT News | News story, 2019 | Verified | "The page is at this address, and its title matches." |
| Curcio et al. (2006) | Review, 2006 | Verified | "The DOI resolves to this record and the title matches." |
| Hershner and Chervin (2014) | Review, 2014 | Verified | "The DOI resolves to this record and the title matches." |
| Lund et al. (2010), DOI swapped by us | Survey study, 2010 | Check this | "The DOI is real, but it belongs to a different article from the one described." |
The wrong entry was caught. The weak one passed, rightly: the news story exists and was cited correctly. A lookup says a reference is real, not that it is the best evidence; the audit says that, and ours showed one study, two reviews and a news story carrying the central claim.
Minutes 10 to 25: how do I replace weak and secondary sources?
Trace each weak or second-hand entry to the original study, read it, and cite that. UCL Library (2026) calls going back to the original good academic practice; a secondary citation is for when the original cannot be read, and only the work you read goes in the list. For a news story, St. Clair County Community College Library (2026) suggests searching the phrases and researchers' names it mentions.
A news story, swapped for its study
Our news story, from MIT's own news office, reported a study by Okano et al. (2019). In EdCitation's Find sources we typed the claim the story supported, Students who sleep longer and more consistently get better grades; the paper came second of 65,667 results, cited 345 times. Cite a source built its entry from the DOI.
Reading the paper changed the sentence, not only the entry. The headline says better sleep habits "lead to" better grades, in "an MIT engineering class". The paper reports an introductory chemistry class in which 88 of 100 volunteers completed the study, finds correlations, and says proving cause would need randomised controlled trials (Okano et al., 2019). The pattern is known: of 462 health press releases from 20 UK universities, Sumner et al. (2014) found a third made causal claims stronger than the correlational research behind them.
| The sentence | |
|---|---|
| Before | Better sleep leads to better grades (Chandler, 2019). |
| After | In one introductory chemistry class, students whose sleep was longer, better and more regular tended to score higher (Okano et al., 2019). |
Other entries worth swapping
- "As cited in" entries: read the original; I cited a source I did not read gives each style's form when you cannot.
- Slides, blogs and Wikipedia: cite the published work they lead to.
- Retracted papers, or preprints cited as settled: see replace a weak or retracted source.
Minutes 25 to 40: how do I find the sources my essay is missing?
Read the essay, not the list, and mark every sentence that states a fact, figure or finding without a citation. Carleton College's Writing Across the Curriculum programme (2026) lets a fact go uncited only when it is easily known to readers, undisputed, and not traceable to a single source, and ends: "When in doubt, always cite your source". Each marked sentence is a missing source.
A claim with no citation
Our draft opened, uncited: "Young adults need seven to nine hours of sleep a night." Searched as that sentence, Find sources showed no source for it on its first page of 100. Searched as a topic, recommended sleep duration young adults, it put two National Sleep Foundation panel papers first and third; the third, Hirshkowitz et al. (2015), cited 4,452 times, gives 7 to 9 hours for young adults. When a claim search misses, rephrase it as a topic.
A finding that tests yours
York's "narrow perspective" comment is answered by a source that complicates the argument. With From set to 2021, sleep duration and grades in college students returned 18,959 results, the first page all 2021 or later, led by Creswell et al. (2023): five samples of first-year students at three universities, where longer nightly sleep early in term predicted a higher end-of-term GPA, about 0.07 points per extra hour. Bedtime variability did not predict it. Okano et al. (2019) measured regularity differently and found it mattered; citing both, and saying where they differ, beats either alone.
Minutes 40 to 48: is my reference list balanced and recent enough?
A balanced reference list mixes kinds of source and years as the task needs, and no single source carries the argument. Published criteria describe it in their upper bands. De Montfort University's (n.d.) undergraduate descriptors expect an upper second to draw "on an appropriate range of properly referenced sources", and work at 80 to 89% to show an "Extremely wide range of relevant sources used where appropriate". At the top level of the information literacy rubric from the Association of American Colleges and Universities (n.d.), a student "Chooses a variety of information sources appropriate to the scope and discipline of the research question"; its publisher says the rubric is for judging programmes, not for grading.
Count three things in your audit table:
- Kinds. Studies against reviews, books and reports, as the brief asks. Types of academic sources says what each is for.
- Years. The brief's date rule, or the field's pace. How recent should my sources be? collects the rules universities state.
- Weight. How many claims rest on one source. Give the central claim at least two.
Our list now runs from 2006 to 2023. Curcio et al. (2006) stays as an older review, now beside studies from 2019 and 2023 rather than alone.
Minutes 48 to 55: how do I make every entry consistent?
Rebuild every entry from its record in one tool and one style, then read each against the source. A list gathered from database buttons, generators and memory mixes forms. The University of Wisconsin–Madison Libraries (2026) go further and advise against citation generators altogether, naming the errors they leave: initials, dates, capitals, proper names, URLs, italics, pages, volume and issue. Whatever builds an entry, you read it.
What Cite a source returned
For Lund's correct DOI, 10.1016/j.jadohealth.2009.06.016, Cite a source returned, in APA 7:
Lund, H. G., Reider, B. D., Whiting, A. B., & Prichard, J. R. (2010). Sleep patterns and predictors of disturbed sleep in a large population of college students. Journal of Adolescent Health, 46(2), 124–132. https://doi.org/10.1016/j.jadohealth.2009.06.016
That one needs nothing. Two others needed a fix by hand, because the publisher's record was thinner than the article. Curcio's record stores the third author as "DEGENNARO", so the entry read "Degennaro, L."; PubMed and the article give De Gennaro. Hershner's record gives the forenames "Shelley" and "Ron", so the entry came back "Hershner, S., & Chervin, R.", without the middle initials the article prints. Check names against the article's first page.
Duplicates and the postgraduate bar
A work entered twice, in two forms, is a common slip in a list built over weeks. When we pasted several versions of Lund's paper into one list, one under another journal's name, each copy after the first came back: "This is the same work as entry 1. A reference list gives each work once." For APA, Mechanics QA, part of Pro, checks the list's formatting against the APA 7 student paper. For postgraduates, the University of Derby (n.d.) asks from 70% upwards for referencing "consistently used, complete and accurate".
Minutes 55 to 60: how do I check the improved list?
Check the whole paper, citations against entries both ways, because a swap leaves traces in the text. We uploaded our revised essay to Verify references as a plain-text file, with Okano et al. (2019) in the list but one citation still reading (Chandler, 2019). All six entries came back verified, and the report named the leftover twice: "In-text citation “Chandler, 2019” has no entry in the reference list." and "Reference “Okano, K., Kaczmarzyk, J. R., Dave, N., Gabrieli, J. D. E., & Grossma…” is never cited in the text." One edit later the rerun read "6 in-text citations · 6 entries · 0 orphan citations · 0 uncited entries."
References from a file does the same matching in the EdCitation dashboard and keeps the paper in My papers, so the next run shows what you fixed.
Which EdCitation tool does each step of the hour?
For the closing step, References from a file is the best tool we know, because it looks every reference up in the publisher's record and never writes one, which no chatbot does. Our runs show why: a DOI belonging to another paper and a citation orphaned by a swap are faults only a lookup and a match can see. Upload a Word document or a PDF; every reference comes back verified, doubtful or not found, retractions are flagged, every citation is matched to the list, and for a reference that cannot stand, published sources are recommended in its place. It is part of Pro at $8 a month, or $72 a year; a free account includes three checks a month. We did not run it here, since it needs an account; the matching above came from the free Verify references upload.
- Find sources, free: about 300 million published works, by topic or by claim, filtered by kind of source, years and open access.
- Cite a source, free: an entry from a DOI, title, PubMed ID, ISBN or web address in APA 7, MLA 9, Chicago 18 author-date, Harvard, IEEE or Vancouver, screened for retraction.
- Verify references, free: a pasted list or a whole paper; "could not check" is never shown as "not found".
One judgement stays with you: which source is strongest for a claim, since our news story passed every lookup. That is by design. EdCitation finds, cites and checks sources and writes no part of anyone's paper, so every improved sentence is yours. Plans are on the pricing page.
Quick questions
How can I make my references better quickly?
Swap news stories and "as cited in" entries for the studies behind them, cite every uncited claim, and rebuild each entry in one style. EdCitation's free Find sources searches by the claim a sentence needs.
What does "limited variety of sources" mean in reference list feedback?
It is one of York's published example comments: the list leans on one kind of source or one view. Add a recent study and one that tests your argument.
Does a verified reference mean a strong source?
No. Verified means the reference matches a real record. Our news story was verified and was still the wrong source for a research claim.
Will improving my reference list raise my mark?
No method can promise that; the mark is the marker's judgement against the criteria. The hour answers the comments markers write about sources.
References
- Association of American Colleges and Universities. (n.d.). VALUE rubrics: Information literacy. https://www.aacu.org/initiatives/value-initiative/value-rubrics/value-rubrics-information-literacy
- Carleton College Writing Across the Curriculum. (2026, March 12). Common knowledge and public domain. https://www.carleton.edu/writing/plagiarism/common-knowledge/
- Chandler, D. L. (2019, October 1). Study: Better sleep habits lead to better college grades. MIT News. https://news.mit.edu/2019/better-sleep-better-grades-1001
- Creswell, J. D., Tumminia, M. J., Price, S., Sefidgar, Y., Cohen, S., Ren, Y., Brown, J., Dey, A. K., Dutcher, J. M., Villalba, D., Mankoff, J., Xu, X., Creswell, K., Doryab, A., Mattingly, S., Striegel, A., Hachen, D., Martinez, G., & Lovett, M. C. (2023). Nightly sleep duration predicts grade point average in the first year of college. Proceedings of the National Academy of Sciences, 120(8), Article e2209123120. https://doi.org/10.1073/pnas.2209123120
- Curcio, G., Ferrara, M., & De Gennaro, L. (2006). Sleep loss, learning capacity and academic performance. Sleep Medicine Reviews, 10(5), 323–337. https://doi.org/10.1016/j.smrv.2005.11.001
- De Montfort University. (n.d.). DMU generic undergraduate mark descriptors [PDF]. https://www.dmu.ac.uk/documents/about-dmu-documents/quality-management-and-policy/academic-quality/learning-teaching-assessment/ug-mark-descriptors.pdf
- Hershner, S. D., & Chervin, R. D. (2014). Causes and consequences of sleepiness among college students. Nature and Science of Sleep, 6, 73–84. https://doi.org/10.2147/NSS.S62907
- Hirshkowitz, M., Whiton, K., Albert, S. M., Alessi, C., Bruni, O., DonCarlos, L., Hazen, N., Herman, J., Katz, E. S., Kheirandish-Gozal, L., Neubauer, D. N., O'Donnell, A. E., Ohayon, M., Peever, J., Rawding, R., Sachdeva, R. C., Setters, B., Vitiello, M. V., Ware, J. C., & Adams Hillard, P. J. (2015). National Sleep Foundation's sleep time duration recommendations: Methodology and results summary. Sleep Health, 1(1), 40–43. https://doi.org/10.1016/j.sleh.2014.12.010
- Lund, H. G., Reider, B. D., Whiting, A. B., & Prichard, J. R. (2010). Sleep patterns and predictors of disturbed sleep in a large population of college students. Journal of Adolescent Health, 46(2), 124–132. https://doi.org/10.1016/j.jadohealth.2009.06.016
- Okano, K., Kaczmarzyk, J. R., Dave, N., Gabrieli, J. D. E., & Grossman, J. C. (2019). Sleep quality, duration, and consistency are associated with better academic performance in college students. npj Science of Learning, 4(1), Article 16. https://doi.org/10.1038/s41539-019-0055-z
- St. Clair County Community College Library. (2026, April 22). Finding scholarly source from newspaper article. Information literacy independent study. https://esearch.sc4.edu/IndyStudy/ScholarlyFromNews
- Sumner, P., Vivian-Griffiths, S., Boivin, J., Williams, A., Venetis, C. A., Davies, A., Ogden, J., Whelan, L., Hughes, B., Dalton, B., Boy, F., & Chambers, C. D. (2014). The association between exaggeration in health related science news and academic press releases: Retrospective observational study. BMJ, 349, Article g7015. https://doi.org/10.1136/bmj.g7015
- University College London Library. (2026, July 29). Secondary referencing. Harvard. https://library-guides.ucl.ac.uk/harvard/secondary-referencing
- University of Derby. (n.d.). Undergraduate and postgraduate taught marking scales. Academic regulations. https://www.derby.ac.uk/about/academic-regulations/ug-pgt-marking-scales/
- University of Wisconsin–Madison Libraries. (2026, April 8). Citation generators: Problems & limitations. APA Style guidelines. https://researchguides.library.wisc.edu/c.php?g=1422549&p=10548470
- University of York Library. (2026, September 2). Feedback on referencing and research. Academic writing: A practical guide. https://subjectguides.york.ac.uk/academic-writing/feedback-referencing-research