Use AI for research, with real citations

A chatbot predicts what a citation looks like, it never looks one up. We resolved 218 model-written DOIs against CrossRef and one in five from the best model was unusable.

What actually cuts fabricated references in ChatGPT: turning on search, supplying your own sources, better prompts, and the one limit no prompt can fix.

Most plagiarism isn't theft, it's drift: a paraphrase that stayed too close, a note you forgot was a quote. The habit that prevents it, and how to handle AI text honestly.

The source exists, the DOI resolves, and it still doesn't support your sentence. Why this failure is more common than fabrication, and how to catch it.

A DOI can be perfectly formatted and still resolve to nothing. Five different causes, and the decision tree that tells them apart in three minutes.

The article exists, but is the journal real scholarship? Predatory journals, hijacked clones, and paper mills, plus a five-minute check on any venue.

How to check retraction status in seconds, how to read a retraction notice, and what to do when a source you already cited has been withdrawn.

One paper in 277 now carries a fabricated reference, up from one in 2,828 in 2023. What the audits measured, and why every number is a floor.

It resolves, the publisher is right, and the article is not yours. We measured how often one wrong digit still lands on a real paper.

An empty DOI field is correct. How to check whether it truly has none, and what APA, MLA and Chicago each ask for instead.

The as cited in rule in three styles, one example worked all the way through, and when a secondary citation is not allowed.

Coursework, a published article, a thesis or your own data. What each looks like, and the bracket that makes reuse honest.

Yes, as often as the reader needs, and it still takes one entry in the reference list. What APA, MLA, Chicago, Harvard and IEEE each do on the second citation, and where two of them flatly contradict each other.

Search four words from the middle, not the sentence. Why quotes fail more often than references, and what to do if you submitted one.

Almost certainly not trouble. What actually goes wrong, the two honest exits, and the measurement of how ordinary it is.

No style manual sets a length. APA, MLA and Chicago give no word count, library guides say about 150 words an entry, and your brief wins. Three worked entries.

Reverse-search a sentence you already wrote to find real backing, using free tools. And what to do when no source supports it.

Nobody agrees and nobody shows their working. What university guides really say, where they contradict each other, and the rule that actually decides it.

The five-year rule is folklore. Seven university guides give different numbers and none shows its working, APA publishes no cutoff at all, and 51% of citations in the humanities go to papers at least ten years old.

The legal ladder, fastest first, with how often each rung works. Plus the tool half these guides still recommend that shut down in 2025.

Researchers pulled 139 papers with undeclared AI use straight out of Scholar results. Where they sat, and what the index does not check.

Usually yes, with two conditions nearly every guide skips, plus the two-minute check for whether it has since been published.

A thesis statement is one or two sentences stating your specific, arguable answer to your question. A simple formula to build one, with before-and-after examples.

About 10 percent of the paper, and five university guides land on the same share. What that is in words at every length, and the one sentence that matters more.

How long should an abstract be? APA caps it at 250 words, and five universities publish caps from 150 to 350. Every number quoted, and the five parts to cover.

How long is a term paper? Eight to fifteen double-spaced pages, about 2,000 to 4,000 words, in five published assignment sheets, plus what the rubric pays for.

An essay argues a point from your own reasoning; a research paper investigates a question grounded in many real sources and cites every one. The difference, side by side.

A calm, hour-by-hour plan for writing a genuinely good essay in one day, even the night before, with real sources and no fabricated citations.

Topic to defense, broken into calm steps: the proposal, the chapter structure, the literature review, methodology, and citing real papers. Plus what changes at master's level.

A clear section-by-section template for a research proposal, plus a worked example, honest length ranges, and the mistakes that get proposals rejected.

AI can draft a research paper in minutes. Whether it's usable, and whether you can submit it, comes down to the citations and whether you can stand behind the work.

A generator can draft fast, but only one that retrieves and reads real papers is usable. What to look for, and how to write a review you can defend.

ChatGPT, Perplexity, Elicit, Consensus, Grammarly and Scribbr, each judged on the one job it owns, with the cases where we would tell you to buy them instead.

Jenni is fast and fluent, but for academic work the references have to be real and your text has to stay yours. Why students switch, and how verify-first citations and reviewable diffs differ.

SciSpace is excellent at reading the literature: chat with a PDF, search millions of papers, compare findings. But writing is a different job. Why writers switch, and how verify-first citations differ.

Writing quality, citations, deep research, and price, compared for a student writing a paper. Plus the problem both chatbots share.

Gemini or ChatGPT for a research paper in 2026? What each one costs a student, what audits measured about their citations, and the check neither one removes.

Grammarly is a general writing checker. Academic work needs journal register, real sources, and reviewable edits. The alternatives, sorted by which problem you have.

Scite shows how a paper has been cited, supporting or contrasting. CiteOwl writes yours. Including what Scite does that nothing else does.

Which one to use for which half of the paper, what grounding really buys you, and the one row where both answers are no.

Rarely the way a chatbot does. What the audits found instead: a real link sitting under a sentence the source never supports.

It rarely breaks the link. It breaks the attribution, and journalists grading 675 answers found a sourcing problem in 72% of them.

The honest answer is "it depends on what you use it for." Here is how universities tend to split allowed help from misconduct, and how to stay on the right side.

What AI detectors actually measure, why they flag plenty of genuine human writing, and what you can and can't read into the score before you hit submit.

Anthropic began marking Claude's text on 2 August 2026 and Google has marked Gemini's for a year. What the mark is, how much of it survives an edit, and why nobody grading you can read it.

Three different things hide in a document: invisible characters, file metadata, and a statistical watermark. How to inspect the two you can, and what the third is worth.

A pattern in the pixels survives almost anything. The signed record beside it dies to a screenshot. What that means for a figure in your paper, and what journals actually ask for.

What the Turnitin AI writing percentage counts, what it cannot see, how it behaves on human writing, and what a score does and does not prove.

What evidence actually helps, what an academic integrity meeting is like, and why the right move differs depending on what really happened.

You cannot prove a negative, but you can show how the paper came to exist. What version history, drafts, a source list and a reading trail prove to a marker.

Does Canvas detect ChatGPT? Not by itself. Canvas keeps page views, an access report, quiz logs and a submission timestamp, and Instructure says in writing not to read them as evidence of cheating.

Does Google Docs show if you copy and paste? It never labels a paste, but the revision history records every edit with a timestamp, and a block that lands in one step looks like one.

Turnitin names seven Claude versions it covers, and stops. What that list tells you, and why Anthropic's own watermark is a second system nobody can read.

Turnitin names seven Gemini versions on its published list. Which ones, the four Google has shipped since, and what the published tests actually measured.

There is no pass mark, and Turnitin says so. What the percentage counts, what your instructor changed before you saw it, and what twelve real universities publish.

Turnitin names QuillBot in its own documentation. What it is really finding, why the one honest study is out of date, and where every circulating percentage came from.

Turnitin answers this on its own FAQ page, and the answer has two halves. Grammar fixes are not what its detector targets. Generative rewrites are a different case.

Your syllabus asks you to disclose AI use and never says how. The six things institutions want, five templates, and which policies would still refuse each one.

Nine university policies, quoted. What needs permission, what does not, and why Turnitin matches you against yourself.

The number on the public benchmark, and why it cannot tell you whether the flag on your own essay was right.

Storage is your instructor setting, not yours. The three options, what another institution sees, and how a paper comes back out.

What published procedures actually say about a first, minor, unintentional case, and the sequence that works from here.

Not by itself, and the line is not about the tool. What universities permit in writing, and where it becomes plagiarism.