Practical write-ups on literature review, reading papers, writing, and putting AI to work in research.
A complete annotated bibliography example — three real entries written out in full, a breakdown of what each sentence in an annotation is doing, and how the format changes between APA, MLA, and Chicago.
There is no single best paper search site. Here is what Google Scholar, PubMed, arXiv, Semantic Scholar, OpenAlex and the national indexes are each good and bad at, and how to combine two or three of them into a search you can defend.
A decision guide for EndNote versus Zotero — institutional licensing and support, what it costs to leave, Word and LaTeX workflows, and collaborating outside your institution — plus the job neither tool is built for.
The fastest routes to a BibTeX entry — one curl command against the DOI, the export links on arXiv and ACL Anthology, or a one-press copy from your own library — plus the six things that are usually wrong with whatever you get back.
Every reference answers the same five questions, and the style only decides the order and the punctuation. Worked examples for journal articles, books, chapters, web pages and theses in APA, MLA and IEEE, plus how in-text citations map to the list.
The two routes to a reference list in LaTeX — a .bib file with BibTeX, and biblatex with Biber — with working code, the cite commands, the natbib styles, and the exact compile order that makes the list appear.
Keyword search finds papers that share your vocabulary. Literature mapping follows citations instead, which is how you find the work that matters but uses different words. Here is the method, and when to stop.
An outline decides the order of your claims, not the order of your papers. Here are the four ways to organize the body, what belongs in each section, a skeleton you can copy, and three tests that catch a broken outline before you write 5,000 words.
What it actually takes to connect a Zotero library to Claude or ChatGPT over MCP — the local API, the third-party servers, a worked config — and when a per-paper handoff is the better fit.
A practical comparison of Zotero and Mendeley for researchers — storage, plugins, LaTeX workflows, group libraries, and who owns your data — plus the one job neither of them does.
A repeatable way to run a literature review: fix the question first, collect in one place, take notes against the question, and only then write the synthesis.
Stop reading papers front to back in one sitting. Read them three times instead — once to discard, once to get the claim, once to check whether you could reproduce it.