Twenty papers. One conversation.
Drop the whole reading list — assigned papers, lecture notes, textbook chapters, preprints. HOOH AI reads each one, remembers every figure, and answers across the stack with the page in hand.
Twenty papers. One conversation.
Drop the assigned reading, the methods chapter, the syllabus and your draft. Ask which two papers disagree on sample size. HOOH crosses every PDF and answers with the page in hand — no skimming, no skipping, no hallucinated quotes.
- PDFPaperSmith 2023
- PDFPaperLiu 2024 (NHIS)
- PDFPaperCreswell 2018
- DOCXThesisCh. 2 draft v.4
- PDFTextbookKvale & Brinkmann (2015)
- PDFPreprintGarcia 2025 (arXiv)
- PDFPreprintOkafor 2025 (SSRN)
- XLSDatasetNHIS 2022 · public-use
- PDFLectureWk 7 — Mixed Methods
- PDFSyllabusSOCI-810 · Spring 2026
- PDFStyleAPA 7th · quick ref
- XLSCodingNVivo export · interviews
- PDFGrantNSF SES-2148321 — proposal
- PDFIRBIRB protocol · Smith lab
- XLSBibliog.Bibliography (.bib)
- PDFConf.NeurIPS 2026 · CFP
- PDFNotesAdvisor mtg · Mar 18
- PDFFig.Fig. 3 — methods diagram
Draft a literature review section comparing the methodologies in my sources (Smith 2023, Liu 2024, Creswell 2018).
2.3 Comparing methodology choices
Recent mixed-methods scholarship has converged on two dominant designs: sequential explanatory and convergent parallel. The relative merits of each depend less on the question itself than on the practical constraints — sample access, funding window, and the analytic tradition of the home discipline.Creswell p.18
Smith (2023) adopts a sequential explanatory design with a mid-scale sample (n=412) drawn from three urban field sites. The strength of this design is depth: the qualitative phase is informed by the quantitative findings, and purposive sampling for interviews maximizes theoretical variation.Smith p.4
Liu (2024) takes a different route, drawing on secondary analysis of the National Health Interview Survey (n=18,302). The gain in statistical power is substantial, but the trade-off is the inability to probe mechanism through respondent self-report.Liu p.12
For the thesis dataset proposed in Chapter 1 (n=520, primary clinical interviews and observation), the Smith design is the closer template. Liu's approach is best invoked where scale claims need defending — not as a methodological model for primary collection.
Draft with your reading list open.
Prompt the HOOH Editor. AI streams paragraphs with inline source badges pulled from your binder — Smith 2023 p.4, Liu 2024 p.12 — directly into your thesis draft. You edit. Exports as DOCX. The four-inputs spine — your sources, the web, the AI, you — embodied here, not just diagrammed.
Deadlines pulled from documents. Remembered for later.
HOOH finds dates, renewals, bills, and follow-ups across your files, then turns them into reminders that stay with your workspace and fire once, not twice.
- 21 · MARPaper draft · SOCI-8103 reading responses pending3 days
- 30 · APRNeurIPS abstractWorkshop track · 6 pp2 weeks
- 18 · MARAdvisor meeting prepBring lit-review drafttomorrow
- 30 · JUNNSF quarterly reportGrant SES-2148321on track
- 12 · NOVThesis defenseCh. 1–5 + appendices6 months
Encrypted. Never used for training.
Your documents, chats, memory, and extracted data are protected with account-scoped encryption. Your uploads and conversations are never used to train models.
- No model training
- Your uploads and conversations are not used to train models.
- Isolated semantic search
- Search signals are transformed per account, so your library is searched only against itself.
Upload. Ask.Edit on the go.
Bring in documents from Files or camera roll, keep them synced, ask questions, and make PDF edits from your phone.
Step in.
No key needed.
No card. No invite. Drop a paper on the desk and talk to it.
Start free