diff --git a/README.md b/README.md
index c3c36ce..c6f7b45 100644
--- a/README.md
+++ b/README.md
@@ -1,3 +1,113 @@
-# research-proposal-literature-review
+# Writing a Research Proposal Literature Review: A Step-by-Step Workflow
 
-Step-by-step workflow for writing a research proposal literature review, with evidence-linked research via Yila (https://yila.ai)
\ No newline at end of file
+A research proposal lives or dies on its literature review. Funding panels, thesis committees and ethics boards read it to answer three questions: does the applicant understand the field, is there a real gap, and is the proposed study the right way to fill it? Many proposals fail not because the idea is weak but because the review is a list of summaries rather than an argument. This article sets out a practical workflow for building a proposal-focused literature review, from scoping to final draft, and shows where an evidence-linked research workbench such as [Yila](https://yila.ai) can save time while keeping every claim traceable to its source.
+
+## How a proposal review differs from a thesis chapter
+
+A thesis literature review can be long and exploratory. A proposal review is short, selective and persuasive. It typically has between one and four pages and must do specific work:
+
+- Establish why the problem matters.
+- Summarise what is already known, with appropriate weight to the strongest evidence.
+- Identify a specific gap, contradiction or limitation.
+- Show how the proposed study addresses that gap and why the chosen methods are appropriate.
+
+Every paragraph should move the reader toward the conclusion that this study needs to be done, and that you are the right person to do it.
+
+## Step one: define the core question and boundaries
+
+Write your proposed research question in one sentence. Then write three to five sub-questions the review needs to answer to justify it. For example, for a proposal on peer tutoring in undergraduate statistics courses:
+
+1. How large are the learning gains reported for peer tutoring in quantitative subjects?
+2. What mechanisms are proposed to explain those gains?
+3. How have previous studies measured outcomes, and what are their limitations?
+4. Which student groups have been under-studied?
+
+Set boundaries early: date range, languages, study designs, populations. Record them so you can justify them later.
+
+## Step two: search systematically, even if the review is not systematic
+
+A proposal review does not usually need a full systematic search, but a documented, structured search makes the review more credible and helps you avoid missing key work.
+
+- Choose two or three databases appropriate to your field.
+- Build search strings from your core concepts and synonyms.
+- Use citation chaining: check the reference lists of key papers and look at who has cited them since.
+- Search for recent systematic reviews first. They summarise large bodies of work and point you to the most important primary studies.
+- Keep a simple log of searches, dates and numbers of results.
+
+AI-assisted discovery tools can widen the net by surfacing papers that use different terminology. Treat their suggestions as leads to verify, not as conclusions.
+
+## Step three: screen and select
+
+Screen titles and abstracts against your boundaries, then read the full text of promising studies. For a proposal, prioritise:
+
+- Systematic reviews and meta-analyses.
+- Large or methodologically strong primary studies.
+- Recent studies that show the current state of the field.
+- Studies that directly motivate your gap, including those with null or conflicting results.
+
+You will usually end up with between twenty and sixty sources, of which perhaps fifteen to thirty are cited.
+
+## Step four: extract evidence into a matrix
+
+Create a spreadsheet or table with one row per study and columns for:
+
+- Citation.
+- Design and sample.
+- Setting and population.
+- Key findings with effect sizes where available.
+- Limitations noted by authors.
+- Your own notes on relevance to each sub-question.
+- The page or section where key findings appear.
+
+The last column is easy to skip and expensive to reconstruct later. When a reviewer asks where a figure came from, or when you need to check whether a finding was about first-year or final-year students, you will want to go straight to the passage. Evidence-linked tools such as Yila can attach source locations automatically as you extract, so each row in your matrix points back to the exact text.
+
+## Step five: synthesise by theme, not by paper
+
+The most common weakness in proposal reviews is summarising one paper per paragraph. Instead, organise by the sub-questions you defined in step one. Within each theme:
+
+- State the main conclusion the evidence supports.
+- Indicate how strong and consistent that evidence is.
+- Note important exceptions, contradictions or methodological concerns.
+- Cite several studies together where they agree.
+
+For example: "Across three meta-analyses and more than forty trials, peer tutoring is associated with moderate gains in quantitative attainment (effect sizes around 0.3 to 0.4). However, most studies measure short-term test scores, and few follow students beyond one semester."
+
+## Step six: make the gap specific
+
+Vague gaps such as "more research is needed" do not persuade panels. A strong gap statement is specific about what is missing and why it matters:
+
+- **Population gap:** evidence comes mainly from selective universities; little is known about open-access institutions where support needs are greater.
+- **Outcome gap:** studies measure immediate test scores but not persistence or later course performance.
+- **Method gap:** most studies are non-randomised and cannot rule out selection effects.
+- **Mechanism gap:** benefits are documented but the mechanisms are untested.
+- **Contradiction:** two strong studies reach opposite conclusions, and the reasons are unclear.
+
+Then connect the gap directly to your aims: "This study will address the outcome and method gaps by randomising tutoring places and following students for two years."
+
+## Step seven: justify your methods with the literature
+
+A proposal review should also support your methodological choices. If you plan to use a particular instrument, cite evidence of its validity. If you choose a mixed-methods design, cite work showing why quantitative results alone have left questions unanswered. If your sample size is modest, cite effect sizes from prior studies to support a power calculation.
+
+## Step eight: draft, check and cut
+
+Write the first draft quickly, following your thematic outline. Then check:
+
+- Every factual claim has a citation.
+- Every citation actually supports the claim it is attached to. Open the source and confirm.
+- Numbers are reported accurately with their context.
+- The argument flows from importance, to current knowledge, to gap, to your study.
+- Nothing is included just to show you have read it.
+
+Proposal reviews almost always need cutting. Remove tangents, merge overlapping paragraphs and shorten descriptions of individual studies.
+
+## Common mistakes to avoid
+
+- Relying on a handful of old or highly cited papers and missing recent developments.
+- Citing reviews for claims they do not make.
+- Ignoring evidence that contradicts your hypothesis. Panels notice, and addressing it strengthens your case.
+- Overstating the gap. If someone has already done something close to your study, acknowledge it and explain how yours differs.
+- Using AI-generated summaries without checking the sources. Fabricated or misattributed citations are a serious credibility risk.
+
+## Summary
+
+A strong proposal literature review is a concise, evidence-based argument for a specific study. Define the question and boundaries, search in a documented way, extract evidence with source locations, synthesise by theme, state a precise gap and link it to your methods. Tools that keep every extracted claim connected to its source, like the research workbench at https://yila.ai, make the checking steps faster and safer, leaving more time for the thinking that makes a proposal convincing.