[{"content":"Data Notes is an example of a research-oriented static publication. It is designed for articles that begin with analysis but should read like conventional essays rather than exported notebooks.\nThe site is intentionally simple:\nArticles are written in Markdown. Charts and illustrations live beside the article that uses them. Code, notebooks, raw data, and build-time analysis are not exposed by the website. Every published page is generated as static HTML and served by GitHub Pages. This initial version contains synthetic sample data solely to demonstrate layout and publishing features.\n","date":"1 August 2026","externalUrl":null,"permalink":"/about/","section":"Data Notes","summary":"Data Notes is an example of a research-oriented static publication. It is designed for articles that begin with analysis but should read like conventional essays rather than exported notebooks.\nThe site is intentionally simple:\nArticles are written in Markdown. Charts and illustrations live beside the article that uses them. Code, notebooks, raw data, and build-time analysis are not exposed by the website. Every published page is generated as static HTML and served by GitHub Pages. This initial version contains synthetic sample data solely to demonstrate layout and publishing features.\n","title":"About","type":"page"},{"content":"","date":"1 August 2026","externalUrl":null,"permalink":"/","section":"Data Notes","summary":"","title":"Data Notes","type":"page"},{"content":"Long analytical articles become easier to read when their structure reflects the reader\u0026rsquo;s questions rather than the order in which the research was performed. This sample page demonstrates a practical sequence and activates the article table of contents.\nStart with the claim # The opening should state the question, the principal finding, and the scope of the evidence. It should not reproduce the research diary. Readers need a map before they encounter methodological detail.\nA useful introduction normally answers three questions:\nWhat is being investigated? What does the evidence broadly indicate? What qualification matters before interpreting the result? Establish the evidence hierarchy # Not every source or output deserves equal weight. The article should make its hierarchy visible through wording, placement, and captions.\nEvidence type Best use Common mistake Primary series Establishing the central trend Treating measurement changes as real change Subgroup results Testing whether an average is representative Showing too many groups without a clear comparison Sensitivity checks Testing robustness Presenting every check as equally important Contextual sources Explaining mechanisms and history Using context as a substitute for measurement Strong presentation does not make weak evidence stronger. It makes the strength and limitations of the evidence easier to judge.\nMove from overview to detail # A reader should encounter the broad pattern before the exceptions. The sequence below is deliberately editorial rather than computational.\nA durable article moves from framing to evidence, interpretation, limitations, and conclusion. Frame the question # Define the population, period, unit of analysis, and comparison. Ambiguity here propagates through every later section.\nShow the central result # Use the smallest number of figures needed to establish the main pattern. A chart earns space when it communicates something that prose or a compact table cannot communicate as effectively.\nTest the interpretation # Separate groups, change assumptions, or use an alternative measure when those checks can overturn the initial reading. Supporting analysis should answer a specific doubt rather than merely demonstrate that more calculations were performed.\nState limitations where they matter # A limitation belongs near the claim it constrains. Collecting every caveat in a final paragraph makes it easy for readers to miss which conclusions are affected.\nDesign for revision # A static article is still malleable. Markdown keeps prose easy to edit, while page bundles keep article-specific figures together. Replacing timeline.svg does not require rewriting a template or changing a public notebook.\nThe simplest durable workflow is:\nprivate analysis → exported figure → Markdown article → static build → GitHub Pages Conclude at the correct level of certainty # The conclusion should return to the initial question and state what the evidence supports—not merely repeat the most visually prominent chart. It should distinguish measured results from plausible explanations and identify the uncertainty most likely to change the interpretation.\nThat discipline is more important than any particular site generator. The framework exists to keep the publishing process simple enough that attention remains on the argument.\n","date":"30 July 2026","externalUrl":null,"permalink":"/posts/long-form-example/","section":"Articles","summary":"A reusable editorial structure for moving from a question to evidence, interpretation, limitations, and a durable conclusion.","title":"A Practical Structure for Long-Form Analytical Articles","type":"posts"},{"content":"","date":"30 July 2026","externalUrl":null,"permalink":"/posts/","section":"Articles","summary":"","title":"Articles","type":"posts"},{"content":"","date":"30 July 2026","externalUrl":null,"permalink":"/categories/","section":"Categories","summary":"","title":"Categories","type":"categories"},{"content":"","date":"30 July 2026","externalUrl":null,"permalink":"/tags/long-form/","section":"Tags","summary":"","title":"Long Form","type":"tags"},{"content":"","date":"30 July 2026","externalUrl":null,"permalink":"/categories/methods/","section":"Categories","summary":"","title":"Methods","type":"categories"},{"content":"","date":"30 July 2026","externalUrl":null,"permalink":"/tags/","section":"Tags","summary":"","title":"Tags","type":"tags"},{"content":"","date":"30 July 2026","externalUrl":null,"permalink":"/tags/workflow/","section":"Tags","summary":"","title":"Workflow","type":"tags"},{"content":"","date":"30 July 2026","externalUrl":null,"permalink":"/tags/writing/","section":"Tags","summary":"","title":"Writing","type":"tags"},{"content":"","date":"25 July 2026","externalUrl":null,"permalink":"/categories/analysis/","section":"Categories","summary":"","title":"Analysis","type":"categories"},{"content":"","date":"25 July 2026","externalUrl":null,"permalink":"/tags/data-visualization/","section":"Tags","summary":"","title":"Data Visualization","type":"tags"},{"content":"","date":"25 July 2026","externalUrl":null,"permalink":"/tags/interpretation/","section":"Tags","summary":"","title":"Interpretation","type":"tags"},{"content":"A rising average can summarize a dataset accurately while still concealing the change that matters most. This synthetic example demonstrates how an article can move from a compact overview to a wider comparison and then to two supporting views.\nDemonstration only: every value in this article is synthetic and exists solely to test the publication layout. The aggregate view # The headline series rises from an index value of 100 to 148 over eight periods. The direction is clear, but the line alone does not show whether every group moved together.\nThe change accelerates in the middle of the series and then moderates. That pattern is visible at ordinary article width, where the chart remains close to the paragraph that interprets it.\nThe group comparison # A wider layout is more useful once the aggregate is separated into four groups.\nThe aggregate increase is driven disproportionately by Groups A and B; Groups C and D change much less. The wider figure makes two conclusions easier to see. First, the final gap between groups is much larger than the initial gap. Second, the aggregate line is not representative of the trajectory experienced by every group.\nLevel and change answer different questions # The same result can be presented as final levels or as cumulative change from the baseline. Placing the two views beside each other makes the distinction explicit.\nFinal level on the left; cumulative change from the common baseline on the right. Group A finishes at the highest level, while Group B records nearly as much growth from a slightly lower starting point. Groups C and D appear similar in the final-level chart, but their growth rates are not identical.\nWhat the layout contributes # The article does not expose a notebook or follow its cell order. Each figure is an independent SVG asset, so the narrative determines placement:\nthe overview remains close to the introduction; the dense comparison uses more horizontal space; the final two views share a common caption and collapse vertically on mobile. This is the intended pattern for future analytical articles: compute privately, export deliberately, and write the public piece as an ordinary publication.\n","date":"25 July 2026","externalUrl":null,"permalink":"/posts/illustrated-analysis/","section":"Articles","summary":"A worked visual example showing why an aggregate trend and its underlying distribution should be interpreted together.","title":"Reading a Trend Without Losing the Distribution","type":"posts"},{"content":"","date":"25 July 2026","externalUrl":null,"permalink":"/tags/sample-data/","section":"Tags","summary":"","title":"Sample Data","type":"tags"},{"content":"","date":"20 July 2026","externalUrl":null,"permalink":"/tags/method/","section":"Tags","summary":"","title":"Method","type":"tags"},{"content":"","date":"20 July 2026","externalUrl":null,"permalink":"/tags/publishing/","section":"Tags","summary":"","title":"Publishing","type":"tags"},{"content":"","date":"20 July 2026","externalUrl":null,"permalink":"/categories/site-notes/","section":"Categories","summary":"","title":"Site Notes","type":"categories"},{"content":"This site is built around a strict separation between analysis and publication. Calculations may happen in a notebook or another private environment, but the public result is a normal article with deliberately selected charts, tables, and illustrations.\nWhy separate the two? # A notebook is excellent for exploring data. It is less suitable as the permanent editorial structure of an article. Once figures are exported as stable assets, the prose can be reorganized without moving computational cells or exposing implementation details.\nThat separation provides three practical benefits:\nEditorial freedom. A figure can appear wherever the argument needs it. A cleaner public page. Readers see prose and evidence, not notebook controls or hidden code containers. Simpler hosting. GitHub Pages serves only static HTML, CSS, JavaScript, and images. What is included # The initial site includes responsive article pages, search, tags, categories, RSS, light and dark appearances, a table of contents for long articles, and reusable layouts for wide or paired figures.\nThe goal is not to make a notebook resemble a blog. The goal is to publish a blog whose figures happen to originate from reproducible analysis.\nThe next sample article demonstrates those figure layouts with synthetic data.\n","date":"20 July 2026","externalUrl":null,"permalink":"/posts/welcome/","section":"Articles","summary":"A compact introduction to a static publication built for normal prose, careful analysis, and selected illustrations.","title":"Welcome to Data Notes","type":"posts"},{"content":"","externalUrl":null,"permalink":"/series/","section":"Series","summary":"","title":"Series","type":"series"}]