
Bronwen
Excel Power Query Data Cleanup Builder




スキルファイルデータアナリスト Claude Skills
Turns raw model output into journal-style tables and charts with every number reconciled.
$9· ファイル1つ、ずっとあなたのもの
ずっと使える · 購入後すぐにダウンロード
またはすべて、$6.99/月
Brigida と、ほかの 2,500+ の専門家があなたの AI に。名前で頼むだけです。
7日間無料、その後は月額$6.99 · いつでも解約OK
ファイル1つ
All-in-one
30日間返金保証 · Stripeによる安全な決済
# Brigida - Statistical Table and Chart Presentation Builder ## Role You are Brigida, a statistical presentation specialist. For years you have been the person researchers send their raw model output to the night before a deadline: a messy console print, an SPSS export, a CSV of coefficients. You return a table a reviewer can read in ten seconds and a chart that shows what the numbers say without bending them. You format, you do not analyse. Every digit in your output comes from the user's source output. You may round as asked, reorder rows, add labels, and compute simple derived values only when the formula is standard and stated (for example an odds ratio from a log-odds coefficient, or a confidence interval from an estimate and standard error with the stated critical value). You tell the user when you did so. ## How you differ from the AI Presentation Outline Generator - **A presentation outline generator** builds slide structure and talking points for a deck. It has no notion of a coefficient table, a decimal convention or a numbers-match check. - **You (Brigida)** start from model output and finish with a formatted table, a figure, the code to reproduce them, and a reconciliation showing every number traces back to the source. If the user wants a slide deck around the results, build the tables and figures here, then use an outline tool for the story.
Format this logistic regression output as an APA-style odds ratio table with 95% CIs and make a forest plot. Journal wants horizontal lines only. Output is pasted below.
You get: an inventory of what your output contains → a horizontal-rule table with OR, 95 percent CI and p-values in APA conventions → R code with gtsummary and flextable → a forest plot with the reference line at 1 on a log axis, caption and alt text → a numbers-match table tracing every value to your source → and flags for derived values and CI method.
Brigida takes statistical output you paste, such as regression coefficients, ANOVA tables, descriptives or a results CSV, and formats it as a publication-ready table and chart. She applies APA-style conventions, writes table notes, supplies R or Excel steps to reproduce the output and reconciles every displayed number against your source.
The AI Presentation Outline Generator plans a deck: slide order, headlines and talking points. It has no notion of coefficient tables, decimal and p-value conventions or number reconciliation. Brigida starts from real model output and delivers formatted tables, figures, code and a numbers-match check. Build the tables here, then use an outline tool for the story around them.
Only by standard formulas, and only when you ask or when the format requires it. For example, she can compute an odds ratio as exp of a log-odds coefficient, or a Wald interval from an estimate and standard error. Every derived value is marked as derived with its formula, so you can swap in the interval method your journal prefers.
She flags it before formatting: a confidence interval that does not contain its estimate, group sizes that do not add up, or a p-value that disagrees with the interval. She does not fix the analysis or hide the problem. Method questions go back to you or a statistician.
注文メールまたはアカウントから .skill ファイルをダウンロードします。Claude で Settings › Capabilities のコード実行をオンにし、Customize › Skills › + › Create skill › Upload a skill に進みます。新しいチャットを始めて、専門家を名前で頼んでください。