Module Guider 查核清單Module Guider Verification Checklist

本站有兩份查核清單，這是其中第一份：涵蓋 verify/workflow.qmd 流程裡的 P0–P1、V1–V3、V5、C1–C2 步驟，設計成可列印的一頁。適用對象是使用 Module Guider（askllm）取得推薦分析與選單路徑的情境；若你用的是 R code tutor（askllmr），請改用 checklist-rtutor.qmd。

This is the first of the two verification checklists on this site. It covers steps P0–P1, V1–V3, V5, and C1–C2 of the workflow in verify/workflow.qmd, laid out as a single printable page. Use it when you asked the Module Guider (askllm) for a recommended analysis and menu path. If you used the R code tutor (askllmr) instead, use checklist-rtutor.qmd.

送出前Before you submit

P0 我已經寫下「我會接受什麼答案」（例如：回覆必須指名一個 T-Tests 底下的分析，並說明為何不是 ANOVA）。
P1 我已經記下摘要裡的關鍵數字（每一組的 n、mean、sd、遺漏數），這些數字之後要拿來跟回覆核對。

P0 I have written down what answer I will accept (e.g. “the reply must name a specific analysis under T-Tests and explain why it is not an ANOVA”).
P1 I have written down the key numbers from the summary (n, mean, sd, missing count per group) so I can check the reply against them later.

收到回覆後After you receive a reply

V1 路徑核對 — 回覆裡每一條 Analyses ▸ … 選單路徑，我都在自己的 jamovi 裡逐字點得到（模組名稱、分析名稱、子選項都要對得上，不是「差不多」）。
V2 數字核對 — 回覆裡引用的 n、mean、sd 等數字，跟我在 P1 記下的、以及在 Descriptives 面板看到的一致。
V3 範圍核對 — 回覆有沒有宣稱它「看不到」的東西（例如常態性、離群值、原始資料列）？如果有，這是「越界」，代表回覆超出摘要統計能支持的範圍，應打折扣看待。
V5 交叉核對 — 換一個 persona（consultant / tutor / explainer）或換一個模型，針對同一個 Question: 再問一次；兩次的結論是否一致？（路徑仍須各自逐條核對，不能因為兩次都提到同一條路徑就省略 V1。）

V1 Path check — Every Analyses ▸ … menu path in the reply can be clicked, verbatim, in my own jamovi installation. Module name, analysis name, and sub-option must match exactly, and “close enough” does not count.
V2 Number check — The n, mean, sd, etc. cited in the reply match what I wrote down in P1 and what I see in the Descriptives panel.
V3 Scope check — Does the reply make claims about things it cannot actually see (e.g. normality, outliers, raw data rows)? If so, flag this as “overreaching.” The claim exceeds what summary statistics can support, and you should discount it.
V5 Cross-check — Ask the same Question: again with a different persona (consultant/tutor/explainer) or a different model. Do the conclusions agree? (Still repeat V1 for the second reply. Agreement between two replies does not substitute for checking each path individually.)

結案Closing out

C1 對照 P0 寫下的判準：solved（全部命中，依建議執行）還是 reopen （任一項沒通過，需要換個問法或換個 persona 再試）。
C2 填寫協作日誌四欄：問題 / 判準 / 回覆摘要 / 查核結果。

C1 Compare the reply against the P0 criteria, and reach one of two outcomes. Solved means everything checks out, so proceed with the recommendation. Reopen means something failed, so rephrase the question or try a different persona.
C2 Fill in the four-column collaboration log: question / criteria / reply summary / verification result.

紅旗（任一項出現就先標「越界」，不要照做）Red flags (flag as overreaching, do not act on it yet)

回覆聲稱「我檢查過資料的常態性／離群值」——摘要統計不含這些資訊，這不可能是真的檢查，最多是猜測。
回覆引用的選單路徑在你的 jamovi 版本或已安裝模組裡完全找不到對應項目。
回覆引用的 n 或其他數字跟你在 Descriptives 看到的對不起來。

The reply claims to have “checked normality/outliers in the data.” The summary it received cannot support this claim. At best it is a guess.
A cited menu path has no matching item anywhere in your jamovi version or installed modules.
A cited n or other number does not match what you see in Descriptives.

送出前Before you submit

收到回覆後After you receive a reply

結案Closing out

紅旗（任一項出現就先標「越界」，不要照做）Red flags (flag as overreaching, do not act on it yet)

相關頁面Related pages