本文由 UniWriterPro 學術規例研究組 verify · Last updated: 2026-08-30 · 本文係按 APA Style Number and Statistics Guide (7th ed.)、APA Style Table setup 同 Figure setup、PolyU ELC FYP Tables and Figures 同 Appendices、PolyU EWrite Project Report Structure、HKUST CSE FYP Final Report、HKU DASE FYP Handbook 2024-25、CityU Library FYP Guidelines、HKU ITS 同 CUHK ITSC 同 HKBU ITO 同 EdUHK OCIO 嘅 SPSS 授權頁、CUHK Library Research Data Management 逐條解讀。冇公開文件嘅位一律標「搵唔到官方文件」，唔搬其他學校套用。
SPSS 結果論文點寫，最常見嘅做法係最錯嘅一種：喺 output 視窗 select all，copy，paste 落 Word，然後逐張表下面寫一句「From Table 4.1, we can see that…」。呢個做法會同時撞爛三樣嘢——章節分工、格式一致性同字數。本文由 output 視窗開始，逐步拆到一份 examiner 讀得明嘅 Chapter 4。
如果覺得自己未必做得好，想買個安心，可以交畀我哋做 —— 我哋提供人手代做服務 FYP 代做、thesis 代做。我哋會先檢查你跑出的分析係咪真正回答 Research Question，再逐個數字對返原始 output；如果問題出喺 Methodology，唔會淨係幫你將錯誤結果排得更加靚。
目錄
- 第一步：SPSS output 分三堆，唔係全部都入正文
- 第二步：表同文唔可以出現同一組數字
- 第三步：每個數字嘅寫法 — 小數位、p 值、斜體
- 第四步：顯著性標示只可以揀一種
- 第五步：表同圖嘅擺位規則
- 四類最常見 SPSS output 嘅搬運模板
- NVivo 質性 output 點寫成文字
- 原始 output 去邊 — 附錄嘅分界線
- 你部私人電腦裝唔裝得 SPSS — 各校條文唔同
- 交完之後數據要唔要留
- 官方文件搵唔到嘅位
- FAQ — 常見問題
第一步：SPSS output 分三堆，唔係全部都入正文
SPSS 跑一次 regression 可以吐八張表。你唔會全部用。所以講 SPSS 結果論文點寫之前，第一件事係做分堆，而唔是排版。
分堆之後仲有一層。一份守得住嘅 Chapter 4，每個數字都答得出三件事：由邊張 output 表嚟、點解揀呢個 test、為咗咩唔放另外幾欄。排版錯可以改，呢三條答唔出，口試就會卡住。
官方文件對「Results 章節放咩」講得最直接嘅係理大 EWrite 嘅 project report 指引：
Present the results in either tabular or graphic format, without discussing them; Indicate meaningful aspects of the data in appropriate order.
PolyU EWrite — How to structure your project report
留意「meaningful aspects」同「appropriate order」兩個字眼。呢句唔係叫你把 output 倒出來，係叫你揀同排。同一份指引講到 visuals 嘅時候用嘅動詞係 synthesize：
Use visuals (i.e. box plots/graphs/charts/tables/diagrams) or sets of formulas to synthesize and present key findings and results, and add captions or explanations to describe them.
PolyU EWrite — How to structure your project report
科大 CSE 嘅 FYP final report 指引就講埋反面：
For boring details, put them in appendices or just leave them out.
HKUST CSE — FYP Final Report
「Just leave them out」呢四個字好多人唔敢做，因為覺得跑咗就一定要放。實際上放晒反而扣分：examiner 讀唔出你嘅重點喺邊。
操作上分三堆。打開你嘅 output，逐張表問一次：呢張表係直接答緊我其中一條 research question 嗎？答係，入正文。答唔係但 examiner 可能要查（例如 reliability、assumption 檢查、完整問卷 frequency），入附錄。答唔係而且冇人會查（你試錯過程跑咗但唔用嘅 model），唔用。
下表係我哋喺實際稿件裡面最常用嘅分堆起點。呢張表係 UniWriterPro 嘅編輯判斷，唔係校規——各系要求唔同，最終以你 supervisor 同學系指引為準。
| SPSS output 常見表名 | 建議去處 | 理由 |
|---|---|---|
| Descriptive Statistics / Frequencies（核心變項） | 正文 | 讀者要知你手上係咩樣本先睇得明後面 |
| Frequencies（每一條問卷題逐條） | 附錄 | 量太大，會撐爆 Results 章節 |
| Reliability Statistics（Cronbach’s Alpha） | 正文一句 + 附錄完整表 | 正文交代量表可用，細節留附錄 |
| Independent Samples Test / ANOVA（答 RQ 嘅） | 正文 | 直接答 research question |
| Levene’s Test、常態性檢定 | 附錄，正文一句交代 | 係前提檢查，唔係發現 |
| Correlations（核心變項矩陣） | 正文 | 通常係 RQ 主體 |
| Model Summary / ANOVA / Coefficients（regression 三張） | 正文，但通常合併重排成一張 | 三張原生表放晒係最常見嘅版面災難 |
| 試錯階段跑咗但最後唔用嘅 model | 唔用 | 放咗會令讀者以為你冇決定 |
呢一步睇落簡單，實際上係整個 Chapter 4 最花時間嘅判斷，因為佢要求你已經知自己邊條 RQ 靠邊張表答。如果你手上 output 一大疊但唔知邊張對應邊條問題，通常唔係寫作問題而係設計問題。我哋做FYP 分階段輔導嘅時候，呢一步一定係同你逐張表對住 RQ 過一次，唔會直接跳去排版。
第二步：表同文唔可以出現同一組數字
呢條係本文最重要嘅一條，而且係最多人違反嘅一條。APA 官方嘅 Number and Statistics Guide 寫得冇任何轉彎位：
Do not repeat statistics in both the text and a table or figure.
APA Style — Number and Statistics Guide, 7th Edition
意思係：如果你已經放咗一張 Descriptives 表，表裡面有 mean 同 standard deviation，你就唔應該喺正文再寫一次「平均分係 X，標準差係 X」。反之，如果你打算喺正文逐個數字讀出嚟，你就唔需要嗰張表。
理大 ELC 由另一個角度講同一件事——表同圖本身嘅功能就係代替重複句子：
Condense detailed information and thus avoid the necessity for complex and repetitive sentences.
PolyU ELC — presentation of tables and figures
所以正文嘅工作唔係讀表，係指出表裡面邊個位重要。理大 ELC 用嘅動詞係 integrate：
Integrate the table or figure into your text by referring to particularly significant results.
PolyU ELC — presentation of tables and figures
一句可以用嘅句型：「Table 4.2 顯示三組之間嘅差異，其中 A 組同 C 組嘅差距最大」——句子指出方向同重點，但唔重複表裡面任何一個數字。相反嘅寫法「Table 4.2 顯示 A 組平均 X、B 組平均 X、C 組平均 X」就係把表讀一次，兩邊都浪費。
要留意呢條規則同 Discussion 章節嘅重複問題唔同層。呢度講嘅係同一章之內「表同文」重複；Discussion 嘅問題係「章同章」重複，即係 Chapter 5 把 Chapter 4 講過嘅結果再讀一次。兩個問題成因唔同，處理方法亦唔同，我哋喺FYP Discussion 只係 repeat results 嘅五個改寫做法拆咗後者。兩篇一齊睇，Chapter 4 同 Chapter 5 嘅分工就清楚。
第三步：每個數字嘅寫法 — 小數位、p 值、斜體
SPSS output 默認出三個小數位，而且成張表都係同一個格式。你唔可以照搬，因為唔同統計量嘅小數位要求唔同。APA 官方指引逐類列出：
Report means and standard deviations for data measured on integer scales (e.g., surveys and questionnaires) to one decimal. Report other means and standard deviations and correlations, proportions, and inferential statistics (t, F, chi-square) to two decimals.
APA Style — Number and Statistics Guide, 7th Edition
即係：問卷（例如 5 點量表）嘅平均分同標準差寫一個小數位；t、F、卡方、相關係數寫兩個小數位。SPSS 給你嘅三個小數位要自己收。
P 值另有寫法：
Report exact p values to two or three decimals (e.g., p = .006, p = .03). However, report p values less than .001 as “p < .001.”
APA Style — Number and Statistics Guide, 7th Edition
呢條解決咗一個好常見嘅困惑：SPSS 喺 Sig. 欄出「.000」，好多人照抄「p = .000」。p 唔會等於零，SPSS 只係顯示位數不足。照 APA 寫法應該寫 p < .001。
小數點前面嘅零亦有規則，而且係兩條相反嘅規則：
Put a zero before the decimal point when a number is less than 1 but the statistic can exceed 1. Do not use a zero before a decimal when the statistic cannot be greater than 1 (proportion, correlation, level of statistical significance).
APA Style — Number and Statistics Guide, 7th Edition
拆開講：相關係數、比例、p 值理論上唔會超過 1，所以唔寫前導零，寫 r = .42、p = .03。t 值、F 值可以超過 1，所以就算今次個數細過 1 都要寫前導零，寫 t = 0.87。SPSS output 一律有前導零，所以呢一項要逐個手改。
符號要斜體：
Use italics for letters used as statistical symbols or algebraic variables (e.g., contained 587 t-test p values; R2 = .12). However, use standard (nonitalic) type for Greek letters.
APA Style — Number and Statistics Guide, 7th Edition
所以 M、SD、t、F、p、r、N、R² 全部斜體；希臘字母（例如 α、χ）唔斜體。Copy-paste 由 SPSS 出嚟嘅嘢一律唔會有斜體，呢個係逐個符號要人手處理嘅位。
APA 亦講明呢啲係一般原則而唔係死規矩：
Keep in mind that these are general guidelines and that the most important consideration when deciding the number of decimal places to use in reporting results is the following: Round as much as possible while considering prospective use and statistical precision.
APA Style — Number and Statistics Guide, 7th Edition
一個重要交代：APA 唔係香港院校校規。APA 係心理學同社會科學最常用嘅格式權威，但工程、商科、護理各有慣用格式。城大圖書館嘅 FYP 指引就直接把決定權交返學系：
Students should follow the guidelines and manuals as specified by the student’s own Faculty or Department when preparing their printed project reports.
CityU Library — Guidelines for Undergraduate Final Year Projects
理大 EWrite 亦有同一句提醒，講明結構同要求要以任課老師為準：
Note that your own subject teacher may require you to include some or all the sections and moves, according to different assessment purposes, subjects and disciplines.
PolyU EWrite — How to structure your project report
所以正確做法係：先問你學系用邊套格式，再用嗰套；如果學系冇明講而你係社科或商科，APA 係最穩嘅預設。
格式一致性呢一層唔靠靈感，靠逐個符號逐個小數位對一次。全份稿三十幾個統計數字，漏一個斜體同漏一個前導零都係同一類扣分。我哋嘅論文校對服務就係逐項對，包括 SPSS 搬過嚟之後最常漏嘅斜體同 p = .000。
第四步：顯著性標示只可以揀一種
好多稿件會同時出現兩種標法：表裡面既有 Sig. 欄嘅實際數值，又有星號。APA 官方 blog 有一篇專講呢件事，標題本身就係答案——Pick one：
The good news is that you have two options, and you need to choose only one.
APA Style — Methods of reporting significance in tables and figures: Pick one!
第一種係喺表裡面開一個 p 欄，寫實際數值：
Reporting exact p values in a column or row of a table or within a figure is thus Option 1.
APA Style — Methods of reporting significance in tables and figures: Pick one!
第二種係用星號或劍號分段，然後喺表底下嘅 probability note 定義每個符號代表咩：
The p < notation involves using asterisks or daggers in the table or figure to identify statistics with significant p values that fall within certain ranges—traditionally, p < .05, p < .01, and/or p < .001—and then defining what each asterisk or dagger pattern means in the probability note that appears beneath the general and specific notes of the table or figure.
APA Style — Methods of reporting significance in tables and figures: Pick one!
喺表同圖裡面嘅 p 值寫法同正文一樣：
In tables and figures, report exact p values (e.g., p = .015), unless p is < .001 (instead write as “<.001”).
APA Style — Number and Statistics Guide, 7th Edition
另外一個常見誤會係用粗體標「顯著」。APA 講明粗體係選項而唔係要求，而且一旦用咗就要喺 general note 交代佢代表咩：
Please note that bold is a formatting option in situations where you want to emphasize significant reported results: Using bold is not a requirement to highlight significant p values themselves or results with significant p values.
APA Style — Methods of reporting significance in tables and figures: Pick one!
操作建議：全份稿統一揀一種，而且喺第一張統計表定咗之後唔再變。表數少（三張以內）用實際 p 值最清楚；表多而且每張有大量係數（例如多條 regression 並排比較）用星號會省好多位。
第五步：表同圖嘅擺位規則
SPSS output 貼落 Word 之後，表就係一堆浮吓浮吓嘅嘢。要處理嘅有四樣：編號、標題位置、方向、同正文嘅相對位置。
編號要連續而且用阿拉伯數字：
Number the tables and figures consecutively throughout the report by using Arabic numbers. Avoid using A,B,C or Roman numbers i, ii, iii as labels because such usage is quite clumsy.
PolyU ELC — presentation of tables and figures
標題位置——呢個係理大 ELC 同 APA 之間有分歧嘅位，所以要睇你跟邊套。理大 ELC 寫：
Give each table and figure a title so as to immediately tell your reader what it is showing. The table number and title should be placed ABOVE the table. The figure number and title should be placed BELOW the figure.
PolyU ELC — presentation of tables and figures
APA 第七版就把表同圖統一，兩者都係編號同標題喺上面：
The figure number (e.g., Figure 1) appears above the figure title and image in bold font. The figure title appears one double-spaced line below the figure number. Note that tables and figures have the same overall setup.
APA Style — Figure setup
兩套都係正式指引，唔存在邊套「錯」。要做嘅唔係揀一套你覺得靚，而係揀你學系用嘅一套，然後全份稿一致。如果你學系冇明講而你用 APA 格式引文，就跟 APA。
編號順序要跟正文提及次序，唔係跟你製作次序：
Number tables in the order in which they are mentioned in your paper.
APA Style — Table setup
而且正文要先提後放，唔可以表突然出現：
Refer to the table or figure BEFORE it appears.
PolyU ELC — presentation of tables and figures
方向要 portrait，唔好橫置：
Present the table or figure in the normal ‘portrait’ orientation, rather than ‘landscape’: it can be quite annoying to have to turn the report sideways in order to read the data.
PolyU ELC — presentation of tables and figures
呢條對 SPSS 用戶特別重要，因為 regression 嘅 Coefficients 表原生闊度好多時放唔落 portrait。正確做法係刪走唔需要嘅欄（例如 Std. Error、Beta 之中揀你要嗰個）而唔係轉橫。理大 ELC 亦講到要收簡：
Keep your table or figure relatively simple. Keep such additions as lines, words, labels and colours to a minimum.
PolyU ELC — presentation of tables and figures
表底下嘅 note 係用嚟講表本身講唔到嘅嘢：
Three types of notes (general, specific, and probability) appear below the table as needed to describe contents of the table that cannot be understood from the table title or body alone (e.g., definitions of abbreviations, copyright attribution, explanations of asterisks used to indicate p values).
APA Style — Table setup
所以你所有縮寫（SPSS 表頭常見嘅 Sig.、Std. Deviation、df）都應該喺 note 裡面定義一次，而唔係當讀者一定識。
四類最常見 SPSS output 嘅搬運模板
以下係四類最常見嘅 test 由 output 到正文一句嘅結構。數字位一律用 X 佔位，唔係真實數據——你要填返自己 output 嘅數，並按上面第三步嘅小數位規則處理。搵唔到官方文件講明每個 test 要點寫成句，所以呢部分嘅句型係按 APA 統計符號規則同一般學術慣例組成，唔係任何一間院校嘅明文要求。
一、Descriptives（描述統計）
SPSS 表：Descriptive Statistics 或 Frequencies。正文要交代嘅係樣本輪廓，唔係每一格數字。句型：先講樣本量同回收情況，再講核心變項嘅集中同離散趨勢，然後指出一個值得注意嘅偏態或極端值。如果表已經放咗 M 同 SD，正文就唔再讀出（見第二步）。
二、Independent Samples t-test（兩組比較）
SPSS 出兩張表：Group Statistics 同 Independent Samples Test。後者裡面有 Levene’s Test（變異數同質性）同兩行 t 檢定結果。三個常見錯誤：一，唔講自己讀咗 Levene’s 之後揀邊一行；二，把 Levene’s 表整張放正文；三，Sig. 欄照抄 .000。正確做法係正文一句交代 Levene’s 結果同因此揀咗邊一行，Levene’s 完整表放附錄，t 檢定寫成 t(df) = X, p = X 嘅形式。
三、One-way ANOVA（三組以上比較）
SPSS 出 ANOVA 表加 Post Hoc 表。ANOVA 表只答「有冇差異」，答唔到「邊兩組有差異」，所以顯著之後一定要交代 post hoc 結果。最常見嘅失誤係只放 ANOVA 表就當講完，或者反過來把 Post Hoc 整張（可能幾十行配對）放正文。做法係 ANOVA 主結果寫成 F(df1, df2) = X, p = X，post hoc 只喺正文點出有顯著差異嘅幾對，完整配對表放附錄。
四、Correlation 同 Regression
Correlations 表如果變項多會變成一個大矩陣，通常保留下三角、刪走重複嘅上三角。Regression 出三張表——Model Summary、ANOVA、Coefficients——原生三張並排放係最常見嘅版面災難。做法係重排成一張：保留每個預測變項嘅係數同顯著性，再喺表底 note 或正文一句交代模型整體嘅解釋力同顯著性。另外要留意 SPSS 嘅 Beta 欄係標準化係數，同 B 欄唔同意思，兩個都放但唔交代邊個係邊個係常見扣分位。
如果你發現自己揀嘅 test 同 research question 對唔上（例如問「有冇關係」但跑咗 t-test），問題唔喺呢一章而喺方法設計。呢種情況重排表格係救唔到嘅，要返上游改。我哋喺FYP 研究方法選擇拆咗點由 RQ 反推 test，喺FYP Proposal 四段寫作流程拆咗方法段點寫到 supervisor 肯簽。寫完 Chapter 4 之後點樣唔喺 Chapter 5 重複一次，就睇Discussion 唔重複 results 嘅改寫做法。
科大 CSE 對 Results 同 Discussion 嘅分工有一句好直白嘅描述，值得抄落自己張紙上：
In your evaluation you summarize the results from testing and state how well you achieved your objectives. This is very objective (客觀的). In your discussion, you interpret the results based on your opinions, so it is more subjective (主觀的).
HKUST CSE — FYP Final Report
官方文件自己用中文寫「客觀的」同「主觀的」，界線就係呢個：Chapter 4 只講發生咗咩，Chapter 5 才講你認為佢代表咩。
NVivo 質性 output 點寫成文字
NVivo 嘅情況同 SPSS 唔同。SPSS 給你數字，你要決定放邊啲；NVivo 連 AI 功能一齊用，係可以由 code、node 同 quotes 生成一整段流暢文字嘅——問題唔係生成唔到，而係佢未必分得清高頻 code 同真正有解釋力嘅 theme，亦未必幫你保留反例、語境同完整分析軌跡。中間嗰段判斷要人補。
搵唔到官方文件講明香港院校要求 NVivo 結果點呈現。港大圖書館有 NVivo 嘅 LibGuide，浸大圖書館亦有研究支援資源，但兩邊都係軟件操作教學，冇規定寫法。所以以下係我哋嘅編輯經驗，唔係校規。
三個最常見問題：
- 用 code 頻次代替分析。寫「呢個 theme 出現咗 X 次」係量化包裝，讀者要知嘅係 theme 之間嘅關係。頻次可以放，但唔可以係唯一內容。
- Node 樹直接搬做小標題。NVivo 嘅 node 層級係你 coding 時嘅工作結構，唔一定係讀者最易讀嘅敘事順序。搬之前要重新排一次。
- 引語冇上下文。放一句受訪者原話但冇交代佢係邊個身分、喺咩情境講，引語就冇證據力。每句引語要有 participant 代號同必要背景。
呢三條唔只係我哋嘅偏好，方法學研究測到同一個方向。2026 年一項研究要人類編碼員逐條評 GPT-4o 生成嘅 inductive code：
In the inductive condition, GPT-4o generated 137 initial codes, of which 31.4% (n=43) demonstrated agreement with human codes, 26.3% (n=36) represented reasonable alternatives, and 42.3% (n=58) were classified as not reasonable.
JMIR Formative Research (2026) — LLM-assisted qualitative coding evaluation
要公道講，同一篇研究對 deductive（跟一份已有 codebook 去 code）嘅評價係正面嘅，作者結論係「With targeted human oversight, an LLM reliably applied an existing codebook and generated additional inductive codes.」——重點落喺 human oversight 呢四個字。另一項比較人手 coding 同軟件 autocoding 嘅研究，記錄到覆蓋率上嘅具體差距：
In one instance, exploring data files, the AI program generated 10 codes including 4 repeats versus 47 researcher defined codes from manual human coding.
International Journal of Qualitative Methods — Artificial Intelligence and Qualitative Data Analysis
同一篇研究嘅結論係「natural language AI platforms provide a novel space for data experimentation, but, for now, not a replacement for the interpretive skills of human data analysts.」實務上嘅意思係：如果你用軟件內建 autocoding 或者 AI assistant 跑完就交，最常出現嘅問題唔係 theme 太多，而係 theme 太扁——code 數目偏少、有重複、大部分停留喺 topic 層面而未到 theme 層面。呢個時候要做嘅係重排同補證據，唔係再 code 一次。
上面第三條——引語——有佢自己嘅量化證據。另一項研究把 GPT-4o 抽出嘅引語，同人類編碼員親手揀嘅 162 段引語做比對：
The average precision score of 0.41 suggests that around 41% of excerpts identified by the LLM as relevant actually were (true positive).
On the other hand, the average recall of 0.53 suggests that just over half of all excerpts that the human coder had identified were also found by the LLM.
AI & Society (2026) 41:731–748 — LLM-assisted thematic analysis
把兩個數字譯成你校稿時睇得見嘅東西：十句自動抽出嘅引語，大約六句其實支持唔到你聲稱嘅那個 theme；而你自己讀過訪談稿會揀嘅引語，佢有一半係漏咗。呢兩條錯嘅方向唔同：前者係 examiner 睇一眼就捉得到嘅「呢句引語證唔到你諗嘅那句話」，後者係你自己睇唔到嘅——最有力嘅那句原話未入到文。該研究作者自己嘅結論係 AI「is no ‘silver bullet’」，同時亦指出真正嘅取捨係節省嘅 coding 時間同人手驗證所需時間之間嘅 trade-off。實務上意思很直：抽得快，但每句引語你都要回到原稿確認一次。
一個可用嘅段落結構：先一句定義 theme，再一句交代佢喺幾多個 participant 出現同分佈情況，再一至兩句引語做證據，最後一句講呢個 theme 同你 research question 嘅關係。四層齊咗，段落就唔會讀落似 coding 報告。
要注意嘅係，Chapter 4 呢一層仍然唔可以做解釋——照科大 CSE 嘅界線，解釋屬 Discussion。質性研究呢條界線比量化模糊，理大 EWrite 亦承認可以合併：
You may combine this section with the Discussions section to allow your explanation to flow better or to develop a narrative to account for the relationship between results and approaches.
PolyU EWrite — How to structure your project report
但同一頁亦提醒分開寫通常更清楚：
Sometimes, making a clear distinction between the three sections; Results, Discussion and Conclusion, can increase the overall clarity of the project report.
PolyU EWrite — How to structure your project report
實務建議：合併定分開，問你 supervisor，唔好自己決定。合併之後改返分開係大改，反過來亦一樣。
質性資料由 coding 到成章文字呢一段，係我哋論文章節寫作服務接得最多嘅一類：coding 你自己做過，但要把 node 結構重排成一條讀得順嘅敘事線，同時保住每句引語嘅證據力，係另一種功夫。
原始 output 去邊 — 附錄嘅分界線
上面講咗好多次「放附錄」。附錄唔係垃圾桶，佢有明文規則。理大 ELC 專頁講得最齊。
定義：
An appendix refers to the inclusion of additional detailed data that is relevant to the content but is too detailed to include in the main body of the Final Year Project (FYP).
PolyU ELC — use of appendices
分界線：
If the data is very important to the understanding of what is being discussed in a chapter, then it should be given in the body. However, if it is not central to the content of the chapter, it should be placed in an appendix.
PolyU ELC — use of appendices
呢句就係第一步分堆嘅官方依據：判準係「讀者唔睇呢張表就睇唔明呢一章」，唔係「呢張表我跑得辛苦」。
典型附錄內容：
The type of information usually included in appendices will include technical notes on method, materials used to collect information such as a questionnaire, detailed results of experiments or observations, illustrations, photos.
PolyU ELC — use of appendices
港大 DASE 嘅 FYP handbook 對附錄嘅描述亦包埋統計細節：
Supplementary information, such as derivations of mathematical equations, details of survey questionnaires, data tables, figures, detailed statistical analyses etc.
HKU DASE — B.Eng. Final Year Project Handbook 2024-2025
最容易漏嘅一條規則：附錄唔可以孤立存在，正文一定要引到。
Every appendix must be referred to in the body of the FYP.
PolyU ELC — use of appendices
即係你把 Levene’s Test 放咗 Appendix C，正文就要有一句「詳細變異數同質性檢定結果見 Appendix C」。冇呢句，個附錄技術上係唔應該存在。
擺位同命名：
Appendices are placed at the end of the FYP, not at the end of chapters. Every new appendix must be given a title plus a number or a letter.
PolyU ELC — use of appendices
理大 EWrite 亦交代咗一個具體判準——會唔會撐長 Results 章節：
Include complex materials, such as a questionnaire template, detailed figures or tables that are not necessary for your analyses and discussions in the main text, or would disrupt the flow of the report or significantly lengthen the Results section.
PolyU EWrite — How to structure your project report
科大 CSE 亦有同樣意思嘅提示，對做 app 或系統嘅 FYP 特別實用：
(If you have a lot of screen shots, it might be best to put them in an appendix.)
HKUST CSE — FYP Final Report
搵唔到官方文件講明係唔係一定要把 SPSS 原始 output 檔或完整 printout 放附錄。理大 ELC 嘅例句提到過「A full computer printout was provided by the company (Appendix F)」呢類做法存在，但冇條文規定學生必須交。要唔要交，問你 supervisor。
你部私人電腦裝唔裝得 SPSS — 各校條文唔同
呢一段同寫法冇關，但同你趕稿嘅實際安排關係好大：你唔一定可以喺自己部電腦跑 SPSS。各校授權條文差異好大，逐間點名。
中大明文限制喺大學管理嘅裝置：
SPSS can only be used by staff and students under University managed devices.
CUHK ITSC — SPSS Campus-wide License
同一頁亦講明用途限制同續期安排：
This program is for staff and students to use on University managed machines in supporting teaching and non-commercial research purposes.
CUHK ITSC — SPSS Campus-wide License
教大就明文寫校內校外都可以用，而且講咗授權年期：
Under the new SPSS license agreement with IBM (from July 2025 to June 2030), staff and students can use the SPSS software both within and outside the campus for academic and research purposes and all SPSS Statistics modules are made available.
EdUHK OCIO — SPSS Software
浸大講明係 JUCC 集體協議，而且逐年續期，到期日固定：
All SPSS licenses are subject to annual renewal with expiration date on June 30 of each year.
HKBU ITO — SPSS
港大 ITS 嘅寫法最簡單：
SPSS is available for use by current staff and students.
HKU ITS — SPSS Products
三個實際影響。一，如果你係中大生而打算通宵喺自己部機跑數，先確認你部機算唔算 University managed device，唔好等到最後一晚才發現裝唔到。二，浸大嘅六月三十日到期日同好多 FYP 補交期重疊，暑假期間要跑數就要留意續期狀態。三，「current staff and students」呢個措辭意味畢業之後授權會停——如果你論文有補交或修改期跨過畢業，要預先諗定。
搵唔到官方文件明文講畢業後 SPSS 授權幾時失效、或者失效之後 .sav 檔仲開唔開得。中大圖書館有一句相近嘅提示，講嘅係另一個工具：「Please note that students lose access to DMP Tool once they graduate from CUHK.」——同類邏輯，但唔可以直接套落 SPSS。要準確答案，問你學校 IT。
交完之後數據要唔要留
好多人交完稿就 delete 咗 .sav 同 NVivo 專案檔。呢個做法有兩個風險。
第一個係制度層。中大圖書館嘅研究數據管理指引對研究型同修課式博士生有明文鼓勵：
Students are strongly encouraged to preserve their research data throughout the entire lifecycle—from its initial creation during the planning stage, through collection and analysis, to ongoing updates as their projects evolve.
CUHK Library — Research Data Management: For Students
同一頁對特定年份入學嘅學生講得更具體：
All RPgs and taught doctoral students admitted in 2025/26 or thereafter are strongly encouraged to preserve their research data, and wherever possible, share it in the CUHK Research Data Repository (RDR) or another trusted open data repository that provides a valid Digital Object Identifier (DOI) or another form of permanent identifier.
CUHK Library — Research Data Management: For Students
注意呢條係講研究型同修課式博士生，唔係本科 FYP。搵唔到官方文件講明本科 FYP 學生要保留原始數據幾久。但同一份指引提到嘅責任原則值得參考。
第二個風險係實際層：你嘅數字會被問。港大 DASE 嘅 FYP handbook 對數據提出嘅要求係代表性：
Ensure that the data collected are representative enough to warrant conclusions
HKU DASE — B.Eng. Final Year Project Handbook 2024-2025
要答得出「我嘅樣本足唔足以支持呢個結論」，你要開得返個檔。同一份 handbook 亦講到報告佔比同 supervisor 溝通嘅角色：
Final Report (70%) … Final Oral Presentation (15%) … Communications and discussions with the supervisor are included in ALL components in the Assessment
HKU DASE — B.Eng. Final Year Project Handbook 2024-2025
呢個比重（該系為例，各系唔同）意味報告本身佔大頭，但口頭環節仍然有實質分數，而口頭環節最常問嘅就係數字出處。科大 CSE 對 methodology 嘅要求把呢件事講到最盡：
The methodology section should serve as an instruction manual with which another FYP or FYT student could completely duplicate your results.
HKUST CSE — FYP Final Report
「Completely duplicate your results」呢個標準嘅實際意思係：你自己都要重跑得出。所以除咗留 .sav，仲要留你嘅 syntax 或者操作步驟記錄。SPSS 嘅 Paste 功能會把你點擊嘅操作轉成 syntax，跑之前按一下、儲存個 .sps 檔，就等於自動幫你留咗步驟。
如果你係碩士或博士層，數字出處會喺口試被逐個問到。呢一層嘅準備方法同本科 presentation 唔同，我哋喺viva 準備：香港 MPhil／PhD 口試流程同答題框架逐條拆過。
官方文件搵唔到嘅位
本文所有格式規則同授權條文都有明文出處。以下五項搵唔到官方文件，唔會用其他學校或者其他來源填數：
- 邊張 SPSS output 表必須入正文。院校指引只講原則（「唔理解該章就睇唔明嘅要入正文」），冇逐張表列清單。本文第一步嗰張分堆表係 UniWriterPro 編輯判斷。
- NVivo 結果嘅呈現要求。港大同浸大圖書館有 NVivo 教學資源，但係軟件操作，冇規定寫法。本文 NVivo 段落係編輯經驗。
- 原始 SPSS output 或 printout 是否必須放附錄。理大 ELC 例句顯示有人咁做，但冇條文規定必須。
- 畢業之後 SPSS 授權幾時失效、.sav 檔是否仲開得。各校 IT 頁只寫「current staff and students」，冇講失效時序。
- 本科 FYP 學生原始數據要保留幾久。中大 RDM 指引明文對象係研究型同修課式博士生，冇覆蓋本科層。
另外要講清楚：本文引嘅香港院校文件來自理大、科大、港大、城大、中大、浸大、教大七間，而且好多係系級或部門級文件（例如科大 CSE、港大 DASE），唔代表全校政策，亦唔代表其他院校有同樣寫法。冇引到嘅院校，本文冇作任何推論。
FAQ — 常見問題
SPSS 結果論文點寫，可以直接 copy output 貼落 Word 嗎？
技術上可以，但會撞三樣規則。一，SPSS 默認三個小數位，APA 對唔同統計量有唔同小數位要求。二，SPSS output 冇斜體，而 APA 要求統計符號斜體。三，如果你連表帶文都放同一組數字，就違反 APA「Do not repeat statistics in both the text and a table or figure」。所以正確做法係重排而唔係貼。
SPSS 出 Sig. = .000，論文要寫 p = .000 嗎？
唔應該。APA 官方指引明文寫「report p values less than .001 as p < .001」。SPSS 顯示 .000 只係位數不足，唔代表 p 真係等於零。
統計數字要幾多個小數位？
APA 分類講：整數量表（例如問卷）嘅 mean 同 standard deviation 寫一個小數位；其他 mean、standard deviation、相關係數、比例，同推論統計（t、F、卡方）寫兩個小數位；exact p 值寫兩至三個小數位。APA 同時提醒呢啲係一般原則，最終要衡量用途同精度。
p 值同相關係數要唔要寫小數點前面嘅零？
唔寫。APA 規則係：統計量理論上唔會大過 1 嘅（比例、相關係數、顯著水平）唔加前導零；可以大過 1 嘅（例如 t、F）就算今次個值細過 1 都要加。SPSS output 一律有前導零，所以呢一項要人手改。
表格標題應該喺表上面定下面？
兩套指引唔同，要跟你學系嗰套。理大 ELC 寫表嘅編號同標題喺表上面、圖嘅喺圖下面。APA 第七版把兩者統一，表同圖嘅編號同標題都喺上面。兩套都係正式指引，重點係全份稿一致。城大圖書館嘅 FYP 指引明文要求跟自己 Faculty 或 Department 嘅指引。
表裡面用星號定寫實際 p 值？
揀一種，唔可以兩種一齊用。APA 官方 blog 講明「you have two options, and you need to choose only one」。用星號嘅話，要喺表底 probability note 定義每個符號代表嘅範圍。
Levene’s Test、常態性檢定要唔要放正文？
搵唔到官方文件逐張表列清單。按理大 ELC 嘅原則判斷：「If the data is very important to the understanding of what is being discussed in a chapter, then it should be given in the body. However, if it is not central to the content of the chapter, it should be placed in an appendix.」前提檢查通常唔係該章重點，所以我哋一般做法係正文一句交代結果同因此揀咗邊條路徑，完整表放附錄。
附錄可以放咗但正文唔提嗎？
唔可以。理大 ELC 明文寫「Every appendix must be referred to in the body of the FYP」。而且附錄要放喺全份報告最後，唔係每章之後，每個附錄要有編號或字母加標題。
我部私人電腦裝唔到 SPSS，正常嗎？
可能係授權限制而唔係技術問題，而且各校唔同。中大明文寫「SPSS can only be used by staff and students under University managed devices」。教大就寫明校內校外都用得。浸大寫明每年六月三十日到期並需續期。港大寫「available for use by current staff and students」。趕稿之前先確認你學校嘅條文。
交完論文之後 .sav 檔可以 delete 嗎？
唔建議。搵唔到官方文件講明本科 FYP 要保留幾久，但中大研究數據管理指引對研究型同修課式博士生明文「strongly encouraged to preserve their research data throughout the entire lifecycle」。實際理由亦好簡單：口頭環節會問你數字出處，而科大 CSE 要求 methodology 寫到「another FYP or FYT student could completely duplicate your results」——你自己都要重跑得出。連 syntax 一齊留最穩。
Output 喺手，但 Chapter 4 寫唔出？
你已經跑完數，剩返嘅係最花時間嘅一段：邊張表入正文、邊張入附錄、每個數字用邊個格式、同一組數字唔可以喺表同文出現兩次。呢一段唔靠靈感，靠逐張表對住 research question 過一次。我哋嘅 FYP 分階段輔導會由你嘅 output 開始做分堆同重排，量化質性都做；如果你係碩士或博士層，論文章節寫作會連埋 NVivo coding 到成章敘事嗰段一齊處理。
我哋交嘅唔止一份排好版嘅 Chapter 4，係一份你答得出嘅 Chapter 4——每個數字由邊張表嚟、點解揀呢個 test、為咗咩刪走某幾欄，我哋會同你講一次，因為 supervisor 同口試都會問。
把你嘅 SPSS output 或 NVivo 專案截圖連 research question 一齊發過嚟，我哋先講得出邊幾張表用得著、邊幾張唔用。
相關文章：FYP Discussion 只係 repeat results？五個改寫做法、FYP 研究方法點揀、FYP Proposal 四段寫作流程、Literature Review 太描述性點改、Viva 準備：香港 MPhil／PhD 口試、FYP 係咩：大學生完整指引、 論文剩返兩個禮拜？即睇點應對