Dissertation Discussion & Conclusion 怎么写：从结果到洞察的完整指南（2026）

完成了 Results/Findings，很多学生以为 Dissertation 已经写完了大半——其实最难的部分才刚开始。

Discussion 和 Conclusion 是导师评分时最有区分度的两个章节：

• Results/Findings 展示你做了什么，得到了什么数据
• Discussion 展示你能理解这些结果，并将它们与学术对话连接
• Conclusion 展示你能综合全部工作，清晰说明研究的价值和局限

这是从"学生完成了任务"到"研究者做出了贡献"的关键跨越。

一、Discussion 的核心逻辑

1.1 Discussion ≠ Results 的重复

最常见的失分原因是：学生在 Discussion 里用自然语言重述了一遍 Results——这是浪费字数，也是错误的。

Results 的问题：What happened?（发生了什么？） Discussion 的问题：What does it mean?（这意味着什么？Why did it happen? What are the implications?）

1.2 Discussion 的四层结构

每个 Discussion 段落（对应一个主要发现）应当包含这四层：

1. 重申发现（Brief restatement）：用一句话说明这个发现
2. 与文献对话（Literature comparison）：这个发现与之前研究是否一致？
3. 机制解释（Mechanism / Interpretation）：为什么会这样？理论上如何解释？
4. 意外发现处理（Unexpected findings）：如果结果不符合假设，如何解释？

示例（定量研究）：

"Contrary to the hypothesis, social media engagement did not significantly predict academic performance (β = .08, p = .24; Table 3). This finding diverges from Smith & Jones (2020), who reported a moderate negative association (r = -.31, p < .01) in a sample of US undergraduates. The discrepancy may reflect cultural differences in platform usage: participants in the present study predominantly used WeChat and Douyin for professional networking rather than passive consumption, consistent with Chen et al.'s (2022) observation that Chinese students exhibit more goal-oriented social media behaviours than Western counterparts. Alternatively, the null finding may reflect a cohort effect — the present sample completed data collection during the post-pandemic period, when both academic demands and social media habits had normalised significantly compared to Smith & Jones' (2020) pre-pandemic sample."

这个示例展示了：

• 不回避 null findings（不显著的结果也要讨论）
• 与文献比较（Smith & Jones, 2020）
• 提出两种可能的机制解释
• 引用支撑性文献

二、处理意外发现（Unexpected / Null Findings）

这是中国学生最不擅长的部分——发现结果与假设不符时，往往陷入慌乱。

重要认知：null findings（不显著结果）不是失败，很多重要论文的核心贡献正是"发现某个假设不成立，并解释了为什么"。

处理意外发现的框架：

1. 解释可能的方法论原因：

   • 样本量不足（power analysis 显示至少需要 N=100，但实际只有 N=62）
   • 测量工具问题（量表信度不足，α < .70）
   • 样本特殊性（大学生样本不代表普通人群）

2. 解释可能的理论原因：

   • 调节变量（Moderator）：也许 X→Y 的关系存在，但只在某些条件下（如高焦虑组）
   • 中介变量（Mediator）：X 不直接影响 Y，但通过 Z 间接影响
   • 文化/情境差异：理论在西方发展，可能不适用于中国/亚洲情境

3. 呈现理论价值：

   • 即使是 null finding，也为文献提供了"边界条件"（boundary conditions）——知道理论在什么情况下不成立，同样有价值

三、Discussion 段落写作范本

定量 Dissertation 范本

假设支持：

"The first hypothesis, that autonomy support from supervisors would positively predict employee intrinsic motivation, was supported (β = .42, p < .001, R² = .18; Table 2). This finding aligns with Self-Determination Theory (Deci & Ryan, 1985, 2000), which posits that satisfaction of the basic psychological need for autonomy is a primary antecedent of intrinsic motivation. The present findings replicate the meta-analytic evidence of Slemp et al. (2018), who reported a mean effect size of r = .38 across 63 organisational studies, suggesting that the autonomy–motivation relationship is robust across cultural contexts. However, the effect size in the present study (β = .42) exceeds the meta-analytic mean, potentially attributable to the nature of the sample — a professional services firm where work is highly cognitive, increasing the salience of autonomy to overall motivation (Parker, 1998)."

定性 Dissertation 范本

"The theme of academic pressure as identity threat emerged as a dominant pattern across nine of twelve participants, resonating with Lazarus & Folkman's (1984) transactional model of stress, whereby an event is perceived as stressful only when individuals appraise it as exceeding their coping resources. Participants consistently framed examination failure not merely as an academic setback, but as evidence of personal inadequacy — a pattern consistent with what Dweck (2006) terms a 'fixed mindset' orientation, wherein ability is perceived as stable and immutable. Notably, this theme was more pronounced among participants from STEM disciplines, potentially reflecting the dominant discourse in STEM cultures that frames intelligence as a fixed, inherent trait (Murphy & Dweck, 2010)."

四、Limitations（局限性）的专业写法

Limitations 是 Discussion 末尾必须包含的内容（有时也放在 Conclusion 里）。

好的 Limitations 写法有三个要素：

1. 指出具体局限（不要泛泛"样本量不够大"）
2. 解释为什么这是局限（影响了什么效度？推论能力？）
3. 说明如何在未来研究中改进

模板：

"A notable limitation concerns the reliance on convenience sampling via a university participant pool, which yielded a predominantly young (M age = 22.4 years), highly educated sample. This limits the generalisability of findings to older working populations and those with lower educational attainment, where the relationship between X and Y may operate through different mechanisms (reference). Future research should employ stratified random sampling across age and occupational groups to enhance external validity."

常见局限类型：

五、Conclusion 章节写法

Conclusion 是独立的最后章节（不是 Discussion 的一部分），通常 500-1500 字，完成以下任务：

5.1 标准结构

1. 研究问题回顾（1段）
2. 主要发现概要（2-3段，无统计数字）
3. 理论贡献（Theoretical Contributions）
4. 实践意义（Practical/Managerial Implications）
5. 局限性与未来研究（如未在 Discussion 详述）
6. 结语（最后1-2句）

5.2 研究贡献的写法

这是 Conclusion 最核心、也是中国学生最弱的部分：你的研究贡献了什么新知识？

理论贡献（Theoretical Contribution）：

"This study makes two theoretical contributions. First, it extends Self-Determination Theory (Deci & Ryan, 2000) to the context of remote work, demonstrating that the three basic psychological needs (autonomy, competence, relatedness) remain significant predictors of motivation in virtual environments, albeit with different relative weights than reported in co-located settings. Second, the mediation analyses provide empirical support for the proposed pathway through which technological self-efficacy translates supervisor support into intrinsic motivation, a mechanism not previously tested in SDT research."

实践意义（Practical / Managerial Implications）：

"The findings carry practical implications for HR managers in professional service firms. The significant effect of autonomy support (β = .42) suggests that managers should prioritise providing employees with flexibility over task sequencing and work methods, rather than focusing solely on outcome monitoring. Concretely, structured autonomy-building practices — such as weekly planning sessions where employees set their own priorities — may be more effective than conventional supervision approaches."

5.3 结语（Final Paragraph）

结语要有力，留下印象，不要只是"总的来说，这个研究很重要"。

好的结语示例：

"In an era of increasingly distributed work arrangements, understanding what sustains employee motivation in remote environments is no longer a peripheral concern. The present study demonstrates that the fundamental human need for autonomy does not diminish in virtual settings; if anything, its fulfilment becomes more deliberate and consequential when stripped of the ambient social cues of a shared physical workspace. As organisations continue to navigate hybrid and remote work models, attention to the quality of the manager-employee relationship — and specifically to how that relationship either supports or thwarts employee autonomy — may be among the most impactful levers available to sustain performance and wellbeing."

这个结语做到了：

• 连接更大的社会背景（remote work 趋势）
• 重申核心发现（autonomy remains critical）
• 给出实践建议的方向
• 语言有力，不平淡

六、Discussion & Conclusion 字数参考

七、提交前 Checklist

Discussion：

• 每个主要发现都有单独讨论（不只是重述结果）
• 所有发现都与 Introduction 引用的文献进行了比较
• 意外发现 / null findings 有理论和方法论解释
• Limitations 具体（不只是"样本量小"），并提出改进建议

Conclusion：

• 明确陈述了研究问题，并给出了回答
• 理论贡献用"this study extends / contributes / challenges..."句式写出
• 实践意义具体，有行动指导性
• 没有引入新的数据或分析
• 结语有力，留下深刻印象

这是 Dissertation 三件套系列的最后一篇。配合Literature Review 写作指南和 Methodology 写作指南一起阅读效果更佳。

需要 Dissertation 全程辅导（从选题到最终提交），欢迎扫码联系学霸辅导导师。