本课为原创大学书面英语能力课程，不对应统一的人教版大学教材或固定大学年级；不含听力。练习目标只反映本站材料掌握情况，不保证实际考试分数。
学习目标
按背景、目的、方法、结果、限制识别一般研究摘要，而不假装掌握专业学科。
用给定虚构数据写结构清晰的摘要，不增加因果结论或显著性。
在本课主题中理解新增词块的语境用法，并把其中相关词项迁移到新的完整句子。
核心讲解
摘要是论文或报告的缩影，但不同学科格式不同。本课只教一般阅读与写作，不替代专业方法训练。常见语篇结构：问题背景→研究目的→方法范围→主要发现→意义与限制。短摘要不一定每项单独标题。
课程虚构摘要：Finding directions in a public building
Clear signs may help visitors navigate an unfamiliar building, but a sign that is visible may still be difficult to interpret. This classroom project compared two written sign designs in a simulated public building. Thirty-eight adult volunteers completed the same four route-finding tasks; nineteen used each design. The group using task-based labels made fewer wrong turns on average than the group using department names. However, the groups were small, and participants were not randomly assigned. The result therefore suggests a possible advantage of task-based wording in this setting, rather than establishing its effectiveness in every building. Further testing should include different visitors and real-world conditions.
目标是比较两种文字标牌；样本38人各19；结果只有“平均走错更少”，未给数字差，所以不能自编“减少40%”；未随机分配意味着组间差异可能有别的解释；suggests可能优势，不是proves普遍效果。
读摘要的四问
研究对象与我关心的对象是否一样？任务是模拟还是现实？结果测了什么、没测什么？结论强度是否与方法一致？读到significant时先看是否明确统计含义；没有显著性检验不能自行补p值。
写摘要的语言
方法常用过去时：Thirty-eight volunteers completed…；一般背景用现在时：Clear signs may help…；结果解释用suggest/may/within this sample。不要把整篇引言复制进摘要，也不要新加正文不存在的发现。
例题解析
问题：能否写“Task-based signs caused all visitors to navigate correctly”？
不能。原文fewer wrong turns不等于zero；small volunteer sample不等于all visitors；非随机组比较不足以排除所有差异。可写：In this simulated setting, the group using task-based labels made fewer wrong turns on average, but the design limits causal conclusions.
本课词汇与搭配
下列为本课实际讲解的词项；词形变化及同一词的不同义项不等于新增独立词汇量。
| 词项 | 本课义项及用法 |
|---|---|
| abstract | n. 摘要；write an abstract |
| simulated | adj. 模拟的；a simulated setting |
| participant | n. 参与者；recruit participants |
| randomly | adv. 随机地；randomly assigned |
| navigate | v. 找路／导航；navigate a building |
| causal | adj. 因果的；a causal conclusion |
| pre-existing | adj. 先已存在的；pre-existing differences |
| generalisability | n. 可推广性；assess generalisability |
词汇工坊
本表补足本课话题的读写词块，含中学承接和大学应用，不将每项都计为大学新词。先遮住中文解释推测例句，再用自己的句子替换人物、时间或对象。理解义项、掌握搭配、能够迁移是三项不同任务；下方两道迁移练习检验部分关键词块，不代表全部词项已自动掌握。
| 词项 | 义项与搭配 | 原创语境例句 |
|---|---|---|
| research question | 研究问题；state a research question | The abstract should make the research question clear. |
| methodology | 方法体系；describe the methodology | A short abstract summarises the methodology without every detail. |
| allocation | 分配；participant allocation | Participant allocation affects how the group comparison should be interpreted. |
| randomisation | 随机分配；report randomisation | Do not report randomisation if the groups were assigned another way. |
| control group | 对照组；a comparable control group | A comparable control group can help evaluate an intervention. |
| intervention | 干预；a small intervention | The intervention changed the wording of the signs. |
| external validity | 外部效度；question external validity | A simulated setting can raise questions about external validity. |
| internal validity | 内部效度；threats to internal validity | Pre-existing group differences may threaten internal validity. |
| replication | 重复研究；independent replication | Independent replication would help assess the finding's stability. |
| statistical significance | 统计显著性；report statistical significance | Statistical significance should not be claimed without the required analysis. |
| effect size | 效应大小；estimate an effect size | An effect size cannot be calculated from a statement that one group did better. |
| confidence interval | 置信区间；report a confidence interval | Do not invent a confidence interval when no estimate is provided. |
| causal inference | 因果推断；limit causal inference | Non-random group assignment limits causal inference in this example. |
| research abstract | 研究摘要；a cautious research abstract | A cautious research abstract keeps results and limitations connected. |
练习
先独立完成，再核对答案；开放题须按要点人工评分，不以是否逐字等同范文评分。
1
分别指出摘要的目的、方法和主要结果，各一句中文。
2
给这项虚构研究写一个不过度承诺的英文标题。
3
为什么不能计算平均错误减少百分比？缺什么数据？
4
把The experiment proved universal success改成准确有边界的结论。
5
用80—110词重写摘要，保留38/19/19、模拟情境和非随机限制。
6
写一句下一步研究建议，须对应现有研究的一项限制。
7
词汇迁移：译：如果参与者分配不是随机的，摘要就不应声称采用了随机分配，也应限制因果推断。
8
词汇迁移：译：在缺少所需分析的情况下，不能捏造统计显著性、效应大小或置信区间。
参考答案与解析 · 完成练习后展开
参考答案与解析
1
目的：比较任务型与部门型文字标牌；方法：38名成人志愿者，分两组各19，在模拟公共建筑中完成四项寻路任务；结果：任务型组平均错误转弯更少。不能补没有给的具体均值。
2
Comparing Two Sign Designs in a Simulated Public Building。Avoiding Every Navigation Error过度承诺。
3
缺两组平均错误数，只有方向性fewer不能计算幅度；38/19是人数，不能充当错误数。
4
The findings suggest a possible advantage of task-based labels in this simulated setting, but the small, non-randomly assigned groups limit broader conclusions. 保留可能、场景和限制。
5
This classroom project compared two sign designs in a simulated public building. Thirty-eight adult volunteers completed four route-finding tasks, with nineteen using task-based labels and nineteen using department names. The task-based group made fewer wrong turns on average. However, the sample was small and participants were not randomly assigned, so the result does not establish a universal or purely causal effect. The findings suggest that task-based wording may be useful in this setting. Further testing with different visitors and in real buildings would help assess how widely the result applies. 20分：目的3、方法人数5、结果3、限制5、结构语言4。
6
Future work could assign participants randomly to reduce the risk that pre-existing group differences explain the result. 或Testing signs in real buildings would help assess whether the simulated result transfers to practical use。针对性明确即可。
7
If participant allocation was not random, the abstract should not claim randomisation and should limit causal inference. 过去事实was与现在写作建议should兼容。6分：信息3，词块2，句法1。
8
Without the necessary analysis, writers must not invent statistical significance, effect sizes or confidence intervals. necessary analysis意为所需分析，而非不必要的分析。6分：信息3，词块2，句法1。