如何阅读文献？

如何阅读文献？

文献阅读目的

①记录主要内容

②扩充知识面

③借鉴表达

1️⃣作者是如何讲故事的？

2️⃣英语句型表达

分类整理文献

借助文献管理软件，推荐zotero

⚠️如果自己的研究课题缺少好期刊或者高被引的相关文献？

（1）或许你所研究的论题是一个学术前沿以至于文献资料欠缺

（2）或许你所选择的学术视角过于狭窄以至于材料不够健全

（3）最后一个不太乐观的问题就是你所选择的论题根本没有学术价值。此时要做的工作就是及时和导师沟通以便做好下一步的修正工作。

阅读文献的顺序

①Abstract（通常包含研究问题、研究目的、研究方法、基本结论）

• 读完abstract后思考：
  
  

✅研究的问题是什么？（问对问题，确定研究的主要构念）

✅研究问题和研究对象要匹配，属于哪一个层面？个人、团体、企业？

✅这个问题为什么重要，为什么值得关注？

✅过去的文献对这个问题的解释有哪些？为什么对这个问题的解释不够好，哪里欠缺？

✅这篇文章与我的研究内容是否相关?

✅采用的方法是否新颖，我是否感兴趣?

✅他的结果，是我想要的么?

• 得到肯定回答后，继续思考：
  
  

✅他是如何把这种方法运用到这类问题中的?（学习）

✅它的结果完美么?是否是充分必要的?保守性强不强?有没有可以改进的地方?（总结+反思）

②Introduction（作者是如何讲故事的）

✅此问题的由来(历史)，作者为什么研究这个课题?
✅此问题在全世界内的研究进展(现状)，目前这个课题的研究进行到了哪一阶段?

✅引申出作者的研究动机，作者使用的理论是基于哪些假设?

③Conclusion（结论）

✅这篇文章存在哪些缺陷?

✅作者关于这个课题的构思有哪几点?

注意点：

⚠️即使通读了他人的现有文献，仍然没有涌现出新观点

⚠️没能简明扼要的概述

⚠️如果读英文文献遇到不认识的单词时，第一遍读的时候可以查询含义并标注，但是不需要高亮，容易掩埋重点

⚠️没必要一开始通读

⚠️非常相关的文献要反复读、反复思考

⚠️定期查看相关研究领域顶刊的abstract

⚠️理论的假设很重要。理论构建的时候最好能涉及到三个层次：宏大、中层、细微。

论文结构

1️⃣摘要Abstract

= ①研究背景
+②冲突（文献综述中提出的批评或研究空白）
+③作者为填补现有文献问题而做的研究（研究方法+研究目的）
+④研究发现

2️⃣前言Introduction（摘要的无结论、加理论依据的扩充版） ——要做什么 & 提出的主要问题

= ①研究背景
+②冲突（文献综述中提出的批评或研究空白）
+③研究重要性的理论依据
+④作者为填补现有文献问题而做的研究（研究方法+研究目的）

3️⃣文献综述Literature Review——目的：解决已有文献和目前发现间的早已存在的冲突（已经知道自己的研究成果是会支持还是抵触其他人已有的研究）

1）对现有文献的综述——对前人研究结果的简要综述内容：

①对研究主题的概括以及例证，突出研究的重要性，如“某某问题/现象已经成为媒体报道、社科研究和道德评论的焦点”；

②概念、定义和理论基础；

③研究方案、技术、结论以及相关问题的比较和分析；

④发展趋势；

⑤值得深入研究的方向；

⑥应用前景预测。

前期工作：对相关文献进行分类（如按照定性/定量、研究目标、研究方法等分类）和整理

2）对现有文献的批评——评论前人研究成果并指出其缺陷或局限性

例如：

✅批评细节不足；

✅理论上质疑作者引入的理论概念的有效性：某一概念的外延是否精确反映了这个概念；

✅方法上质疑自变量和因变量的测度以及意图能否反映某一特定概念的外延的准确性和等效性）

✅忽视了与先有文献的理论、概念或分析方法上的某一联系

📕常见词：“尽管”、“但是”、“然而”

3）空白——现有文献中缺失的成分（可以其他文献中的未来研究趋势和现有文献进行对比）

4）研究重要性的理论依据（研究意义）——提出依据以证明其研究是必要的，有理由的
⚠️如果未能充分阅读已有文献，那么我们提出的观点可能沦为自说自话或做无用功。

4️⃣研究结果Results

⚠️其他文献的研究成果是自己论文的文献综述部分的重要组成

5️⃣讨论Discussion

1）研究目的：旨在···
2）与现有文献观点一致的研究发现——当前研究和已有研究间的联系
3）与现有文献观点相反的研究发现——有没有提供新的洞见？

6️⃣结论Conclusion

1）本文已经回答的主要研究问题
2）对未来研究的建议——针对目前仍存在的研究空白提出路线图

EVALUATING A STUDY

This outlinewill be used for discussion of the papersin the small groups and the exam.

Summary of the findings

1.           What is the problem and what is known about it so far? (Background)

2.           Why did researchers do this particular study? (State thehypothesis.)

3.           Who was studied? (Describe the population including size.)

4.           How was the study done? (Describe the studydesign and duration, outcomes and risk factors/treatments and how they weremeasured.)

5.           What did the researchers find? (Present the major resultsof the study referring totables and figures. Report the measure(s) of effect/association (comparison ofrisks) between risk factor/treatment and disease/outcome (RR, OR, AR.)

Interpretation of the findingsand limitations of the study

Are the findingsbelievable (not the result of chance,bias or confounding)? Are the findings useful (generalizable and practical)?

6.           How may chance (randomerror) have affected the results? (Describe the methods used to evaluatethe probability of Type 1 error (confidence intervals, hypothesis testing) or Type 2 error (power).)

7.           How may bias (systematicerror) have affected the results? (Name the strategies used to minimizeinformation bias (blinding, objective measures of riskfactors and outcomes) or selection bias (randomization, complete follow up).Note the size and direction of any bias.)

8.           How may confounding(association with other risk factors) have affected the results? (Note the provisions made to minimizeconfounding (randomization,matching, stratification, multivariate analysis). Note also the size anddirection of any confounding effects.)

9.           What is the generalizability of the study findings? (Discuss the individuals or groups that may or may not benefit from the findings.)

10.        Are the findings practical? (Discuss the applicabilityof the findings to non study individuals or groups with reference to effect size, acceptability, availability, cost and obsolescence of thetreatment.)

Check list for specificstudy designs

(see Dr Levine’sLecture notes and Chapter 5 High-yield Biostatistics)

Case control

1.           Were the caseschosen from a population that was in all ways similar to thecases except for the disease or outcome being studied?

2.           Was the exposure/risk factor/intervention present beforethe outcome?

Cohort study

1.           Did the cohort exclude those already with the disorder?(i.e. an inceptioncohort)

2.           Was there completefollow up?

3.           Was the population representative

4.           Were there an adequate numberof events?

5.           Was there adequatetime between exposureand event?

6.           Was there appropriate control for confounding factors?

Controlled trial

1.           Was there randomization?

2.           Was there completefollow up?

3.           Was the analysisby “intention to treat”?

Evidence that an association reflectscause and effect

1.           Temporality: cause precedeseffect

2.           Strength of the association: large relative risk

3.           Dose response: increasing risk results in increasing disease

4.           Biological plausibility: consistent with biological knowledge of the time

5.           Consistency: similar to other studiesin different persons,places and times

6.           Reversibility: decreasing or removing risk reduces disease

Guidelines for article critiquesFirst: Summary
Summary of article. Try to be clear but concise. This should not be much more than 2 pages. This summary should include the following points, assuming they are relevant to the article under discussion.

1) The author's purpose in writing the article and/or the question(s) this article was intended to address. This should include the general questions that inspired the article as well as the specific questions the author seeks to answer.

2) The major points of background, including information on context, theoretical assumptions or definitions the author is using. This might include: background of speakers, theoretical perspective, theoretical constructs (e.g., social networks, priming), definitions of technical terms used.

3) A summary of the kind data that the author is using. Possible things to note: language, range of data, example sentences and where these examples come from. Try to focus on the general nature of the data. Is this a meta-analysis of several studies or details of a set of the authors' own research studies?

4) The major arguments of the article. Most articles will have two - three points that they are making, and several sub-arguments to support those points. Try to focus on the general structure of the arguments, not the details of the examples used to support those points.

5) The major conclusions of the article.

Next: Analysis

This section should display your own critical reading of the paper. Think about and discuss as many of the following questions as you can. Note that only some of these questions may be relevant to the article you are reading.

1) Does the author achieve his/her stated purpose? Does the data support the author's conclusions?

2) Are there any major flaws in argumentation? Are there any counterarguments? (Does the author deal with existing counterarguments? Can you come up with new ones?)

3) Are there any issues that the author has left unadddressed?

4) Can the arguments be modified to make it more explanatory? Extended to account for other data/languages?

5) What contribution does this article make (despite any potential flaws) to the field?

6) How does this article fit in with the larger issues in the field?