如果作为学生，你曾被告知：你的研究“描述性太强”，你需要“更具分析性”；
 
如果作为一名教师，你曾告诉你的学生：你们的作业“描述性太强”，缺乏“分析深度”。
 
那么，你应该来读读这篇文章。

事实上，尽管我们普遍鼓励让作业和研究论文“更具分析性”，但老师可能从未让学生知道，什么是好的分析标准。分析，可以说是社会科学殿堂中的一头大象，包括所有的那些带有“数据分析”字眼的相关课程——例如我所开设的“话语分析”。

分析是一个转化的过程：原始数据和信息本身并不具有意义，是我们通过分析的过程赋予其意义。
 
分析也是一个创造的过程：我们通过分析所创造的是，关于世界的新话语，它帮助人们以不同的方式感知世界，从而理解他们在阅读我们的研究之前，不知道或不理解的事情。
 
但分析是如何生效的呢？分析又是如何使世界变得有意义的呢？笼统地说：分析，通过聚焦、整理、命名以及建立范式与关系，来帮助其他人建立对这些范式与关系的感知，从而以不同方式理解世界。
 
有一个关于分析的比喻很好，那就是天文学家和其他观星爱好者识别星座的工作。天空布满了星星，但是尽管如此，有些人还是在天空中画出了图案，给这些图案命名，甚至围绕着这样形成的形状创造了一系列故事。这种感知天空的方式代代相传，被人们传授并写在书上，了解过这些知识的人在看星星的时候，都会不自觉地注意到星座。

以下，是来自方法论专著中的两个关于分析的定义：
 
“从字面上看，分析是将复杂的东西‘分解’成较小的部分，并以这些较小的部分的属性和它们之间的关系，来解释整体” (Robson 2011: 412)
 
“分析是给数据带来秩序的过程，它将已存在的东西组织成模式、类别和描述性单元，并寻找它们之间的关系；‘解读（interpretation）’涉及在分析中赋予意义和重要性，从而解释模式、类别和关系...”（Brewer 2000: 105）
 
这些定义，突出了分析工作中的双重过程。一方面，分析工作是对复杂混乱的现象进行分解和简化；另一方面，分析工作会在选定的元素之间建立模式，以产生一种解释世界的新的混合思路。
 
在研究的实际操作中，一个小时的访谈，转录记录可能会多达40页纸，而10个访谈的转录记录，就相当于400页纸。我们需要将我们的材料分解成可管理的部分，并专注于某些元素而放弃其他元素，以便我们能识别出隐藏在这些数据中的最有趣的知识。

作为社会科学家，从我们的数据中创造任何意义和话语是可以的吗？我说过，分析是一个创造性的过程，但这是否意味着可以创造一切？
 
答案当然是否定的。意义不是原始数据所固有的，而是研究者在培育经验材料的过程中，与其共谋的结果；因此，这一观点并不意味着从社会科学的目的出发，我们能从数据中随意创造任何话语。
 
针对我们需要怎么做才能实现良好的分析的问题，我们需要制定一些明确的标准，才能帮助我们确保我们的分析目标处于正轨。总结一下就是说：分析的目的是在基于经验数据的基础上，产生一个令人信服的证明，以描述、解释和解读一个社会现象

实际上，上述定义的每个词都有其重要性。在下面的“车轮状”示意图中，我对分析的这个定义进行了详细的阐释，以明确良好分析的标准。

让我们进一步解读这些标准：

• 描述：

​即为讲述清楚发生了什么事？记得确保在描述证据之前不要妄下结论。

• 解释和解读：

对经验材料的描述是分析工作的一个必要层面，但这还不够。读者需要我们再往前走一步，向他们解释正在发生什么。根据我们的研究问题，举个例子，它可能是情况背后的原因、它们出现的条件、行动者如何在不同情境中对其进行意义的解释等等。我们需要超越简单地描述数据中的内容，通过将我们的材料与现有的理论和文献联系起来，综合解释和解读我们所审查的议题和情况。

• 具体的研究对象：

通过选择相关材料进行分析，为研究问题提供答案。我们需要的，只有与分析相关的材料！我们有一个目的，就是我们的分析需要坚持与研究对象相关。我们的数据中会有一些有趣的东西，但并不直接回答我们的研究问题，此时，我们必须无情地牺牲掉它们！意大利面非常好，但如果我们是要去烤蛋糕，我们就不要把意大利面放进蛋糕里，否则我们就会把蛋糕弄坏，前功尽弃。总结一下就是说：我们的数据中可能有一些非常有趣的东西，但如果它们脱离了我们具体的研究主题，它们就需要被淘汰。

• 为别人的论述：

分析工作，不是我们为自己写数据总结，它是我们传达给别人的东西。它需要清晰和具有吸引力，它是一项论证工作，我们需要说服读者；想象一下，我们正试图说服我们认识的人!

• 呈现：

实现令人信服的分析工作的一个非常好的方法，是呈现我们分析的严谨性，创造一些不仅仅是能让人产生深刻印象的东西。我们需要超越择优挑选的轶闻般的案例，争取使我们的论证变得透明和可追溯，使读者能够理解我们是如何得出我们的解释的。我们需要对假设、程序和步骤提供一个清晰的说明，从而努力实现一个系统性的论证。我们还需要制定有步骤的分析策略和程序，并始终如一地遵循，这也意味着我们要用同样的分析框架和同样的问题来处理我们的所有数据（例如我们的所有访谈记录）。


我们每个人都有天然的优势和劣势。例如，我们中的一些人在清晰地表达他们的论点方面没有问题，但在使他们研究课题环环相扣方面却很费劲；也有人擅长提供丰富的解读，但却不太能呈现他们是如何得出这些结论的。我们需要做的，首先是了解清楚哪些分析维度是我们的弱点，或者哪些维度是我们经常会忽视的，随后，将我们所意识到的我们的劣势作为优先事项而加以努力。
 
“车轮状的分析过程示意图”，或许可以作为一个罗盘来指导我们的工作，使其“更有分析性”，而使用这个示意图检测我们研究的关键时刻，是在我们完成第一稿之后。我们可以把车轮中所标注的流程作为一个检查表，把它的不同要素与我们迄今取得的结果进行比较。我们是在提供解读，还是仅仅描述数据中的内容？我们是把我们的分析作为自己的总结来写，还是作为旨在说服听众的论述来写？我们需要在这个车轮状的示意图与我们的分析之间反复检查

This article is also available in English here, and as pdf.
Puede leer el artículo en español aquí.

-“How do I find a research topic when I have no ideas?”
- “I am trying to identify a research project, where should I start?”
-“How do I know which of my research ideas is the best adapted for a research assignment?”
​
These are questions I am often asked by students in the modules assessed by autonomous research projects that I teach at the London School of Economics and Political Science. In this blog post, I have put together for you a three-step process to help you a find a research topic:

1. Starting points to identify a research topic
2. Criteria to help you select the most adapted research topic
3. One thing to keep in mind: research topics are not set in stone and they evolve over time

• Starting point 1: Something that interests you in the socio-political world. Is there a social or political phenomenon that sparks your curiosity? Events that you would like to know better? Problems that you would like to understand? These are all relevant starting points. Your next step then would be to identify the key words and concepts through which social scientists talk about such phenomena so you can identify the relevant literature.
• Starting point 2: Analytical frameworks (theories and concepts) that make sense to you and that you would like to use. Are there theories or concepts that you would like to explore? That you would like to apply to a case or phenomena to help people understand it differently? This is a common way of starting a research project. You next step then would be to identify a specific case to which you can apply your framework, and make your ideas more concrete.
• Starting point 3: There is a piece of research that inspired you and that you would like to emulate. Are there articles that you've read that really made you think and stayed with you for a long time? An academic conversation/debate that you would like to contribute to? Then go for it. Your next step would be to identify what exactly in this publication/conversation you want to focus on and find an angle through which you can contribute through your project (you can write something that is inspired by a published piece but this must be different to a simple copy of it, be careful of plagiarism!).

​
Take a step forward: Brainstorm using the three starting points above and list 3-5 potential research topics. For each of these options, address “the next step” identified (1. finding key words in the literature, 2. identifying a potential case, 3. pinpoint specific contributions/positioning).
At the end of this exercise, you should have 3-5 potential ideas for research topics in front of you in writing.

1. Module’s/programme’s expectations. If you are developing a research project for a graded assignment, the first thing you should do is to read the documentation for your module to make sure you understand the objective(s) of the assignment, its potential structure, and marking criteria. Make sure that you pick a research topic that enables you to showcase the skills you will be assessed on. For example, if your module is about text analysis, pick a research topic that enables you to analyse textual data. Or, if your module is about producing a research design for empirical work, make sure that your topic enables you to do just that.
2. Skill set. Chose a topic that you have the capacity to investigate. By skillset, for example, I refer to your methodological and linguistic skills. Basically, you need to make sure that you have the skills required to deliver your project to completion. If the topic requires the use of certain methods of data collection and analysis, you need be comfortable using them. For example, if you have to produce an empirical analysis but you have never done statistics or text analysis, avoid choosing a research topic that requires advanced methods in these domains – let’s say multivariate analysis or discourse analysis – if your deadline is a few months away and you are not taking specialised modules to learn these methods. Producing research takes time and if you work under time constrains, it might be more strategic to use methods you already know or that are easier to learn on your own (let’s say descriptive statistics or thematic analysis if we follow the previous example). The same goes for learning a new language. If a project requires learning a new language, that might work for a PhD where you have several years to develop your skill set but, considering all the other things you have to do in a shorter programme, it might not be reasonable to assume you will be able to do so.
3. Personal interest. Another important criterion when it comes to selecting a research topic is to consider what sparks your interest and curiosity. Better to pick a topic that you personally enjoy learning about rather than something that you would do just because a lot of people are doing it. For many students, the research projects they do during their studies are the only ones they will do in their life, so you might as well pick something you are passionate about. Also, you will be more motivated to spend time working on your project if you have a personal interest in it and therefore more likely to produce better work!
4. Feasibility. Make sure that you pick a project that you can actually deliver. This criterion builds upon the skill set point I made above and encompasses other elements. Quite often, students come to my office hours saying things like: “It is the perfect topic but to be able to do it, I would need to have access to secret military material that is classified and which therefore I cannot access”. By definition, if your project is impossible for you to do, it is not a perfect project! A perfect project is a project that shows the readers something they did not know before and that you can finish on time. You need to find the right balance between going out of your comfort zone and forcing a project onto yourself that you cannot actually deliver.
5. Planned trajectory. Finally, I also encourage students to choose their project in relation to what they want to do after the end of their programme. Depending on the professional path you want to take, producing a piece of work that you can showcase on your resume or send to your future employer might be useful. Here I think in terms of your research question/research problem, case study, and methodological skillset demonstrated. Between the different options in front of you, is there one that is more likely to be seen as a bonus for the type of interviews/applications you are considering?

It is likely that no research topic will perfectly tick all these boxes. So rather than searching for a research topic that surpasses all other research topics, it is about you comparing the different options in front of you, assessing the trade-offs and making an informed choice on this basis.

Finally, it is important to consider that the topic you have chosen will likely not be exactly the same topic as the one you end up submitting. By this I mean that it is normal that your research project evolves throughout your research journey, which includes transforming/tweaking some aspects of your initial research topic. That research topics evolve is a process inherent to research; letting go of our previous ideas, perceptions and anticipations is an integral part of being a self-critical researcher!

​That being said, transforming a research topic through research is different to changing a research topic all together. By changing topics, I mean that you have started working on one of the options selected through the criteria above, but a couple of months down the line you decide you want to work on another option. This is a different scenario. If this is something you consider doing, you need to ask yourself the reasons behind this desire (as some might be strategic and relevant but many are ... not). For example, if you realise that you didn’t anticipate a feasibility issue, it might be wiser to reorient the research question instead of changing the topic altogether. Indeed, if you have a deadline, time is precious and it might be risky to completely change a project after a certain point.