Selecting the relevant papers for a survey paper I have to write?

Question

I have to write a paper as an assignment for one of the courses that I take in University.

The main outcome of this lecture is having a good final paper in the end. As we are Masters students, the paper that I am going to write resembles more a survey paper in a given topic, rather than a research paper presented by the authors of a specific research.

I have my topic assigned already. I have done my literature search and organized all the papers that I want to use. Unfortunately, at this point I am not that deep into the topic. But, everything I read seems to me like relevant for the topic, and somehow useful.

I have had difficulties with a similar situation before. As I was trying to include as much as possible relevant material to the survey paper, to make is complete. But in the end I was blamed that the paper loses its coherence.

Now I am stuck. In fact I am reluctant, because reading multiple papers, multiple times, writing them and then getting bad remarks does not feel nice.

What criteria should I use to filter my papers all over again, and what approach should I use when writing a survey paper.

Answer 1

The answer may differ from field to field.

First of all ask yourself "What is the central question" or the goal of you paper?

Do you want to give an overview on all aspects of the theory / concept? Or is it more about the central aspect of the theory / concept and extensions / applications are not that interesting? Or something else?

Then think about the story you want to tell. If its about the theory itself you might want to talk about the development of the theory, central aspects, proofs, and maybe some applications or extensions at the end.

If you want to give a general overview, try to create groups, are there some main movements, aspects, fields of application? Where are the links between these things?

Then select your paper according to this "central story".

While writing think about the audience of you paper, which concepts are common knowledge? Which are important to explain? Which depth of information is appropriate?

And, of course, try to get more feedback: Read the feedback of your old paper, then read your paper and try to find the flaws. Ask the person who graded you to give you some advice. Read some paper of your fellow students which got good marks, find out what you like about them.

in short:

1. Identify the central question of your paper
2. Create a structure / central story    
3. Think about the audience
4. Fill the structure, create links between your sections.

Answer 2

To add to Frederick's excellent answer, after you have settled on your central question/goal/thesis AND have created at least a rough draft of your paper, use the 'So what?' method. For every section, idea, and paragraph, ask yourself "So what?". Why does this need to be here? What purpose does this serve? would the narrative still flow without this piece? Don't automatically throw it out if you cannot answer the questions immediately, but use this as a way to assess what you have written, find the weak spots, and begin to understand how to improve them. This also helps weed out irrelevant material, so that you can avoid a bloated, incoherent paper.