Fwd: How to do a literature search

[Sagi Jaffe-Dax]

A thread I received from Leon's lab.

Might be useful for those of you who are in the midst of writing something.

---------- Forwarded message ----------
From: <lia...@gmail.com>
Date: Fri, Feb 13, 2015 at 6:05 AM
Subject: Re: How to do a literature search

Hi all,

One more thing to add: while doing the search, you also need to document what you've found somehow. I usually do it in one of two methods. If I am mostly looking to nap findings, I start a ppt and copy the main figure of each paper I found (with the paper's name) into a slide. 

If I am trying to review the theories in the field, I open a word file, copy the relevant paragraphs from the paper (again, with the paper's name) while reading/reviewing it, and highlight key sentences. I found that this can substantially ease the writing process, which in many cases comes long after you've searched the literature (hopefully, before designing the experiment...) - it is then my easier to remind yourself what you actually found. 

In both cases, I divide the papers according to sub-areas (e.g., object processing, scene processing, object-scene interactions etc).

I don't know if this will help anyone, but it works pretty well for me :-)

Liad

Sent from my iPhone

On Feb 12, 2015, at 3:52 PM, Sagi Jaffe-Dax <sagi....@mail.huji.ac.il> wrote:

My two cents:

1. Pubmed might be useful for searching keywords.

2. Searching through recent meeting abstracts to get an idea on who's currently working in the field.

On Thu, Feb 12, 2015 at 3:42 PM, Tamar Regev <tama...@gmail.com> wrote:

Thanks Edden!

Another two thing that I would add, a bit less systematic but that I remember as key moments for me in familiarising with a new field - 

1. Choose a fairly new review paper and actually read it... -- it will help to get the gist of the field, map and group the key questions and evolution of ideas.

2. If you know a friendly expert in the field - you can either try to meet for a discussion or just ask for a few references by mail. This is of course only if you feel comfortable, or perhaps if you have a specific question or issue to discuss. Has been very helpful for me in the past. 

Tamar

On Thu, Feb 12, 2015 at 3:11 PM, Leon Deouell <deo...@gmail.com> wrote:

I would add

1- the process should be iterative, that is, the right key words to use are expanded and refined through looking at the search results

2- it is not just the 'cited by' but also the 'cited in' which leads you to relevant papers

3 - through the process you also map the groups or specific individuals  who work in the field and it is a good idea to survey their work and look for relevant studies. 

4 - in some cases there are mailing lists and things like research gate where you can post queries about relevant papers. This is handy when your search turns up fewer results than you expected. If you post questions, you should be somewhat careful not to disclose nature-worth ideas, but in general it is more important to not miss a whole branch of results you missed.

Leon 

On Thu, Feb 12, 2015 at 2:04 PM, Edden Gerber <edden....@gmail.com> wrote:

Hello lab and lab friends, 

This message is not about lab meetings, but about a topic which, it occurred to me now, is a big part of our work but rarely discussed explicitly - researching a topic in the literature. I thought it would be interesting just to bring up the topic and see what people have to say about it from their experience. 

Working on my research proposal over the past several weeks, for which I've had to get a general picture of a few areas of research which I was not familiar with, I've noticed that I have developed a more or less systematic approach to this task - so I will share it here as a starting point. Here it is:

1. Start with a general keyword-based search, and find a few papers which fit the topic I'm looking for. 

2. Among these papers's references, try to identify a small group of "canonical" studies or popular review papers. 

3. Produce the "cited by" list for these papers. Assuming that I went back far enough in time and correctly identified the papers everyone is citing, I probably achieved pretty good coverage of the field with this list. 

4. In case more "central" papers turn up in the process, repeat step 3. 

* I use the Web of Science search engine for all these step, or sometimes Google Scholar just for step 1 as it can sometimes do better with looking for words or phrases. 

Will be happy to hear your thoughts - 

Edden 

--

---------------------------------------------------
Prof. Leon Y. Deouell, MD, PhD
Department of Psychology

Edmond and Lily Safra Center for brain sciences (ELSC)
The Hebrew University of Jerusalem
Jerusalem 91905
Israel

URL: http://hcnl.huji.ac.il