Skip to content
Home » The Effortless Blog » A Visual Strategy for Finding Research Gaps

A Visual Strategy for Finding Research Gaps

Finding a research gap is the prerequisite that no scientific paper can exist without. But what does it mean, and how do you find a gap? While there is no secret AI shortcut to finding one, strategies and tools can help you find it faster and with more confidence. Let’s dive in.

Join the Effortless Newsletter. Receive free tips on note-taking, literature review, AI tools, and other productivity topics specifically tailored for academics and students.

There are many definitions of a research gap: Unexplained phenomena, unanswered questions, mechanisms that are not understood, methods that are underexplored, and so on. But they all have one thing in common: They can be described as a missing connection between two things. For example, if you can show that “climate change increases tree mortality”, you have answered a question or filled a gap. The emphasis here is on “increases” because it connects all our information about climate change and the literature on forest/tree mortality. Trees die independently of climate change for many reasons (drought, insects, Christmas), and climate change happens independently of forests (e.g. human emissions). So, your paper on “Climate change increases tree mortality” connects these two, and finding a research gap is equivalent to finding a connection or link. In this post, we will use this strategy to find missing links in scientific research.

To find this gap, we first need to lay out all the topics around our area of research. The research gap will be almost directly visible in the diagram. We will use Draw.IO, one of the best tools for making diagrams. The main reasons I prefer it over others (especially online tools):

  • Your diagram is inside an editable image. You can embed it into your notes, edit it later and have all the notes updated.
  • It is complex enough to allow for any layouts (unlike, for example, Scrintal, Obsidian Canvas or Heptabase).
  • It is free and runs offline, making it pretty fast. You can synchronize all your work in the cloud using it in combination with Google Drive or Obsidian Sync.

Visualizing your research

We need to zoom out to see research gaps or connections. Let’s look at the result and see why this is a good idea before getting into the tutorial and exploring how to do it. Here is a video of a complex research topic that I have been exploring.

  • Each color and shape has a meaning
  • Different arrow colors/shapes also have meanings.
  • There are a lot of annotations referring to papers, e.g. “Doak 2010.”
  • There is a logical flow from left to right (which is not always the case).

This example summarizes a month’s worth of reading (just note how many papers there are). This means that what you read in a month fits into a relatively small overview. Now, I can access my entire reading in a few seconds. Visual diagrams like this give you massive amounts of information, improving memory and synthesis skills.

Here is another static example from a different domain. Notice the same ideas here as well: Dozens of papers are usually associated with connections, as connections are the research gaps, and thus, publications will always be related to connections in these diagrams.

The key to using visual thinking effectively is to develop a language and keep refining it until you can express everything your research requires. This language will be personal and work only for your domain. We will create our language step-by-step. To get an idea, here is the legend I used for the diagram above:

While it looks complicated now, it started with a few simple elements. Over time, more and more connections and node types have joined. You might already get the idea of how to spot the research gap: Find two unconnected elements and think of a possible new connection between them. Once you have mapped out all connections in the literature, any new and feasible connection will be a research gap. It will be almost obvious to spot. The work is in creating these diagrams. Luckily, making them is a straightforward step-by-step process requiring little thinking. We effectively broke down the enormous task of finding the research gap into a series of small (and enjoyable) steps!

Getting started with

DrawIO is a free tool that you can access online at I recommend downloading the tool and using it offline.

After downloading, you can create a new diagram. To embed this diagram into Obsidian (and many other tools), use the SVG (or PNG) format. The difference is that SVG is a vector format, which means your images can be scaled up without losing any quality.

You can then store these images directly inside your note-taking tool (I use Obsidian). In Obsidian you will store the SVG file inside your vault and then embed it like you would any other image using the ![[filename.svg]] format. Adding a |800 behind it will limit the image’s width, if necessary.

Obsidian embedding a DrawIO SVG diagram.

Creating a visual language for your work step-by-step

Let’s get started creating a visual language. I will run you through an example in my domain. It will be very different from you. Note, however, that we start simple and become increasingly complex; this is the key takeaway.

My example is about Species Distribution Models (or SDMs). A type of algorithm that takes observational data about a species occurrence and then learns the climate this species “likes”. The result is these maps of ranges that we have seen:

We identify 3 core elements here:

  • Green box: Symbolizes data of some sort
  • Gray shape: Symbolizes and algorithm that does something withthe data
  • Purple shape: Output of some sort.

The connections, too, have meaning.

  • Red: Means something is required for something.
  • Teal: Something creates something.

Now, let’s add some Problems. Every approach or solution has problems. Here are two problems that species distribution models have. Note that one problem is on the data side and relates to bad data, while the other is on the prediction side and affects a more fundamental problem with this algorithm.

Let’s add some papers related to species distribution models (SDMs). I add three types of papers:

  1. Yellow box: A paper that comments on a specific topic. Here on the data that goes into SDMs. The box allows to add a comment. [Simmonds 2020]
  2. Subtitle to Shapes: I added [Norberg 2019] to the SDMs box. From the placement, you might suspect that this is a review of different types of SDMs, and you would be exactly right.
  3. Arrow annotation: These annotate processes. For example, [Lee Yaw 2020] explore different problems SDMs. Therefore the annotation goes on an arrow.

These annotations with papers are the key to building great diagrams. You can use the placement inside the diagram to immediately remember what the paper was about. When looking at “Norberg 2019” above, there is no doubt that this is a review paper on SDMs, while “Lee Yaw 2020” explores something related to the problems of SDMs, right? You can, of course, mention papers multiple times for clarity.

One optional step that will help you spot a research gap is to add questions to your diagram. These questions can be copied and pasted from a review paper or your questions. They can be something nobody has answered or something you don’t yet understand. Here are three example questions related to species distribution models.

If you are unsure where to get questions, check for papers that make it easy for you. Here is a paper that helped me a lot define my own question. The author’s could not be any clearer on this:

The importance of research questions can’t be overstated, as I painfully learned myself in the first year of my PhD.

We have created an excellent and simple visual language that will allow us to map a large chunk of our research. Here is the legend:

Legend for a visual language using Drawio.

It is now time to read, take notes and document our research visually.

Mapping out your research

This will take you some time. Below is a screenshot of the final product of my research. It is a complicated diagram, but it was created step by step, slowly. I broke down a very complex scientific field into a series of simple 1-to-1 relationships. Each little relationship was a paper I read. It would be impossible to keep the entire field in my head, but it is easy to keep it on an endless canvas.

As a side effect, you will remember your research much easier once you have created a diagram, and you can discuss it with colleagues when trying to find a gap. If you have an iPad, this approach will significantly leverage its capabilities! If you are curious about an iPad note-taking approach, check out the Notero article.

Research Gap Type 1: Adding a step

To add a novel contribution, you could add a new step into your diagram. For example, I could add new types of data that will make my predictions much better than what was available previously. This adds an additional loop and some new data to the diagram.

Research Gap Type 2: Skipping a step

One type of gap is skipping a step in your diagram. This is called a methodology gap. Here, we develop a novel method as a shortcut in the diagram. In our example, this might eliminate a step and make predictions: better, faster, more precise, and so on.

Both of these gaps are not limited to computer science algorithms. Think of the grey boxes as “experiments” and the purple boxes as “insights” instead of algorithms creating outputs. Now, adding a step would mean refining someone’s experiment and making the insights more nuanced, and skipping a step might be combining two experiments into one and eliminating some errors from running consecutive experiments.

Research Gap Type 3: Missing connections or conflicts

Lastly, you can add a new insight by identifying conflicting connections. If one paper tells you one thing, but that new connection somehow contradicts what you already have in your diagram, you have found a conflict. Resolving this conflict with an experiment might be

In the example below, the light gray-dotted line hints at such a conflicting connection. The blue pills in the image are proteins inside the cell that form large interactive networks. One paper states that eIF3E binds to YTHDF, while another states that YTHDF is not needed as a binding partner. Maybe it was my mistake, or maybe it is a new mechanism that is not fully understood. Researching this small fragment can lead to a valid new research idea/gap.

Sometimes, you might also find connections missing in your diagram and ask: Why? Answering this question might again lead you to new and essential research questions.


In this article, we used to map out large portions of our knowledge. Dozens of papers gave rise to one large diagram. We created a “visual language” and defined what the different colors of boxes and arrows meant. This language evolved slowly over time, and we added elements as needed. This allowed us to break down a very complex and large topic into a series of small and manageable steps that together form a large diagram.

This diagram is a tool for spotting gaps by adding new steps, finding shortcuts, or uncovering conflicting connections. Each new connection is a potential research gap. You can also use this diagram for discussions or to memorize and summarize your research.