When to switch from reading to implementing in your PhD?

To read or to code is the question

In the early days of your PhD, you are required to read, a lot and there is hardly a magic formula to tell you how much is too much to read. So, how do you decide when you need to switch from reading to implementing?

Several PhD scholars have asked me this question over the years in one form or another so here is a framework I use to help guide the early stage researchers (ESRs).

Disclaimer: This is an agile framework that can easily be expanded and modified to fit most of the PhD research niches.

Step-1: Literature Review Database

One of the ways for an ESR to figure out what is it that they need to do, is to figure out what others are currently doing (and not doing, aka limitations).

Think of it as spying on your competition though with good intent also known as literature review, an important building block of any PhD proposal.

To help with this, I released a FREE Notion database template along with a complementary video to walk you through with the set up.

The idea with this database is to do away with the pen and paper scribbling because things get real messy when you want to query something: Which of the authors said X and all you have is a pile of papers.

I switched to Notion a little too late but I got through just fine as I have photographic memory. But, unless you have a photographic memory too, you would benefit from maintaining an active literature review database.

This is hardly a one time thing: this needs to be consistently updated.

Step-2: Assimilation of State-of-the-art (SOA)

Now, as you are populating the literature review database, you should also keep in mind that at some point in your PhD journey, you would also be required to write a Research Proposal. This would need you to have a strong grip on the following aspects:

• What is the current state of art?

  • The lit review database should have most of these works listed already

• What are there limitations?

  • As you read the SOA, make notes of their limitations and future work in the database

• Which of the limitations are worth addressing right now?

  • Can you rank the works in terms of the criticality?

• Which of these limitations would become a part of your thesis?

  • Are you planning on doing incremental or ground breaking research?

All of this information is then presented neatly in the form of a research proposal though by now, you should clearly see the importance of the literature review database.

Step-3: Research Blueprint (Block Diagram)

This step will serve as the blueprint to plan and monitor the progress of your research. Typically, I ask my mentees to create a block diagram wherein,

• Block: represents which research question is being addressed

• Links: represented as arrows, these have a dual purpose:

  • Causality: clarify the causal relationships between various blocks

  • Outputs: clarify what output(s) in the form of publications are to be expected. There can be more than one publications as an output for a block which would require commensurate amount of time and effort.

Step-4: Research Timeline (Gantt Chart)

Take the blocks you have set up in the step above, and identify what kinds of tasks would they need to come up with the outputs you have labeled and what would be the (measurable) milestones for progress tracking.

Lay out such tasks and milestones as a waterfall model/ Gantt chart (see sample below) along with the intended time required.

To Read more or to implement more?

We started out with the million dollar question: How do you know that you have read enough and need to start implementing stuff?

Given the breadth and depth of your niche, there might be numerous researchers you might be doing this exercise too. There is a good chance somebody else might beat you to the punch, i.e., the SOA is likely to keep changing and with time you will know how frequently.

The frequency of “significant” changes in your SOA determines the frequency of updates to your literature review database and also SOA assimilation.

If you can say with a significant level of confidence that given the list of SOA in your literature review database, you are fairly confident with your assimilation, you can get to implementing stuff.

Though, you would always needs to keep an eye open for competitors :)

Also, beware of the fact that when you are doing this exercise in the early days, there are a lot of assumptions and idealistic settings you are banking on.

As time unfolds, a lot would change and hence you need to be proactive in consistently updating this framework (see a visual representation below).