How a New Research Technique Becomes Standard: The 7 Stages of Published Papers

If you follow research in one field for a while, you notice a pattern in what gets published. A new technique appears in one paper, a few groups test it, and then many groups start using it, so a lot of papers end up looking similar. Eventually someone compares all the versions. A few years later the technique is in textbooks and people stop publishing about it because it has become a standard tool.

This pattern is often drawn as seven stages in two phases. In the first phase, peer review, the idea is tested and debated. In the second, standardisation, it becomes a trusted tool. Knowing which stage a topic is at helps you judge new papers on it, and decide whether it is worth writing one yourself.

Academic papers rising with the wind

I will use CRISPR gene editing as the running example, because it went through the whole peer review phase in about ten years and the history is well documented.

Phase 1: peer review

1. Technique papers

Someone introduces a new method and shows that it works, usually in a controlled setting. For CRISPR, that was the 2012 paper from Jennifer Doudna and Emmanuelle Charpentier's groups, showing that the Cas9 protein could be programmed to cut DNA at a chosen spot. It later earned them the 2020 Nobel Prize in Chemistry.

Papers at this stage are exciting, but they usually only show that the idea can work. Whether it is practical comes later.

2. First to measure papers

Other groups pick the technique up and test it on real problems, measuring how well it works and where it fails. Within a year of the 2012 paper, labs had shown CRISPR editing in human cells, and by 2015 it had been tried, controversially, in human embryos. These papers give the first independent numbers on how efficient and reliable the method is.

3. "Us too" papers

Once the method has been shown to work, many labs apply it to their own problems and publish, often with small changes. In the second half of the 2010s there were a huge number of CRISPR papers on crops, animal models and genetic diseases. Plenty of them were useful, but many were close to repeats of each other, which is where the slightly mocking name for this stage comes from.

If you are writing a paper at this stage, you need to show what it adds that the previous papers did not.

4. Evaluation papers

When there are enough results, researchers start looking closely at where the method goes wrong. For CRISPR, the main concern was off-target edits, cuts in places that were never meant to be cut. That led to work on measuring off-target effects and to more precise versions of Cas9.

5. Intercomparisons

By this point there are many versions of the technique, and people need to know which one to use. Intercomparison papers test them side by side, under the same conditions. For CRISPR that meant comparing Cas9 variants and the different ways of getting the editing machinery into cells, such as viral vectors versus direct delivery. They get less attention than the early papers, but they are often more useful in practice.

Phase 2: standardisation

6. Reference method

The technique is now the default. It is in textbooks and taught in courses, and new methods are compared against it. PCR, the method for copying DNA developed by Kary Mullis in the 1980s, is a good example. People do not write papers about using PCR anymore. They mention it in one line of the methods section.

CRISPR is moving into this stage. The first CRISPR-based therapy, Casgevy for sickle cell disease, was approved in the UK and the US at the end of 2023, a strong sign that the technique is no longer only experimental.

7. Public data

In the last stage, the protocols and data are public and anyone can build on them. The Human Genome Project, finished in 2003, released its sequence through public databases like GenBank, and the data has been the foundation for a huge amount of genetics research since.

Why it is useful to know this

This is a general pattern and fields move through it at different speeds. In machine learning it can take a couple of years. In medicine it can take decades because of clinical trials and regulation. Still, it helps in a few ways:

When reading papers, be more sceptical of claims in a technique paper than in an intercomparison, because early results often come from the best case.
When choosing what to write, keep in mind that late in the "us too" stage another application paper is hard to publish, while a careful evaluation or comparison is often welcome.
When choosing a method for real work, pick one that has reached the standardisation phase.

Here is the whole cycle with the CRISPR example in one table.

Stage	What happens	Example
Technique papers	A new method is introduced and shown to work	Programmable Cas9 cutting (2012)
First to measure	Others test it and report real performance	CRISPR in human cells (2013), human embryos (2015)
Us too	Widespread use, often with small variations	CRISPR for crops, animal models and diseases (late 2010s)
Evaluations	Critical look at weaknesses and risks	Off-target editing studies and more precise Cas9 versions
Intercomparisons	Versions compared side by side	Comparisons of Cas9 variants and delivery methods
Reference method	Accepted as the standard	PCR; CRISPR heading there with Casgevy (2023)
Public data	Data and protocols open for anyone	Human Genome Project data in GenBank (2003)

How a New Research Technique Becomes Standard: The 7 Stages of Published Papers