Credibility of research depends upon duplication

I am often asked why there is so much "duplication" in research. Those who ask this clearly think there is a lot of duplication going on, and that this is very undesirable.

Some of this supposed duplication is a mirage. A politician once asked me why Children's Medical Research Institute and another well-known research institute should both be supported because they both do cancer research and therefore it is obvious that they must be doing the same things! The reality is that there is so much to be done to understand cancer, so many types of cancer, and so many facets of this disease that need to be investigated, that the chance of two researchers unknowingly doing exactly the same research is quite small. Besides, cancer researchers spend a lot of time and effort trying to keep themselves as up-to-date as possible with what their colleagues are doing. The same applies to research into other diseases.

There are some aspects of the research endeavour where duplication certainly needs to be avoided. For example, there may be little point in funding two separate tumour banks containing the same type of tumour but with incompatible databases, when combining all of the tumour specimens would provide a more powerful resource for research. It would be wasteful for neighbouring research institutes to each buy the same type of expensive research equipment when they could effectively share the one instrument. The research community clearly recognises this. For example, for more than a decade now the research institutions on Sydney's Westmead campus have coordinated the purchase of all major research equipment which is then made available for use by all of their researchers, and by researchers elsewhere in Western Sydney. Research facilities are also shared well beyond a geographical location when this is possible. CellBank Australia is the only facility of its kind in Australia, and Children's Medical Research Institute (CMRI) operates it on behalf of researchers throughout the country. CMRI is also building a unique, specialised mass spectrometry facility that will also be available to researchers throughout Australia. In turn, CMRI benefits from the availability of research facilities that have been built and maintained in other research organisations.

But the most important point to make about duplication is that it is an essential part of science. Unnecessary duplication of research equipment is wasteful, but it is imperative for research results to be duplicated and failure to do this can be disastrous to the research enterprise. In fact, duplication is usually not enough - results usually require replication many times over. This needs to happen at several levels. At the first level, an individual researcher or team usually needs to replicate an experiment many times before the results will be accepted by their colleagues. It is then essential for the results to be confirmed independently in one or more additional laboratories. This applies to clinical research as well: we need to see the same result in multiple independent clinical trials before recommending a new treatment to patients. This level of caution - this checking and double checking - is the bedrock foundation upon which the success of science is built.

It is not duplication but the lack of sufficient duplication that most needs to be guarded against. Researchers can easily be impatient and too keen to move on to the next experiments before properly validating the last set. This can be like building a house of cards that later falls down when a card at the bottom is pulled away. Resources are wasted if research is built on results that were not adequately checked. A complicating factor here is that there are few rewards for doing the essential work of replicating results from another laboratory. Researchers who first make a discovery rightly get the kudos, because making it can involve a large number of failed experiments before finding what works, whereas repeating someone else's successful experiment is usually much easier. In many ways, the ideal situation is where two or more research groups independently come up with the same results at the same time. Superficially, this seems like a waste of resources - multiple groups doing the same thing and achieving the same results. But, in reality, what this means is that at the time the results are first known the essential work of replication has already been done, and all of the research groups can share in the credit.

In research, lack of appropriate duplication is ultimately very wasteful. Duplication needs fostering and applauding.

Roger Reddel
Lorimer Dods Professor and Director
Children's Medical Research Institute