Best time to publish research revealed

It is one of the great dilemmas of the research world - to publish a working paper or not?

Now David Gill, an Oxford University economist, has created a model that answers the conundrum of whether to share interim research results or to play your cards close to your chest.

Dr Gill, who revealed his findings at the World Congress of the Econometric Society, said that the key determinants were how good your research was and how tough the competition was.

"By publishing earlier in the process, you're signalling to rivals in the field that you're serious about the research project," Dr Gill told The Times Higher .

"This may well put potential rivals off, which is good for you, but it may have an information transmission effect where they can see what you've done and potentially catch up."

Dr Gill used a form of game theory to analyse the trade-off in disclosing intermediate research results. "There are other strands of literature that look at other reasons to disclose but I'm the first to talk about trade-off," he said.

If the initial research results are poor, his model suggests jettisoning the research project rather than investing in a second stage.

Where competition is weak, even if the results are good but not brilliant, Dr Gill's model advises disclosure in a bid to put off rivals who have not yet reached the same stage.

The risk of helping them catch up is outweighed by signalling commitment to the project.

But where a research team is leading its field, and there is strong competition, its goal is to be the only group to get to the final stage.

"In that case, I find that you shouldn't disclose at all," Dr Gill said.

Competitors already know that the leader is investing in this area, and they may be effectively deterred simply because they do not know how far behind they are. If another research team is carrying out complementary research, it is valuable for each team to have access to the other's interim findings.

Dr Gill admitted that he had had to think about whether to publish a working paper on his research before presenting it at the conference.

"I did decide to disclose because I felt the spin-off effects would be quite weak. It was such a novel approach, I didn't have too much of a follower effect."

olga.wojtas@thes.co.uk