Picking a pithy title for your research paper could significantly improve its impact, new analysis suggests.

While many researchers might not worry too much about finding a memorable or eye-catching title for a new journal paper, a recently published study indicates that a longer and more convoluted title could seriously harm its chances of doing well.

According to an examination of the titles of around 155,000 journal articles submitted to the UK¡¯s most recent research excellence framework (REF), papers with briefer titles tend to be cited more frequently.

Across 10 of the 11 units of assessment covering 36 disciplines in the 2014 REF, ¡°citations significantly decline with title length¡±, according to the paper, ¡°An analysis of the titles of papers submitted to the UK REF in 2014: authors, disciplines, and stylistic details¡±, by John Hudson, professor of economics at the University of Bath, in a recent edition of the journal Scientometrics.

ÖйúAƬ

That effect was more marked in certain subjects, including clinical medicine, biology and physics, and less so in disciplines such as computer science and economics, according to the study.

¡°A longer title generally means a paper is less likely to receive citations,¡± Professor Hudson told Times ÖйúAƬ.

ÖйúAƬ

Using a question mark in a journal paper¡¯s title reduces the number of citations it received, he added.

However, using a colon tended to improve the citations received by a paper, Professor Hudson said.

¡°With a colon you have to divide up the title into two parts, so you have two messages for readers,¡± he said, noting that a dual title could broaden the appeal of a paper.

Overall, the discipline with the longest journal titles on average was public health, whose titles averaged 117 characters including spaces, followed by clinical medicine (113) and agriculture (110). The shortest titles tended to come in social sciences and humanities, such as economics (66 characters), Classics (69) and English language (74).

ÖйúAƬ

Political scientists and law academics were most likely to use question marks in their journal titles (almost a fifth of papers in these disciplines used them) but these were used by less than 1 per cent of papers submitted in maths and electrical engineering.

The average number of authors per paper differed significantly between disciplines; physics papers had 131 co-authors on average, but for history, philosophy and theology the average was 1.1 in each case, the study found.

Consistent with previous studies, those papers with multiple authorship tended to receive higher citation rates, but numerous authors also led to longer titles, Professor Hudson said.

Those longer titles may be the result of negotiation between numerous authors, he suggested.

ÖйúAƬ

¡°If they are compromising on titles, are they compromising on other aspects of research?¡± Professor Hudson said.

Disciplinary variation in the number of authors per paper also highlighted how it had been much more difficult for academics in certain disciplines to generate the four pieces of work required for the 2014 REF ¨C a factor policymakers might wish to consider when finalising future research audits, he added.

ÖйúAƬ

jack.grove@tesglobal.com