Women¡¯s research output appears to have bounced back in the latter part of 2020 as the lockdowns that closed schools and nurseries in many parts of the world were eased, new data suggest.

According to figures from a database of both peer-reviewed and preprint publications, the share of accepted papers with female first authors reached almost 37 per cent for articles submitted in October.

This figure was the highest proportion for any month going back to 2015, according to the analysis by Digital Science of publications in its Dimensions database, and 2 percentage points higher than for the same month in 2019.

It contrasts with earlier in 2020, when the share of female first authors fell compared with previous years in some months, something that has been attributed to women taking on a disproportionate amount of childcare during lockdowns.

³ÉÈËVRÊÓƵ

While the Digital Science data are still incomplete?owing to the lag between article submission date and acceptance, the analysis tallies with other sources that saw a pick-up in women publishing research in the latter months of 2020.

?set up by researchers following the issue that shows the share of preprints with a woman as first author suggests that three major repositories saw a rise in the autumn, particularly medRxiv, which carries articles in medicine and clinical research.

³ÉÈËVRÊÓƵ

The proportion of women first authoring articles on medRxiv fell to almost 20 per cent in April 2020, according to the data, but recovered to nearer a third in the autumn, although this was still below levels seen in 2019.

Megan Frederickson, associate professor of ecology and evolutionary biology at the?University of Toronto,??on the preprint trends, said it was possible women were ¡°playing catch-up¡± in the autumn as lockdown measures eased in many countries.

¡°In the northern hemisphere, schools reopened in September in many places, so the timing fits,¡± she added,? showing that while about 84 per cent of children were out of school on 31 March, this had fallen to 18 per cent by the end of November.

However, there were other possible explanations, she said, such as academics altering their behaviour in response to media coverage over the summer of the impact lockdowns could be having on women¡¯s research, Dr Frederickson said.

³ÉÈËVRÊÓƵ

She also pointed??of the under-representation of women in research on Covid-19, which may have contributed to the changes seen in the spring ¨C when such research was a large share of preprints ¨C and so also explain the autumn bounce-back.

Vincent Larivi¨¨re, professor of information science at the University of Montreal, one of the researchers behind the data-monitoring site, also said this was a possible factor.

The preprint data suggest ¡°that part of the reason for the decline in the percentage of women authors is that they were less likely to jump on the bandwagon and start new research projects related to Covid¡±, he said.

¡°In other words, the relative decline of women authors can also be associated with the gender asymmetry in the availability to start new research projects during the pandemic.¡±

³ÉÈËVRÊÓƵ

Researchers following the issue will now be keenly watching data for the first weeks of 2021, when lockdowns have again ramped up in parts of the northern hemisphere.

¡°Given that there have not been drastic changes in the organisation of families over the last year, any additional lockdown in the future should have similar effects¡± to the spring, Professor Larivi¨¨re suggested.

³ÉÈËVRÊÓƵ

simon.baker@timeshighereducation.com