Thursday, November 29, 2012

More Gender Bias Uncovered: Conference Symposia Organizers

Think about the last scientific conference you attended.  Did the speaker roster reflect the composition of your science society?  At a recent meeting I attended, several of us (women) were discussing the fact that there seemed to be a preponderance of male speakers, a proportion that did not reflect the gender makeup of our society.  When we pointed this out to the (mostly male) organizers, they basically pooh-poohed our concerns and waved their arms at all the female participants milling around (mostly grad students)....as if that addressed the point we were trying to make.

Turns out, there is now some evidence that the gender of session organizers has a significant influence on who ends up speaking in the more prestigious symposia sessions.

A recently published paper in PLOS One, "Stag Parties Linger: Continued Gender Bias in a Female-Rich Scientific Discipline", examined participation by women and men at conferences in a field in which women are in the numerical majority.  Based on an analysis of 21 annual meetings of the American Association of Physical Anthropologists, they found that in the subfield of primatology:

1. Women gave more posters than talks; men gave more talks than posters.

2. The proportion of female participants was lower in symposia organized by men (29%) compared to those organized by women only (64%) or by both men and women (58%).

They found the same pattern for 12 annual meetings of the American Society of Primatologists.

Primatology is a field that has had more female than male participants since the early 1970s and where women have achieved substantial peer and popular recognition for their achievements (Jane Goodall, Dian Fossey).  The results of this study suggest that even numerical dominance by females in a scientific field does not prevent gender bias. By looking at symposia participation, which is typically by invitation of the organizer, in relation to the gender of the organizer(s), the authors were able to say something about whether the underrepresentation by women at conferences might be due to lack of assertiveness by females or to bias.  Whereas posters and regular talks are self-motivated, the more prestigious symposia reflect a choice made by the organizer.  The results suggest that when men are the organizers, fewer women are invited to participate, but that men and women are equal participants in symposia organized by women (despite the larger numbers of women in the field).

Such bias may be unconscious on the part of the male symposia organizers.  I'm just guessing that this is the case as the authors have no data either way.  The authors of this paper, however, do state that they "cannot rule out" the possibility that male primatologists are more homophilic (homophily is preferential interaction with others who have similar attitudes, beliefs, or personal characteristics) than women.  The idea is that male symposia organizers have to spend more time interacting with their invitees, both before and after the symposium (e.g., to jointly prepare a conference proceedings) and simply prefer to work with other males.  Perhaps.  The jury is still out on that one.  But my guess is that this behavior is not restricted to male primatologists.

The finding that women gave more posters than talks and men the reverse is troubling and suggests the possibility that women are preferring the less visible presentation mode.  Although the study indicated that men were 12% more likely to request an oral talk over poster than women, which might explain the gender difference, the authors suggest that there still might be some bias occurring at the presentation selection/assignment stage (by the conference organizers).  Their data were pretty weak on this point, however.

In any case, this is yet another example of why it is important for women to be assertive in promoting themselves and their science.  We may need to encourage female students to give oral presentations. As for bias in inviting speakers, one solution is for more women to organize conference symposia and ensure that women are well represented in the speaker lineup.

Sunday, November 25, 2012

Using Science Networking Sites to Increase the Online Visibility of Women in Science

In this series of posts, I'm talking about various ways women in science (as well as other minorities) can increase their online visibility.  In case you are wondering why this is important, see this post, this post, and this video.  In brief, male scientists tend to have a greater online presence than female scientists.  I'm talking about personal websites, profiles on science networking sites (e.g., ResearchGate, Google Scholar Citations, Epernicus), science blogs, Twitter and Facebook (used specifically to spread the word about scientific accomplishments), writing articles for Wikipedia (and being the subject of articles), to name a few things.

It's not clear why the gender difference exists, but the lower visibility of women can have a range of negative effects....from perpetuating the idea that women cannot be successful in science...to discouraging girls and young women from envisioning a career in science.  In the past, women in science were overlooked or actively excluded by authors of textbooks, organizers of science conferences, and editors of journals. Because such platforms were controlled by others, women had no voice in promoting their contributions to science and little recourse when their work was overlooked.  Even today, major science journals discriminate against women, although apparently not consciously (see this surprising admission by the editors at Nature published 4 days ago).

Today, we have many options and opportunities to show how women are contributing to science.  The internet has helped to put us on equal footing with male colleagues in having a voice in describing our contributions to science.  We should take advantage of it, not shy away. Various efforts are in fact underway to encourage more women in science to improve their online visibility...from editing Wikipedia articles to establishing blogs chronicling their experiences in science.

Some may be reluctant, thinking that they need IT skills to create an online presence.  However, it's ridiculously easy to set up a scientific profile on a networking site, a blog, or a personal website where you can showcase your publications, your ongoing science projects, and your research group.  If you've got a Facebook page, you already have the know-how to set up a profile on ResearchGate, Google Scholar Citations, or similar sites.

In the last post, I described Google Scholar Citations as one place to establish an online profile.  In this post, I'll focus on ResearchGate, which is designed like a social networking site....for scientists.  Their website claims over 2.2 million members. On their homepage, the interactive piechart indicates that medicine and biology have the most members and the most publications.  To see how the site works, you are forced to create an account.  So, I'll give you a brief tour that will give you a peek inside without having to sign up.

Once you log into ResearchGate, it automatically identifies publications that you've authored and asks you to check off those that are yours.  Once it identifies you, ResearchGate then searches for other pubs as well as the full-text files and uploads them if available.  You can add a photo and other information to your profile.  Your main contributions page will look like the screenshot below (using the top-cited ecologist from the previous post as an example):



At the top of the page is your "RG Score", which is calculated "based on how other researchers interact with your content, how often, and who they are. The higher their score, the more yours will increase." Selecting the RG Score will take you to a page with additional analytics of your interactions and how you compare to other ResearchGate members. 



Below the profile are two tabs: contributions and info.  Your publications are listed under the contributions tab.  To the right is a running tally of total publications and numbers of articles, books, book chapters, conference proceedings, and other items.  Below that is a listing of followers and people you are following.  Your co-authors are also listed at the bottom of the page. 

Selecting the pub title or the document image will open up another page (below) where additional information is presented about that article, including the abstract, full-text article (if available), datasets, additional text files, and media.  You can upload a pdf of your published article...if you have permission to do so.  ResearchGate seems to add these automatically if they are available somewhere on the internet.  On my profile, I've additionally uploaded "research briefs" (described previously) for each publication; these can be viewed full-screen online or downloaded (and you can see on your analytics page how many times your pubs are viewed or downloaded). 



Topics for discussion are listed under the Info tab; here, people can ask questions and discuss topics.  There are general topics such as "climate change" and "botany", which thousands of people are following, but you can add specialized topics of interest to you.  Anyone can post a question, and others reply, providing, in some cases, links to relevant articles or simply their opinion.  You can see examples of questions in the screenshot below.



Another option is that members can start a "project" with "benches", which allows you invite others to share files and post comments...this would be convenient for a research group with members scattered around the world.

Overall, I found ResearchGate easy to navigate and to upload and organize my publications and supporting materials.  If you prefer your profile to be private and only interact with a select set of colleagues, you can do this by changing the privacy settings.  ResearchGate differs from Google Scholar Citations in that it quantifies a scientist's reputation based on a combination of factors, including publications (and their impact factor) as well as how many others are following that scientist and their interactions with others with high RG scores.  This aspect is interesting because it means that someone with fewer pubs or lower-impact pubs, e.g., early career folks, can raise their RG score by being more social and networking with colleagues, something that women often excel at.  If you're already skilled at socializing on Facebook, then you'll likely be at an advantage here.  Several of my colleagues have profiles on ResearchGate, and from what I can tell, the site has a more equal representation of women (about 48 % according to this reference) compared to Google Scholar Citations.  I also like the option of uploading full-text articles, media, and datasets to share with others as well as a way to track views of your profile or pubs.

On the downside, ResearchGate does not fully explain how they calculate the RG Score (or the contribution "total impact")...at least I couldn't find it.  Also, the emphasis on ranking scientific standing promotes competitiveness, which is can be off-putting, especially to women.  ResearchGate also has been criticized for sending spam to members' coauthors (inviting them to join). You are also encouraged to invite your co-authors to join.  But that's how Facebook and other social networking sites attract members; all you have to do is ignore the invitation if you receive one and don't want to join.  If any of your co-authors has a profile and articles in ResearchGate, however, then you're already in their system, and anyone searching for your name will likely find a page listing some of your publications.

One additional worry might be that such sites will end up being another online time-sink (as Facebook is for some).  I suppose that might be a possibility, but it doesn't seem likely (for me at least).  One could choose to establish a profile on a site like ResearchGate and only periodically visit to add new publications, much like updating a CV.  But for those who like to socialize or want to expand their professional network, this type of site might be a good option.

In addition to ResearchGate, there are other science networking sites such as linkedin.com, and epernicus.com.  Some people have profiles on multiple sites, which increases their overall visibility.  Of course, there are other ways to increase the online visibility of women in science.  In coming posts, I'll talk a bit about those and my experiences with them.   


Wednesday, November 21, 2012

How to Increase Visibility of Female Scientists

In this series of posts, I'd like to focus on how women in science can increase their online visibility.  This strategy was a topic of discussion at the SpotOn London 2012, held November 11.  In the last post, I listed some of the resources SpotOn was gathering to assist women in improving their online science profiles, which are here, in case you missed it.

One of the simplest ways to create a professional online presence is to develop profiles on websites such as ResearchGate, Google Scholar Citations, and Epernicus.  Although someone can have a website or webpage that summarizes their professional accomplishments, these networked sites provide some options and advantages that a personal website cannot.  They are similar to social networking sites in that others can "follow" a specific scientist whose work is of interest. If you want to keep up with what the lead scientists in your field are publishing, you can "subscribe" to their profiles on Google Scholar or ResearchGate and they can subscribe to you. Subscribers are notified of new articles.  If you have a profile, you will be notified of new citations to your published papers.  ResearchGate offers the option to post full-text articles and supplementary information alongside each citation, which a follower can read online or download.

I have profiles on the first two and can describe how they work.  In this post, I'll focus on Google Scholar Citations.

Google Scholar Citations (GSC) is a fairly recent offering (2011), which is described as "a simple way for authors to compute their citation metrics and track them over time".  GSC is more than a tool for tracking one's citations, however.  Authors can create public or private profiles with options to show a photo, primary affiliation, email address, key words, and links to their professional websites.  Below that is a summary of citations, which shows the total number of citations your publications have received (all time and previous 5 years), your h-index, and your i10 index (the h-index is the largest number h such that h publications have at least h citations; the i10 index is the number of publications with at least 10 citations).  There is a summary graph showing numbers of citations per year over your publishing career.  Below that is a list of all your publications, each one showing the title and citation details, year published, and number of times cited.  The title is hyperlinked to a more detailed page showing the pattern of citations for that publication over its lifespan.  The number of citations is hyperlinked to a search page listing all sources that cite that publication.

As I've described previously, GSC does a good job of identifying publications belonging to a specific individual.  However, it's essential to carefully examine all publications ascribed to you to ensure they are correct.  Any incorrect entries are easily deleted, however, and missing publications can be added to your profile. On the downside, citation metrics are notoriously prone to error and various statistical biases.  The search option by discipline results in a listing of author profiles ranked by total number of citations, which encourages comparisons based on a single number (total citations).  See the screenshot below for the first ten profiles in the field of ecology:



Notice they are all male.  In fact, there seemed to be fewer women with online profiles in GSC than men, even in fields where women are well represented.  I conducted an in-depth analysis to see if my impression was correct. I searched for all individuals with public profiles in GSC who listed ecology as their field of expertise (search term = label:ecology).  I scanned through all 1700 entries starting with those individuals with the highest number of citations and counted all male and female entries (note that there is a third category (indeterminate) for individuals whose gender could not be conclusively ascertained from their photo or name).  The following breakdown shows the absolute number (top panel) and proportion (bottom panel) of male vs. female profiles by citation category (ranked from lowest to highest).



Out of 1,700 profiles for ecologists with public profiles in GSC, only 332 women are listed (19.5%).  In the top 100 authors (i.e., people with the most citations), there is only one female; there are no females in the group with >10,000 citations.  However, the proportion of females increases with decreasing numbers of citations.  In the group with fewer than 100 citations, including students and postdocs, the proportion of females is 41%.

It's not possible to determine why there are so few females with profiles on GSC.  Ecology is a discipline that has many female participants.  The Ecological Society of America, for example, has a 40% female membership (greater than 50% for age groups 21-30 years old). The profiles for physics, mathematics, astronomy, zoology, botany, and other sciences were similarly skewed toward the male gender (especially at the highly cited end of the spectrum).  In some cases, the 50 to 100 top-cited scientists were all male.

The impression one gets is that there are few females impacting science. This is certainly not true.  It's possible that women are less inclined to post public profiles on sites such as GSC.  Perhaps males are more inclined to compare themselves to their competitors using citation metrics, especially if they've totted up a high citation ranking.  Perhaps papers authored by women are less frequently cited (not supported by analysis of ecology papers, however).  Perhaps men tend to submit papers to higher impact journals that are more likely to be cited.  Whatever the reason, the paucity of female profiles on such sites adds to the overall impression that women are not major players in science.

In the next post, I'll look at ResearchGate, which is quite different in its emphasis compared to GSC.

Friday, November 9, 2012

Women in Science Resources

The SpotOn Blog is developing a toolkit of useful resources for female scientists wanting to raise their online profiles (note that some of these resources are useful for male scientists as well).  This effort is a really good one because it encourages women to get their names, their work, and their stories online.  As I've tried to emphasize on this blog, women must be their own advocates and make sure their voices are heard.  The Web is a great place to promote your accomplishments, advertise your skills, and generally make yourself more visible.

The resource pack has links organized into the following categories:

Important websites where every female scientist should have a profile

Other websites where it could be interesting for a female scientist to have a profile/be listed

Women in science on Twitter

Good tips about science communication

Where to create my blog

Good places to get a post hosted

Online communities

Everyone is invited to add links to useful blog posts, websites, twitter lists, and other online resources.