By Brooke Struck
Would it be useful to have a tool to quickly measure the disciplinary diversity of your team? At Science-Metrix we’ve created a widget for just such a purpose. In this post, I’ll explain what the disciplinarity widget does, how to use it, how to interpret the measurements and how we are refining the tool.
How is disciplinary diversity measured?
For several years, Science-Metrix has maintained a classification of research into a three-level taxonomy, arranging research into domains, fields and subfields. We have also developed several approaches to assess the conceptual proximity of these subfields to each other, based on how often material from these subfields is used in combination.
With the taxonomy in hand, and a proximity matrix relating the subfields to each other, we can calculate disciplinary mix using a three-dimensional approach. The first dimension is simply the number of different subfields integrated, the second dimension is the balance between the subfields being represented, and the third dimension is the conceptual distance between them.
For example, a team that consists of five biologists and one chemist is considered less diverse than a team of three biologists and three chemists, because the latter team is more balanced between the subfields involved. Similarly, a team with five biologists and one chemist is considered less diverse than a team with five biologists and one performing artist, because biology and chemistry are conceptually more proximate to each other than are biology and the performing arts.
How do you use the widget?
Using the widget is intended to be very simple. Each team member needs to be tagged for the relevant subfield that they represent. In order to collect this information, the widget asks you to name your team and identify the sector in which you’re working, and it then presents you with a menu to navigate through our three-level taxonomy and identify your subfield.
Once you’ve inputted your own information, a link is provided for information to be supplied for your teammates as well—a link that you can send to your teammates, or that you yourself can click through in order to enter information on their behalf.
How do you know what subfield to associate yourself with? At Science-Metrix, we generally recommend using your highest level of education (or the degree most recently completed, as a tiebreaker). However, in some cases, people have ventured into completely new intellectual areas since finishing their studies, so it is perhaps more relevant for them to identify their new area of expertise instead.
For now, each person can only choose one subfield to represent themselves in the measurement. If you find this particularly constraining, let us know, as allowing multiple subfields per person is a feature we can consider building if this challenge is widespread.
What do the results mean?
The scores reflect:
- Number of sub-fields represented
- Balance between the disciplines represented
- Intellectual distance between the subfields represented.
Scores on this indicator range from 0 to 1, 0 being completely monodisciplinary and 1 being maximally diverse. The ranges of these scores can be interpreted as follows:
- 0 means totally disciplinary, everyone from the same background.
- 0.1–0.2 is a low score, meaning that there is one “home” discipline with a few “secondary” areas also included.
- 0.3–0.5 is a mid-range score, meaning that there is a balance amongst the disciplines represented but that they’re all still quite clustered in one intellectual area.
- 0.6 and above is a high score, meaning that several different disciplines are involved, they’re relatively balanced (rather than a “home and guest” model), and they’re drawn from a broad intellectual diversity.
Collecting widget data
We’re collecting anonymized data through the widget to see how diverse the teams are out there, what kinds of disciplinary combinations might crop up, and so forth. If we can characterise patterns broadly enough, they can contribute valuable information for users in interpreting their own scores.
The main data we collect are the disciplinary diversity scores, which we’ll be able to slice by sector (inputted manually) and by geographic area (collected via the Internet Protocol (IP) address stamped on the submission). We’ve left a space for you to input your email if you’d like to receive updates about new features built into the widget, and findings that we’ve uncovered looking at patterns in the inputted data. We designed the widget to log your email independently from your team’s data.
Widget development: Seeking feedback!
We’re currently working to improve the widget, so we’re looking for feedback! At a User Experience (UX) event in Montreal in early 2018, audience members commented on the smooth workflow of the tool, but pointed out that we’ve rolled together two distinct functions:
- Calculation: to measure the score of an existing team
- Exploration: to experiment with various permutations of potential teams.
Data collected to date bears out this remark as about half the teams entered have the word “Test” in their name.
What’s the value to you of knowing about disciplinary diversity? If you’ve tried the widget, is there something about the tool that feels clunky or unintuitive to use? Is there another tool that you’d like to see this widget integrate with, like maybe your Open Researcher and Contributor ID (ORCID) (and the ORCIDs of your colleagues)?
This blog post is adapted from a longer version “Team diversity widget: how do you measure up?” which appeared on the Science-Metrix blog: http://www.sciencemetrics.org/team-diversity-widget-measure/
Biography: Brooke Struck is a senior policy officer at Science-Metrix Inc. in Montreal, Canada, where his role includes contributing to project design and management, research and analysis, reporting, and communication with clients. He also leads research projects for the development of new indicators to measure scientific activity. These projects integrate research from the history and philosophy of science with emerging policy priorities and bibliometric innovations to ensure that new indicators are both strategically relevant to client needs and methodologically robust. Additionally, he contributes synthetic and critical assessments of science governance and policy developments through the ScienceMetrics.org blog.