Main discussion thread: Gender Diversity Index

This is the main discussion thread. If you haven't done so already, you need to logon with your GenPORT account (or register if you do not have an account) and join the discussion group.

1. Post your comments. New comments appear at the end of the page.

2. You need to refresh your browser window in order to see most recent posts by others.

3. Main discussion is scheduled for Tuesday 12th of June 2018 from 11:00 to 13:00 Central European Time. However, posts are welcome after that date as well.

4. Please introduce yourself shortly in your first post.

Comments

Elisa Guenther's picture

Welcome to everyone participating in this online discussion. In this discussion we want to (1) introduce briefly the Gender Diversity Index and (2) discuss with you its potential use as well as its limitations. We have prepared some short statements to summarise the main ideas behind the Gender Diversity Index. These initial statements are the result of a collaborative process and authored by both Anne Laure Humbert and Elisabeth Anna Guenther, even though they can only be posted by one author.

For further information, please look at our report on our project website: https://www.gedii.eu/wp-content/uploads/D3.2_GenderDiversityIndex_28FEB2018.pdf

 

We would kindly ask all participants to briefly introduce themselves, when posting something. Thank you.

 

A few words about Anne Laure and Elisabeth

Anne Laure Humbert, PhD, is a Reader in Gender and Diversity and Director of the Centre for Diversity Policy Research and Practice at Oxford Brookes University. Anne is very experienced in gender equality research at national and EU level, policy analysis and assessment as well as gender statistics. She specialises in applying advanced quantitative methods to comparative social and economic analysis, particularly in relation to work and organisations, entrepreneurship, and the integration of work and life. She has also developed several composite indicators at international level. She has worked both at EU level and in academia, with previous positions at Cranfield University and Middlesex University London. Anne is a regular public speaker on gender equality and she enjoys the opportunity to make connections between theory, practice and activism.

 

Dr Elisabeth Anna Guenther is a social scientist. Elisabeth research focusses on gender and diversity in the area of science and technology from an organisational perspective. Drawing on sociological theory, amongst others Bourdieu and intersectionality, Elisabeth applies qualitative and quantitative methods to capture complex phenomena. She presents her work regularly in academic settings and engages with practitioners and policy makers as well as in different outreach activities. Elisabeth worked, amongst others, at Cranfield University and TU Wien.

 

alhumbert's picture

The Gender Diversity Index is a composite indicator which captures gendered processes within teams. In doing so, it goes beyond counting the number of women and men. Instead, the Gender Diversity Index combines the gendered composition of a team along different diversity dimensions. To this end, two metrics are calculated for all seven identified pillars of gender diversity: representation and attrition.

In operationalising representation, parity is considered as ideal state. This means that teams with equal numbers of women and men will get a score of 1, the highest possible score. Teams that are made up of all women, or all men, get a score of 0. The metric penalises unequal representation regardless of whether it is women or men that are under-represented.

Attrition, the second metric used, captures whether the representation improves or worsens at different levels. A distinction is made between levels in relation to how desirable and/or inclusive they are. For example, it is desirable to ensure that women and men are equally represented in senior positions or as carers.

We calculated these metrics for all identified layers of gender diversity. Seven grounds of diversity were retained following multivariate analyses: age, education, care responsibilities, marital status, team tenure, contractual arrangements, seniority. These form the seven pillars of the Gender Diversity Index. Aggregation was first conducted within each pillar, before the overall aggregation. At both levels, PCA weights were used. These weights were chosen after a multi-modelling procedure in which different assumptions were tested.

The data used for constructing the Gender Diversity Index was obtained through an international web-based survey. To minimise the standard errors of estimates, only teams which met the following criteria were used to compute the Index: 100% response rate for teams with 4 respondents, 50% response rate for teams with 5 to 9 respondents, and 40% response rate for teams with 10 or more respondents. The Gender Diversity Index was computed using data for 101 teams.

In line with the COIN best practice recommendations and the OECD/JRC established methodology for constructing composite indicators (Nardo et al. 2008), the robustness of the Gender Diversity Index was assessed by examining the correlation structure between the indicators, between the indicators and the pillar/index scores, and between the pillars and the Index itself. This shows that there are some very strong correlations between some indicators within the same pillars, such as those associated with senior roles, team tenure and age. However, this is likely related to the indicators picking up on the same phenomenon. Since there is no evidence of strong collinearity among indicators not grouped together, with correlations not exceeding 0.6 and no negative coefficients, the correlation structure of the indicators is deemed satisfactory. The association of the pillars to underlying indicators shows that all indicators contribute to both the pillars and the overall Gender Diversity Index since all correlation coefficients of interest are above 0.5.

Elisa Guenther's picture

Gender diversity can have a different meaning in different contexts. For the Gender Diversity Index, we looked at gendered processes within teams along different diversity grounds. This allowed us to go beyond a sex-disaggregated analysis of groups. Instead, we wanted to capture the effects of gendered processes within teams.

In doing so, we focused on demographic and functional diversity, as there is an extensive literature on how these aspects are gendered and influence team dynamics.

Demographic diversity comprises the following indicators:

  • Age: The number of wo/men below or above team average
  • Education: The number of wo/men without or with a doctorate
  • Care responsibilities: The number of wo/men without or with current care responsibilities
  • Marital status: The number of wo/men without or with a cohabiting partner or a spouse.

 

Functional diversity includes the following indicators:

  • Contract: The number of wo/men with fix-termed or permanent contracts
  • Seniority: The number of wo/men in junior positions or in leadership roles
  • Team tenure: The number of wo/men below or above average team tenure

This approach has some limitations: in our conceptual framework, we also strived to include disability as well as ethnicity in the Gender Diversity Index. However, we could not include these dimensions, due to the low figures of people stating they are a member of an ethnic minority group or consider themselves to have a disability or chronic illness. Moreover, due to ethical concerns we did not ask questions about sexuality. This is because in an organisational context, nobody should be asked to ‘come out’. Finally, our operationalisation relies on coding gender as a binary category, although we also offered a third option “other” (two respondents selected this third option). Further work should examine to what extent all those issues could be included in the Gender Diversity Index.

Despite these limitations, we believe the Gender Diversity Index provides a useful measure of gender diversity within teams by going beyond counting heads. As the results illustrate, gender diversity is more than just parity. This is because parity within a team does not necessarily mean that decision-making power within this team is equally distributed or that the team is more inclusive.

What do you think about this? What would you see as benefits or challenges for your work?

namsor's picture

Hi !

We've been involved with several research that used scientific databases (ex. published papers, pattents, grant applications/budget approvals, etc.) which is a different approach to surveys, of course. These databases normally don't have gender information, but we enrich them with the likely gender based on personal names - with a high precision and recall. 

Do you think the two approaches (surveys vs. existing database mining) could be combined? 

That would be especially interesting to analyse both the dimensions of gender and ethnicity/cultural identification, which is an other output that NamSor provides based on personal names. The error rate for ethnicify/cultural identification is higher than for gender, but it's also more complex to estimate - so combining with surveys is a great way to confirm and estimate the precision. 

Thx,
Elian

Elisa Guenther's picture

Welcome to all participants. The floor is open to your questions and comments.

Are there any questions regarding the operationalisation?

What do you see as benefits or challenge in measuring gender diversity on the team level?

How could such a tool enrich your work?

 

 

joerg's picture

I think it would be a welcome addition to the list of indicators that are used for monitoring the implementation of Gender Equality measures. To my knowledge there are no instruments that address the GE on the team level in particular. The advantage would be that data for each of the 7 dimensions is relatively easy to obtain. Initial diagnostics of GEPs are most likely to collect these socio-demographic information, which then can be compiled into a composite score to measure inclusivity of teams but also monitor its change over time.

 

arroyo_lidia's picture

Hello, 

I am Lidia Arroyo, researcher in Gender and ICT Research Group (IN3-UOC) and part of the GenPORT team. I find your work on the Gender Diversity Index very useful for being applied in different research orgamisations to measure gender processes. It could be exportable? Are you thinking in ways to be applied in different types of research organisations? 

Lidia

 

 

Elisa Guenther's picture

Dear Lidia,

thank you for your question. One way the Gender Diversity Index can be exported to another context is a self-assessment tool. As part of the GEDII project, we developed a first tool in this direction:

https://www.gedii.eu/self-assessment-tool/

We are thinking about further developments for such a tool. Do you or any other participants have perhaps some remarks in this regad?

best wishes

Elisabeth

dimitra christakou's picture

Hello Lidia,

I am Dimitra Christakou, Memberhsip Director for WISE Campaign in the UK. I think the development of such tools is a great asset, at WISE we work closely with members that need them urgently to capture quantitative information about gender diversity within their teams. We often collabraote with other organisations in the development of tools, we would be more than happy to help you introduce the tool to our members.

 

Best Regards,

Dimitra

Elisa Guenther's picture

Dear Dimitra,

thanks for joining the discussion. It would be great, if you could share the self-assessment tool (https://www.gedii.eu/self-assessment-tool/) with your member organisations.

How do you feel about its usability for your work? Do you think this might be something usefull or do you have any suggestions that would improve it's usability?

alhumbert's picture

Dear Lidia

Many thanks for your interest in the GEDII work, and on the Gender Diversity Index. It was developed in the context of STEM research teams in this project, however, there is plenty of scope to take it outside of this context or setting.

For example, the Gender Diversity Index could be used within any other subject areas, although it is true that team-based research (where teams are organisational defined, e.g. as a research labs) are more common in STEM. However, within Higher Education but also in industry, there are many teams that could look at such a tool to measure how gender diverse they are.

Another possibility would be to use the Gender Diversity Index outside of a research and innovation setting, for example in large organisations to measure gender diversity. There has been a lot of management research looking at the link between gender diversity in teams and performance, and an adapted tool such as the Gender Diversity Index could contribute to that area.

Thanks for your interest again

Anne

arroyo_lidia's picture

Thank you very much for sharing the possibilities of the Gender Diversity Index to be used in different context and organisations. Definetively theiself-assessment tool will be a great resource for measuring gender processes within research organisation.

 

 

sandraklatt's picture

Dear all. My name is Sandra Klatt and I am with the project GEDII. I wonder, do you have maybe an example at hand, e.g. for a fictional team, how is the index being built and what does the GDI tell me as a team leader? Could I compare the index with other teams or can I monitor the progress of my team by calculating the GDI after certain periods of time again?

Elisa Guenther's picture

Welcome Sandra,

thank you for your question. Yes, team leaders - as well as equality officers or HR - can use the Gender Diversity Index. The best way to do so, is to examine both, the overall score of the Gender Diversity Index as well as the different metrics. Because, if a team leader or HR looks at the different metrics they can discover areas where they might want to implement policies. For instance, if they look at different metrics and see they score low on team tenure they might put in place some policies to increase retention of the under-represented gender group which then would increase the overall score in team tenure. So yes, the Gender Diversity Index and its metrics can be used to monitor progress over time. It can also highlight potential issues and therefore inform necessary policy steps.

 

kateclaytonh's picture

Hello - I'm Kate Clayton-Hathway, Research Fellow at Oxford Brookes with a particular interest in gender within organisational contexts. I have some thoughts about some of the aspects of the index as they relate to demographic diversity, specifically:

- care responsibilities: how were these defined, for example, are full-time caring responsibilities only taken into account (children, eldercare), or were other caring scenarios included such as part-time parental or family responsibilities?

- ethnicity and disability: do you envisage that there will there be scope at some stage to incorporate these as factors in the index?

Kate

 

 

 

alhumbert's picture

Dear Kate

For care responsibilities, this is the question asked in the survey: Do you have/ have had care responsibilities for children under 16 years of age or for dependent adults? Select all that apply. When then gave sub-options: Yes, for children under 16 years; Yes, in the past for children under 16 years; Yes, for dependent adults; Yes, in the past for dependent adults. In the Gender Diversity Index, we take into account current care responsibilities for both children and dependent adults. In the survey, more specific questions are also available, although we have no yet analysed this in greater depth.

Ethnicity and disability: including these two diversity aspects was problematic from a methodological point of view because of the very small proportion involved. In the work we did on setting up a measurement framework, it was not possible to include them. However, a case could be made that the seven pillars are sufficiently robust methodologically, and that it is worth adding a further two pillars for conceptual reasons. This will leave some questions to be answered: what score to give a team that does not have anyone with a disability in their team? Currently, this would be '0' on the basis that there is no inclusivity. However, with much smaller proportions, this debates needs to take place. I would very interested in thoughts on this.

Anne

joerg's picture

Hello, my name is Jörg Müller. I'm part of the GEDII project as coordinator. I think the Gender Diversity Index address an important lacuna in current Gender Equality policy and research. One of the most consistent global trends in science over the last decades concerns the move from individual efforts of scientists to collaborative projects among several researchers. Research is done in teams. At the same time, our current understanding and policy measures regarding Gender Equality is either geared towards the individual or happening on the organisational level. One of the main instruments of the European Commission for example consists of funding the implementation of Gender Equality plans in higher education and research performing organisations. The GE are planned and implemented on the organisational level, often neglecting what is happening on the team level.

In this sense, the GDI is an important and pretty unique contribution to make first of all visible what happens on the team level in terms of Gender Equality. I see it as a first step which now needs to become part of other initiatives and measures to see really what difference it will make to have very gender inclusive teams.

Elisa Guenther's picture

Dear Jörg,

thank you for your comment. I would agree that it is important to look at what is happening at the team level, as this constitutes the immediate working enviornment and is therefore crucial when it comes to including and retaining (new) members.

At the same time, this also poses some challenges, as teams sometimes are quite small and every change can show big effects in regard to numbers. By this I mean, it is important to not just focus on the score of the Gender Diversity Index but to also look at the different pillars/metrics and their development - when it comes to policies.

How do the other's feel about the benefits or challenges looking at the team level? Is it the logical next step for pursuing gender equality in science and technology?

 

joerg's picture

I guess, part of the question is really how you define the team. The Gender Diversity Index in its current form addresses organisational based teams, i.e. teams whose members have some form of contractual link with their institution. Although the way the GDI works is independent of the criteria of group membership - it could be calculated for any group -  I think this is important to stress. The focus on organisational teams is an advantage when considering current GE practices which clearly target the organisational level. But at the same time it remains to be seen in which way the GDI relates to wider definitions of team collaboration: for example bibliometric studies do not consider organisational team; they can't because information on team membership is not contained in the relevant databases such as Scopus. However, cross-organisational collaborations among researchers in research is almost the norm and crucial for research performance. It remains to be seen how these different levels of analysis - the organisational team and the wider, co-authorship "team" - relate to each other in terms of advancing for example science careers.

Elisa Guenther's picture

agreed, it very much depends on the definiton of teams. I think the Gender Diversity Index works good within organisational setting. For author-teams some aspects of functional diversity might not work (for instance how do you define team tenure or seniority within author teams), but demographic diversity would work. 

Elisa Guenther's picture

Thank you for your contribution to this online discussion. So far we addressed questions on the potential use of the Gender Diversity Index, for instance how team leaders and HR could use it to discover potential scope for action, or the self-assessment tools for teams to monitor their own diversity. We furthermore clarified some of the demographic variables. Moreover, we discussed the importance of team level analysis and some of its challenges. We are happy to receive any comments you might have in the future.

alhumbert's picture

Thank you so much to everybody that participated (and those reading this thereafter). Please keep sending us suggestions for how to use the Gender Diversity Index as a tool to measure gender diversity in teams, or in further analyses!

Anne

Elisa Guenther's picture

Dear participants,

please join us as well next week (20th of June 2018, 13:00 – 15:00 Central European Time) for the e-discussion on Sensor-based data for gender research following this link:
http://www.genderportal.eu/group/e-discussion-sensor-based-data-gender-r...

Content:

Research has revealed gender differences in non-verbal behavior within interactions and conversational styles. Status and gender matter during interactions. Sensor technology and computational approaches are rapidly evolving, potentially allowing for the detection and measurement of these behavioral cues. This online discussion brings together researchers working with sensor-based data in order to discuss current possibilities and challenges for the social sciences in general and gender research in particular.

The discussion is convened by Jörg Müller, Open University of Catalonia, Spain.

Add new comment

Active Threads:

2018-Jun-12

1 month 1 week ago
Posted by: joerg
This is the main discussion thread. If you haven't done so already, you need to logon with your GenPORT account (or register if you do not have an account) and join the discussion group.1. Post your comments. New comments appear at the end of the page.2. You need to refresh your browser window in order to see most recent posts by others.3. Main discussion is scheduled for Tuesday 12th of June...
Comments: 23