Closing the Leadership Gender Gap in European …...Closing the Leadership Gender Gap in European Athletics Project number EAC/21/2009/145 FINAL REPORT Submitted by: Royal Dutch Athletics

Closing the Leadership Gender Gap in European

Athletics

Project number EAC/21/2009/145

FINAL REPORT Submitted by: Royal Dutch Athletics Federation Papendallaan 60, 6816 VD, Arnhem Postbus 60100, 6800 JC, Arnhem [email protected] www.atletiekunie.nl

- Page 2 of 50 -

Table of Contents

Introduction .................................................................................................................................... 5

Part I ............................................................................................................................................... 6 Work Package 1 – Survey ............................................................................................................ 6

Objective ......................................................................................................................................... 6 Introduction .................................................................................................................................... 6 Content and implementation ......................................................................................................... 6 Results............................................................................................................................................. 8 Follow-up ...................................................................................................................................... 15 Recommendations ........................................................................................................................ 15

Work Package 2 – Awareness Training for the Athletics Population ........................................... 17 Objective ....................................................................................................................................... 17 Introduction .................................................................................................................................. 17 Content and implementation ....................................................................................................... 17 Results........................................................................................................................................... 19 Follow-up ...................................................................................................................................... 22 Recommendations ........................................................................................................................ 22

Work Package 3 – Awareness Training for Federation Presidents and CEOs ............................... 23 Objective ....................................................................................................................................... 23 Introduction .................................................................................................................................. 23 Content and implementation ....................................................................................................... 23 Results........................................................................................................................................... 24 Follow-up ...................................................................................................................................... 25 Recommendations ........................................................................................................................ 26

Work Package 4 – Skills Training for Future Women Leaders ..................................................... 27 Objective ....................................................................................................................................... 27 Introduction .................................................................................................................................. 27 Content and implementation ....................................................................................................... 27 Results........................................................................................................................................... 31 Follow-up ...................................................................................................................................... 34 Recommendations ........................................................................................................................ 35

Concluding Remarks ...................................................................................................................... 36 Part I - Survey: ............................................................................................................................... 36 Part II - Gender awareness interventions: .................................................................................... 36

Appendix A – Survey homepage .................................................................................................... 38

Appendix B – Sample press coverage of the survey ....................................................................... 39

Appendix C – Text of an e-mail from a Member Federation Communications Manager (female) ... 40

Appendix D – Perspective taking exercise from the awareness training for the general athletics population. ................................................................................................................................... 41

Appendix E – Gender quiz from the awareness training for the general athletics population. ........ 42

Appendix F – Diversity quiz from the awareness training for the general athletics population. ...... 44

Appendix G – Volcano forces change to CEO Conference plans ...................................................... 46

Appendix H – Future Women Leaders Seminar Programme ........................................................... 47

- Page 3 of 50 -

List of Partners

Partner Representative Level

Dutch athletics federation Atketiekunie

Rien van Haperen [email protected]

II

Austrian athletics federation Österreicher Leichtathletik-Verband

Helmut Baudis [email protected] I

Danish athletics federation Dansk Atletik Forbund

Jakob Larsen [email protected]

I

Estonian athletics federation Eesti Kergejõustikuliit

Sirje Lippe [email protected]

I

French athletics federation Fédération Française d’Athlétisme

Jean Gracia [email protected]

II

Maltese athletics federation Malta Amateur Athletic Association

Anthony Fava [email protected] I

Slovenian athletics federation Atletska Zveza Slovenije

Boris Mikuz [email protected]

I

Spanish athletics federation Real Federación de Atletismo

Jose Luis De Carlos [email protected]

II

Swedish athletics federation Svenska Friidrotsföbundet

Mikael Peterson [email protected]

I

European Athletics European Athletic Association

Bill Glad [email protected]

II

University of Lausanne John Antonakis

[email protected] II

Note: The key tasks of Level I partners were to promote a) the online survey and b) the awareness training intervention to the general population of the sport. Moreover, Level I partners are also committed to using the tools developed through the project in the future and to making special efforts to a) disseminate the results of the project within their countries through their websites and publications and through contacts with other national-level organisations and b) supporting an ongoing European network for exchange information on matters related to the project. In addition to the commitments of Level I partners, Level II partners a) participated in the delivery of one or more project activities and b) provided staff time for project assistance. In addition, the University of Lausanne will a) use data gathered as the basis for scientific papers and b) conduct the follow-up studies that are not included in the current project.

- Page 4 of 50 -

List of Main Contributors

Name Partner Affiliation Role

Rien van Haperen Atletiekunie Project Leader

Dr. Philippe Jacquart Atletiekunie Researcher Project Manager Leadership Skills Trainer

Prof. Dr. John Antonakis University of Lausanne Scientific Consultant Leadership Training Consultant Leadership Skills Trainer

Prof. Dr. Raphael Lalive University of Lausanne Scientific Consultant

Prof. Dr. Christian Zehnder University of Lausanne Scientific Consultant

Prof. Dr. Jean-Philippe Bonardi University of Lausanne Leadership Training Consultant

Marika Fenley University of Lausanne Leadership Skills Trainer

Dr. Jane Khayesi University of Lausanne Leadership Skills Trainer

Bill Glad European Athletics Project Assistant

- Page 5 of 50 -

Introduction

The Closing the Leadership Gender Gap in European Athletics project focuses on gender equality in the sports of athletics. Generally, we wanted to see if we could create effective tools that could be used by the sport to alter the culture of its community and thereby improve the conditions for the emergence of more women leaders at the local, national and international levels. We believe that if available such tools could also be used by other sports and thereby make sport a significant driver in promoting gender equality in the whole of European society.

This project was conducted from January 1st, 2010 to March 31st, 2011. It was led by

Atletiekunie—the Royal Dutch Athletics Federation—with a consortium of initially 10 partners (the athletics federations of Austria, Denmark, Estonia, Malta, Slovenia, Spain and Sweden plus the Swiss-based European Athletics and University of Lausanne), which grew to 11 during the project (with the French athletics federation joining later).

This project comprised two main parts. The aim of Part I was to document the

attitudes (and biases) of individuals within the sport of athletics towards women in general, and towards woman leaders in particular. The aim of Part II was to test the efficacy of three interventions: i) a bottom-up intervention in the form of a gender awareness training for the general athletics population, ii) a top-down approach through a gender awareness training for federation leaders (presidents and CEOs) and iii) a targeted intervention designed to provide leadership skills to potential future women leaders. With the knowledge and understanding gained from the results of these tests we will be better able to develop the tools for altering the sport’s culture.

In the following pages we describe the content and implementation of the different

work packages related to parts I and II of the project. We then present and discuss the results of these different work packages. Finally, we make recommendations for future work to support organizations in athletics and other sports.

- Page 6 of 50 -

Part I

Work Package 1 – Survey

Objective

Design and conduct an online survey to measure attitudes and awareness in the athletics population.

Introduction

To change the culture of a sport or any other group in a systematic way it is important to have indicators that describe that culture. For us, the key points of interest in the culture of athletics in Europe are the attitudes within the athletics community towards women and women leaders in the sport. In order to obtain a picture of these attitudes, we developed an online, multi-language athletics population survey from existing tools used to measure explicit and implicit forms of sexism.

We anticipated that the results would have strong policy implications and serve as a guide for designing our interventions and tools. In particular, the figures obtained were essential for testing the efficacy of the Awareness Training for the Athletics Population Tool that we developed as a part of the project (see Work Package 2).

Furthermore, we can use the figures we obtained as a baseline to compare the athletics population (and sub-groups within the population) with other sports and with society in general, to assess the efficacy of policies and interventions and to measure progress or lack thereof towards a culture of gender equity.

Content and implementation

The online, multi-language athletics population survey measures sexism using two scientifically validated instruments:

i. The Gender-Career Implicit Association Test (IAT)1: Measures automatic associations held in memory (predicting unconscious bias and implicit discrimination);

ii. Ambivalent Sexism Inventory (ASI)2: Measures hostile and benevolent attitudes towards women.

We also included questions from the European Value Survey (EVS)3 measuring attitudes

towards women in society and towards family. This set of questions will allow researchers from the University of Lausanne to assess the extent to which the surveyed population is

1 Greenwald, Anthony G.; McGhee, Debbie E.; Schwartz, Jordan L.K. (1998). Measuring Individual Differences in Implicit Cognition: The Implicit Association Test, Journal of Personality and Social Psychology 74 (6): 1464–1480 2 Glick, P., & Fiske, S. T. (1996). The Ambivalent Sexism Inventory: Differentiating hostile and benevolent sexism. Journal of Personality and Social Psychology, 70, 491-512. 3 www.europeanvaluesstudy.eu

- Page 7 of 50 -

representative of their country’s general population (for the purpose of scientific research). These analyses will be conducted by October 1st and the results will be communicated to the project partners and to the European Commission.

In order to maximise the chances that the survey would be well received by the

athletics community, we included a number of customised athletics-related questions. These served to ground the survey in issues relating not only to gender but also to athletics. Indeed, a first version of the survey with only gender related questions was criticised by some because it was perceived as unrelated to athletics.

The inclusion of the IAT in a web-based multi-language survey presented a number of technical challenges, which were met by running the IAT4 simultaneously on multiple online platforms and by developing a custom questionnaire as a substitute to the IAT for Macintosh users.

The survey was developed using an online survey application5 provided by the

University of Lausanne. The University of Lausanne also hosted the surveys on their own servers. These services were contributions to the project. A screen shot of the survey homepage is shown in Appendix A.

The survey was promoted by the project partners as well as by a number of other European Athletics Member Federations who enthusiastically supported the project. Information about the survey was communicated through press releases (see Appendix B for sample press coverage), mailing lists, and on the partner federation websites. Both Atletiekunie and European Athletics created permanent pages on their websites to provide information about the survey and the overall project. Furthermore, Dr Sylvia Barlag, a vice-president of Atletikunie and a member of the European Athletics Council, was named Project Ambassador to further promote the project and used her contacts throughout the sport.

A number of incentives were offered to ensure participation in the survey and the Awareness Training for the Athletics Population (Work Package 2). The French athletics federation and European Athletics offered a chance to win two all expenses paid trips for two people to the March 2011 European Athletics Indoor Championships in Paris (flights, accommodation, VIP tickets) to all participants completing the survey (and the awareness training for the athletics population—i.e., Work Package 2) by February 1st, 2011. European Athletics further contributed by offering secondary prizes in the form of merchandise. A second wave of incentives was proposed to further foster participation. This time European Athletics offered a chance to win an all expenses paid trip for two to the 2011 summer's European Team Championships to everyone who completed survey (and the awareness training for the athletics population) by March 31st, 2011.

A delay caused by the eruption of the Eyjafjallajökull volcano in Iceland prevented a meeting to brief the project partners on their role in this work package meant that the survey was launched later than originally planned. This did not affect the data gathering

4 www.millisecond.com 5 www.limesurvey.org

- Page 8 of 50 -

process but it meant that the original incentive prize offer of trips to the July 2010 European Championships in Barcelona had to be substituted with prizes mentioned above.

In addition, some unforeseeable difficulties were encountered during the

implementation of the survey, which hindered the collection of data from certain countries. For instance, the Estonian athletic federation was unable to promote the survey because of a major change in their web servers and the French athletic federation was unable to promote the survey because of its late entry as a partner in the project. The result is that the number of participants in these two countries was lower than we planned for these two project partners. As described in our follow-up plan below, we will be collecting additional data from these two countries before the end of the year.

Results

A total of 2’068 individuals took part in, and completed, the athletics population survey. We first present demographic information about participants and then present information on the levels of sexism.

Participants reported coming from a broad array of countries. The Figure 1 graphic represents the distribution of participants by country. Surprisingly, a small number of individuals based outside of Europe participated in the survey. Given that there was no promotion targeting these individuals, we can only speculate that they learned about the survey from the partner websites or heard about it through word-of-mouth.

Athletes represent more than half of the respondents (54%). Coaches come in second

position in terms of largest group with 20%. Figure 2 displays the distribution of respondents by role in the sport.

In terms of age and gender, we have broad representation of the population. Figure 3

presents the distribution of respondents by age group. Men are slightly over-represented in this sample with respectively 55.66% vs. 44.34% of total respondents.

- Page 9 of 50 -

Figure 1 – Distribution of respondents by country

Note: Countries are listed in the legend by decreasing number of respondents (countries with less than 37 respondents are not listed in the legend).

Figure 2 – Distribution of respondents by role in athletics

- Page 10 of 50 -

Figure 3 – Distribution of respondents by age and gender

We now present and discuss a statistical picture of attitudes towards women and

women leaders in the athletics community across countries. We report data obtained from 16 EU member states for which we have more than 10 observations. Additionally, we report data obtained from 9 non-EU member states (8 of which are nevertheless members of the European Athletics Federation).

Figure 4 presents the mean levels of explicit sexism as measured by the ASI in the athletics community by country in decreasing order of magnitude. These results are in line with what the existing literature would suggest. For instance, northern European countries in general, and Scandinavian countries in particular, tend to have more gender egalitarian values and this is also what our data indicates, with Denmark, the Netherlands, and Sweden averaging the lowest values of sexism. Previous research6 has reported levels of sexism for some of the countries included in our sample (i.e., Italy, Spain, Germany, the Netherlands). It is noteworthy that we report lower levels of sexism for these countries. However, the results from such comparisons should be interpreted with caution as the data collected in previous research was not derived from representative samples of the population. Thus, the differences we observe might result from differences in terms of the population sampled. For example, because men are on average more sexist than women, the proportion of men included in a given sample should affect the overall sample level of sexism. Nevertheless, the data we report here has great value as it can serve to monitor the evolution of sexism over time in athletics and help to assess the effectiveness of interventions and programs designed at reducing sexism.

6 Glick, P., S. T. Fiske, Mladinic, A., Saiz, J. L., Abrams, D., & Masser, B. (2000). Beyond prejudice as simple antipathy: Hostile and benevolent sexism across cultures. Journal of Personality and Social Psychology, 79, 763-775.

- Page 11 of 50 -

Figure 4 – Mean levels of explicit sexism in the athletics community by country

The Ambivalent Sexism Inventory conceptualises sexism as comprising two components: hostile sexism and benevolent sexism. Hostile sexism represents what is traditionally thought of as prejudice, while benevolent sexism represents attitudes that, while being positive in the mind of the perceiver, are nevertheless sexist in that they carry and convey stereotypical representations of women. Figure 5 shows the levels of these two facets of sexism in the athletics community per country. It is noteworthy that, with the exception of Germany and Italy, hostile sexism prevails over benevolent sexism. This result is congruent with existing research on sexism and reflects the fact that across cultures, the role of women has always carried less social value. Consequently, policies and tools aimed at reducing sexism should in priority focus on changing these hostile attitudes towards women.

Figure 5 – Mean levels of hostile and benevolent sexism in the athletics community by country

- Page 12 of 50 -

Table 1 – Mean levels of explicit sexism in the athletics community by country

Country Sexism7 Benevolent

sexism Hostile sexism

Number of respondents

Austriaa, b 3.04 3.07 3.01 37

Belarusb 3.59 3.15 4.09 68

Bulgariaa, b 3.33 2.99 3.72 77

Croatiab 2.99 2.70 3.32 48

Denmarka, b 2.39 1.98 2.86 140

Estoniaa, b 3.13 2.75 3.57 13

Francea, b 2.89 2.50 3.33 12

Germanya, b 3.39 3.46 3.33 306

Great Britaina, b 2.63 2.17 3.14 46

Hungarya, b 3.02 2.77 3.31 103

Icelandb 2.42 1.80 3.13 21

Irelanda, b 3.10 2.81 3.42 78

Israelb 2.63 2.49 2.79 110

Italya, b 3.36 3.46 3.26 106

Maltaa, b 2.97 2.59 3.40 75

Netherlandsa, b 2.36 1.99 2.78 198

Norwayb 2.01 1.36 2.75 19

Polanda, b 2.98 2.60 3.41 106

Russian Federationb 3.50 3.03 4.03 73

Serbiab 3.13 2.80 3.52 40

Sloveniaa, b 3.10 2.79 3.45 97

Spaina, b 2.73 2.80 2.66 336

Swedena, b 1.98 1.48 2.56 94

Ukraineb 3.56 3.08 4.11 101

United States 2.88 2.61 3.18 14

Note: Countries with less than 10 respondents are not shown here. This data is available on request. Item relating to the ASI scale were rated on a 1 to 6 rating scale. a European Union member state. b European Athletics Member Federation.

Furthermore, we find that these sexist attitudes are negatively associated with the prevalence of women leaders. Indeed, we find statistically significant associations between these measures of explicit sexism and variables measuring women’s empowerment in Europe in general and in the sport of athletics in particular. We considered two measures of gender empowerment. The first is the United Nation’s Gender Empowerment Measure (GEM) which measures the degree of participation of women in political and economic forums. The second measure is the percent of women executives sitting on boards (Council,

7 Sexism was measured using the Ambivalent Sexism Inventory (ASI)—see footnote 2.

- Page 13 of 50 -

Executive Board, etc) of European Athletics Member Federations. Zero-order correlations between explicit sexism and these measures of gender empowerment are presented in Table 2.

A cautionary note is here warranted: these correlations measure association and do not imply any causality. It is however noteworthy that these associations are in the direction we would expect to find with one exception. We find a positive association between country levels of hostile sexism and the percentage of women executives sitting on the boards of European athletic federations. We leave it to our partner, the University of Lausanne, to establish in future scientific studies the nature of the relationship between these variables.

Table 2 – Correlations between country levels of sexism and gender empowerment

United Nations' Gender Empowerment Measure

Percentage of women executives sitting on the boards of European Athletics Member

Federations

Sexism -.54† -.44

Benevolent sexism -.34 -.91**

Hostile sexism

-.65** .90*

Note: Standard errors clustered at the country level † p < .1, * p < .05, ** p < .01

We now report and discuss the (implicit) measures of sexism we collected using the implicit association test (IAT). The number of observations for this measure is smaller (i.e., 540) than the number of observations collected for the ASI, and this is for several reasons. Firstly, the IAT could not be used for all respondents as it does not support Macintosh operating systems, and could not (at the time we conducted the survey) support all alphabets. Secondly, in its present form, we found the use of the IAT to be problematic when using computers with little internal memory. In order to correct for these problems, we had to pull down the IAT from the survey on two occasions. Thus, on these two occasions participants taking part in the survey did not complete the IAT. Coincidently, it is during these two occasions that participants from Austria took part in the survey, which is why we do not report any data on the IAT for Austria. Country mean levels of implicit sexism are displayed in Table 3 along with the number of respondents.

- Page 14 of 50 -

Table 3 – Mean levels of implicit sexism in the athletics community by country and number of respondents

Country IAT N

Bulgaria 0.209 33 Denmark 0.162 58 Germany 0.158 107 Hungary 0.119 43 Ireland 0.330 36 Italy 0.103 8 Malta 0.053 12 Netherlands 0.136 87 Poland 0.077 54 Russia 0.163 10 Slovenia 0.127 31 Spain 0.290 4 Sweden 0.112 43 Ukraine 0.149 14

Note: Greater values indicate stronger stereotypical associations in the sense that “men” were more easily associated with positive attributes than were “women” (i.e., indicating greater levels of implicit sexism).

Figure 6 – Mean levels of implicit sexism in the athletics community by country

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

- Page 15 of 50 -

As we noted above, the numbers we report for explicit sexism are lower than those reported in previous research (this concerns Italy, Spain, Germany, and the Netherlands). However, because the numbers reported by previous research were based on samples non-representative on the general population, any comparisons should be made with caution. No country level data has ever been published on implicit levels of sexism, thus comparisons cannot be made here between our data and previous data. However, our data can serve as initial reference point for future surveys studies of implicit sexism.

Finally, we note that at this time it is not possible to compare sexism in athletics with sexism in other sports, again because of a lack of published data. However, should data on levels on sexism in other sports become available, such comparisons could easily be made.

Follow-up

As mentioned above, the research team of the University of Lausanne will conduct a supplementary survey in the two project partner countries, France and Estonia, where the number of survey participants was less than originally hoped.

Furthermore, the University of Lausanne will conduct analyses of the European Values Survey by October 1st to assess the extent to which our sample is representative of the general population. They will then communicate the results to the project partners and to the European Commission.

Recommendations

1. We strongly recommended conducting similar surveys in the future in order to study the evolution of sexism in athletics throughout Europe. Indeed, the information collected over time through these surveys would allow decision makers to optimally decide when and where to implement interventions to reduce sexism. Furthermore, regularly conducting such surveys would also allow assessment of the effectiveness of these interventions.

2. Conduct similar surveys in other sports for the same reasons mentioned above and to make a comparison between sports possible. This would assist with the identification of sport specific factors in sexism in sport and with assessing the effectiveness of interventions

3. On the practical side, implementing a multi-language, multi-platform (with one php based interface and one custom programmed interface in the case of the IAT) has proven to be very challenging. We strongly suggest that future research efforts willing to take a similar path ensure have the necessary specialized support to implement this type of survey.

4. Another challenge associated with this survey was to balance the content between the scientific and political requirements within a sport community. A first version of the survey designed solely with the scientific aim in mind was found to be too remote

- Page 16 of 50 -

from athletics for a number of participants. Moreover, certain questions8 had to be removed from the survey because some federations found them inappropriate. One federation (not a partner) found the whole line of questioning and the general idea of the survey so objectionable that it refused to use its channels to communicate to the athletics population in its country (see Appendix C). It is noteworthy that i) the removal of the questions did not directly affect the project; ii) these questions have been asked thousands of times throughout Europe as they are part of the European Value Survey, which is a cross-national, large-scale, and longitudinal survey research program on basic human values. Our recommendation is that future researchers ensure the aim and value of surveys to the sport as a whole are well communicated and agreed with those who will be involved. The researchers need to be aware that this task requires a certain time and resource allotment.

8 e.g., “Which of these statements comes closest to your beliefs? (a) There is a personal God; (b) There is some sort of spirit or life force; (c) I don’t really know what to think; (d) I don’t really think there is any sort of spirit, God or life force”.

- Page 17 of 50 -

Part II Work Package 2 – Awareness Training for the Athletics Population

Objective

Design and test an online training task to increase awareness and change attitudes towards women and women leaders.

Introduction

We believe that raising awareness and challenging prejudices can lead to a change in individual attitudes and behaviour. Moreover, if the attitudes of enough individuals in a given group change, the culture within that group will change. Our aim in this work package was work towards a tool that could be used cost effectively to raise awareness with sub-groups in the athletics population, including athletes, coaches, competition officials, club leaders, federation leaders, etc.

The Internet seems to present an opportunity in this area. However, conducting awareness training online represents a major challenge in that the setting inhibits the use of traditional methods like role playing, group discussions, and so forth. On the other hand, if it could be developed, an effective online tool would provide the opportunity to reach a greater number of a sport’s population and thereby impact the culture of the sport more quickly.

The awareness training for the athletics population of this work package was designed as an online scientific experiment. To the best of our knowledge, this is the first time an intervention for this purpose been done this way. Only through this procedure can we rigorously test for the effects of the training we are trying to provide, and without it we would be only left speculating about its effectiveness.

Content and implementation

All individuals who took part in the athletics population survey (Work Package 1) were automatically contacted to take part in the awareness training. Participants who had completed this task were then invited to go online a third and last time, during which we measured their levels of prejudice towards women so as to test the effectiveness of our awareness training. In order to ensure participation, only those participants who completed the athletics population survey (Work Package 1), the Awareness Training for the Athletics Population and the follow-up measurement were entered in the draw and had a chance of winning one of the incentive prizes offered by the French athletics federation and European Athletics.

For this awareness training, we used two primary tools:

i. a perspective taking exercise; ii. a series of questions and answers relating either to gender or diversity (or both).

- Page 18 of 50 -

Participants were asked to read a story in which they had to imagine they had landed on an alien planet along with fellow humans and had to get accustomed to the social rules prevailing in this new environment. The aim of this perspective taking exercise was to put individuals in a situation in which they were faced with situational constraints similar to those with which women are faced when they seek leadership positions. In one version of the story, the main characters, a human, tried to climb the leadership ladder in an alien organization, is ridiculed during a meeting for wanting to act in a tough and decisive manner. In another version of the story, the same behaviour is well received. The two versions of the story used for the perspective taking exercise are presented in Appendix D.

After reading the story, participants answered a series of questions about how they would feel in the situation they had just read. We adapted an existing perspective taking exercise, which had originally been developed to reduce prejudice against homosexuals9and has been shown to successfully change attitudes in a classroom setting10. None of the participants were informed that the situation described in the perspective taking exercise can mirror the situation faced by women. Our intent was for the participants to make the link themselves when they engaged in either one (or the two) of the following quizzes:

i. the gender quiz, which aims at making participants aware that women are underrepresented in leadership positions, that this situation is not due to what is commonly believed but rather to stereotypical expectations about leaders and women;

ii. the diversity quiz, which serves to verify whether any effect obtained by the gender quiz should be attributed to gender specifically, or diversity related questions would produce the same effect.

The gender and diversity quizzes are presented in Appendix E and Appendix F. The design of the awareness training was thus a 2 x 2 x 2 fully crossed between

subjects experiment11. In order to run such a design we programmed 8 different surveys one for each experimental condition. Participants were randomly assigned within country to one of the surveys.

After completing the awareness training, we contacted the participants one last time and asked them answer a questionnaire that would allow us to measure sexism and thus evaluate the efficiency of our intervention. In order to not make participants weary, we chose to use a different measure of sexism than the one which had been used in the athletics population survey. The measure we used this time defines sexism as taking on two forms: an old fashioned and a modern form12. Conceptually, this measure of sexism has some overlap with the ASI used in the athletics population survey and could, in the context of this awareness training, serve as a pre-intervention control measure.

9 Hillman, J., & Martin, R. A. (2002). Lessons about gay and lesbian lives: A spaceship exercise. Teaching of Psychology, 29 (4), 308-11. 10 Hodson, G., Choma, B., & Costello, K. (2009). Experiencing Alien-Nation: Effects of a simulation intervention on attitudes toward homosexuals. Journal of Experimental Social Psychology 45, 974-978 11 The design is the following: perspective taking (situational constraints vs. no situational constraints) x gender quiz (present vs. absent) x diversity quiz (present vs. absent). 12 Swim, J., Aikin, K., Hall, W., & Hunter. (1995). Sexism and racism: Old-fashioned and modern prejudices. Journal of Personality & Social Psychology, 68, 199-214.

- Page 19 of 50 -

Results

We estimated three statistical models to evaluate the effectiveness of the awareness training for the athletics population. We first estimated a baseline model (model 1), which looks at the effects of the experiment on sexism without controlling for any other variables. Given that we randomly assigned participants within country to a given condition, and given that we would expect to find permanent difference between countries, we estimated a second model (model 2), which tests for the effects of the experiment while controlling for any unobserved country-level heterogeneity. Finally, in a third model (model 3), we additionally accounted for the effects of participant age, gender, and pre-intervention levels of sexism (as measured in the survey). Table 4 below presents the results from these three models.

We find a three-way interaction between our experimental manipulations in models 1 and 2. The pattern of this three-way interaction is the same for both models and is graphically depicted in Figure 7. It can be understood as follows: taking either the diversity quiz or the gender quiz reduces sexism. Furthermore, we find that taking both quizzes has an additional effect over taking only either one of them but only for those participants for whom the perspective taking exercise involved situational constraints. For participants who were asked to take part in a perspective taking exercise in which the situational constraints were nonexistent, taking both quizzes did not have any additional effects beyond the effects of either one of the quizzes.

In model 3, we find that the main effects of both quizzes are significant and are qualified by a two-way interaction, which is graphically represented in Figure 8. The three-way interaction is no longer statistically significant (although we observed a similar pattern of effects) and would suggest that this model specification is too strong a test. Following these results, as a cautionary measure, we tested whether our experimental manipulations would predict sexism measured during the survey (i.e., prior to the intervention). In other words, we tested whether the average level of participants’ sexism was equal across conditions, and thus, whether we had successfully randomized participants across conditions as far as sexism was concerned. Our intervention successfully passed this test.

Finally, it is noteworthy that the gender and age of participants have highly significant effects. The researchers from the University of Lausanne will conduct sub-sample analyses in future studies to further probe these results.

- Page 20 of 50 -

Table 4 – Experimental effect on sexisma

(1) (2) (3) Variable Sexism Sexism Sexism

Perspective 0.179 0.141 0.0789

(-1.85) (-1.6) (-0.94)

Gender quiz 0.239 0.255* 0.261*

(-1.92) (-2.14) (-2.37)

Diversity quiz 0.281* 0.275* 0.228*

(-2.24) (-2.37) (-2.39) Perspective x Gender quiz

-0.17 -0.142 -0.109 (-1.23) (-1.05) (-0.74)

Perspective x Diversity quiz

-0.298* -0.215 -0.101 (-2.54) (-1.85) (-1.06)

Gender quiz x Diversity quiz

-0.306 -0.308* -0.264* (-1.91) (-2.19) (-2.12)

Perspective x Diversity quiz x Gender quiz

0.434** 0.328* 0.228 (-2.71) (-2.12) (-1.46)

Sexism (time 1) -0.206*** (-4.04)

Age 0.00498** (-3.29)

Gender 0.378*** (-8.84)

Country fixed effects

Included Included

Intercept -0.245** -0.409 -0.000159

(-2.95) (-1.26) (-0.00) n observations 998 998 980

Note: t statistics in parentheses. All models were estimated using OLS estimator and country level cluster robust standard errors. a The dependent variable sexism is measured using Swim et al.’s (1995) measure of old-fashioned and modern sexism. The scale is constructed such as that lower scores on this variable denote higher levels of sexism. * p < 0.05, ** p < 0.01, *** p < 0.001

- Page 21 of 50 -

Figure 7 – Three way interactional effects of perspective taking, gender and diversity quizzes on sexism (from model 2)

Perspective taking without situational constraints (control)

Perspective taking with situational constraints (experimental manipulation)

Note: Dependent variable is Swim et al.’s (1995) old-fashioned and modern sexism. The scale was constructed in such a way that greater values indicate less prejudice.

Figure 8 – Two way international effects of gender and diversity quizzes on sexism (from model 3)

Note: Dependent variable is Swim et al.’s (1995) old-fashioned and modern sexism. The scale was constructed in such a way that greater values indicate less prejudice.

-0.5

-0.4

-0.3

-0.2

-0.1

0

No gender quiz Gender quiz

No diversity quiz Diversity quiz

No gender quiz Gender quiz

No diversity quiz Diversity quiz

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

No gender quiz Gender quiz

No diversity quiz Diversity quiz

- Page 22 of 50 -

Follow-up

The research team from the University of Lausanne will conduct sub-sample analyses of the gender and age of the participants in future studies to further probe the results presented here. They will also communicate the content and results of this intervention to the scientific community through publications in academic journals.

In the context of its ongoing commitment to promoting the development of women leaders in the sport of athletics, European Athletics will be working with the project partners and others to develop the knowledge and experience collected here into a tool that can be used in practical settings in the sport.

Recommendations

5. The results we obtained indicate that we have put together elements of an online tool which can be easily be developed for use by sport and other organizations to provide gender awareness training to large groups of people. We know that combining a perspective taking exercise with questions and answers designed to go against preconceptions about gender and leadership (and diversity) is effective in reducing sexist attitudes. Thus, we recommend that young athletes, both boys and girls, coaches, club leaders and others take part in gender issue awareness training interventions based on these findings. An online platform should be developed extending what we have done by including animations, videos, and the such.

6. Organizations that may use such a tool in the future should be aware of the fact that running this intervention requires up to four hours per day of work for each day the intervention is active in order to collect data from each subsection of the survey/awareness training and randomly assign participants to the subsequent subsections (this may vary as a function of the number of people taking part in the intervention per day). Consequently, we recommend developing a custom survey so as to be able to automatically run the 9 surveys that form this intervention (8 experimental conditions, plus one post-intervention survey) and automatically randomly assign participants to conditions within country.

7. Finally, comments from the participants made it clear that the time required to complete the initial athletics population survey (Work Package 1) and take part in the Work Package 2 intervention (including the awareness training and the follow-up survey) were felt by some, at least, to be excessive. It is certain that a number of people will have abandoned the initial survey before completion, but it is hard for us to know how many, and then we know that some completed the survey but abandoned the process at a later point. Although the intervention tool we have in mind would not be as time consuming, we recommend that future researchers take into account the sport community they are working with and in particular its expectations and tolerance when designing their online survey and intervention projects.

- Page 23 of 50 -

Work Package 3 – Awareness Training for Federation Presidents and CEOs

Objective

Design and deliver workshops for federation leaders to increase awareness and change attitudes towards women and women leaders.

Introduction Federation leaders obviously play key roles in their organizations. Indeed, because of

the managerial discretion available and their function as role models, federation presidents and CEOs to are able to directly and indirectly impact their organizations, including their internal culture. Therefore ethical and values-based leadership at the top of federations is a sine qua non to fostering gender equality at both the higher levels of the organizations and, through a trickle-down process, in the attitudes seen at the other levels of the sport.

The purpose of this work package was to promote, and give an understanding of ethical and values-based leadership to athletic federation leaders in the hope that it would impact the culture of the organizations and the sport. Just as importantly, we wanted to develop our understanding of the most effective way to teach these. In other words, to create a top-down approach to changing attitudes within sports organizations towards women in general, and women leaders in particular. If we can identify and confirm effective teaching techniques, they can be developed for use in a variety of situations and in other sports.

Content and implementation

This intervention was delivered during European Athletics’ biennial CEO conference. The CEO conference was originally scheduled to take place in Barcelona, Spain, in April of 2010. However, a day prior to the beginning of the conference, the European air space was shut down following the eruption of the Eyjafjallajökull volcano in Iceland and two thirds of the participants were unable to attend. Because of this force majeure the CEO Conference was rescheduled to November 2010.

Federation presidents and CEOs took part in this conference along with a few of their top managers. The intervention per se took place within sessions devoted to strategic leadership training. After having taken part in one session on leadership in general, the participants were randomly split into two groups. Each group was led by a training consultant who facilitated group discussions focused heavily on ethical and transformational leadership—a highly effective component of leadership based on morals and vision. The consultants demonstrated how to effectively communicate a vision through the use of rhetoric and they put a strong emphasis in the value of ethical leadership (from both pragmatic and moral standpoints). In one of two groups, a strong emphasis on gender issues and leadership was placed.

- Page 24 of 50 -

Participants (n = 66) were asked to complete a pre-conference survey in which they completed the ambivalent sexism inventory (ASI)13 and a short version of the questionnaire measuring the big-five dimensions of personality14. During the conference, we also obtained measures of fluid intelligence15, as well as measures of implicit sexism (IAT)16 for some of the participants.

Unfortunately, we were unable to conduct a post-intervention study to assess the effectiveness of this intervention. Our original intent was to sample the work colleagues in the organizations of all the seminar participants to test whether any observed systematic difference in organizational-level sexism could be attributed to group assignment. However, because of the volcanic eruption in Iceland (which delayed the intervention) and mainly the workload inherent to the position of the seminar participants, we were not able to get sufficient numbers to participate in a follow-up to the intervention at a distance.

In order to assess the effectiveness of this training, over the next couple of years, European Athletics will be monitoring the proportion of female employees in the different federations, especially at the higher-echelons, in order to observe any systematic difference which could be attributed to participants’ group assignment in this intervention.

Results

As mentioned above, we cannot, as of now, assess the effectiveness of this intervention in a scientific manner. Nevertheless, we have reasons to believe that it has had an effect. First, this intervention was carried out by seasoned trainers who are accustomed to giving such interventions—and especially to such a high profile audience. Indeed, the researcher who devised the intervention and delivered it to one of the groups has published scientific papers showing the effectiveness of this leadership training, and is highly proficient on matters relating to gender and leadership (he has been nominated for several awards because of his engagement in this direction). Second, we informally surveyed a sample of participants to gauge their feelings about the intervention. The feedback obtained was by and large very positive. The participants reported that the leadership training was extremely interesting and interactive, and that the discussion on gender and leadership was particularly pertinent. Thus, we are confident that the intervention was effective.

Using the data collected in the pre-conference survey we can test whether country level factors can predict sexism amongst top-level leaders in athletics organisations. Societal norms convey expectations of men and women and consequently influence sexism attitudes. Thus, we estimated a model regressing explicit measures of sexism (ASI) on the following country-level factors:

13 Glick, P., & Fiske, S. T. (1996). The Ambivalent Sexism Inventory: Differentiating hostile and benevolent sexism. Journal of Personality and Social Psychology, 70, 491-512. 14 Donnellan, M. B., Oswald, F. L., Baird, B. M., & Lucas, R. E. (2006). The Mini-IPIP Scales: Tiny-yet-effective measures of the Big Five Factors of personality. Psychological Assessment, 18(2), 192-203. 15 Vernon, P. A. (1993). Der-Zahlen-Verbindungs-test and other trail-making correlates of general intelligence. Personality and Individual Differences, 14, 35-40. 16 Greenwald, Anthony G.; McGhee, Debbie E.; Schwartz, Jordan L.K. (1998). Measuring Individual Differences in Implicit Cognition: The Implicit Association Test, Journal of Personality and Social Psychology 74 (6): 1464–1480

- Page 25 of 50 -

Hofstede’s cultural dimension of masculinity which measures the extent to which there are role gender expectations in a given culture;

Atheism which is directly related to the prevalence of religion which may convey expectations about gender.

Because this country level data is not available for all the countries in our sample, our

sample size for this analysis is effectively reduced by half to 31 observations.

Table 5 – Experimental effect on sexism

Variable Sexism

Hofstede’s Masculinity dimension

0.48*

(2.20) Atheism 0.38 (1.96) GDP -0.22 (-1.72) Age -0.17 (-0.93) Gender -0.09 (-0.46) Intercept 1.59 (1.50)

R2 29.09% N 31

Note: t statistics in parentheses. Estimates are beta standardized coefficients. Standard errors are robust to heteroskedasticity. * p < 0.05

We find Hofstede’s country level measures of masculinity, collected several decades ago, predicts individual levels of sexism of athletic federation CEOs and presidents. The fact that these measures are predictive of current levels of sexism is indicative that sexist attitudes are still rather entrenched and that the present project is highly relevant to advance gender equality in European sports. The University of Lausanne researchers and the project partners continue to believe that through interventions such as the present ethical leadership training sexist attitudes at the top of organizations can be changed and we will be working to devise ways that these can be scientifically shown to be effective.

Follow-up

European Athletics will continue to monitor the number of women in leadership and management positions in the different athletic federations to see if there is any change that might be attributed to the intervention.

- Page 26 of 50 -

The University of Lausanne and European Athletics will discuss the development of the training method tried in this work package, the possibility of a second seminar and new approaches to working with federation leaders to raise awareness and change attitudes towards women leaders in the sport.

Recommendations

8. Focus specifically on the “trickle-down” hypothesis mentioned above by surveying the different federations to test to which extent changes in the attitudes of federation leaders carry over to their followers as a result of the present intervention or future projects.

9. Stage a second workshop with federation CEOs and presidents. As we have not been able to get access to federation leaders from a distance for important follow-up actions, having them at hand on two occasions will probably be a necessary condition to i) fully assess the effectiveness of the intervention on the participants and ii) investigate the extent to which changes in the attitudes of federation CEOs and presidents carry over to their followers.

10. Offer additional transformational and charismatic leadership trainings to male leaders who have been identified as non-effective and/or problematic with regard to gender equality issues in the sport.

11. Ensure future workshops are tailored to the participants in order to be well received. It is likely that federation presidents and CEOS would not “buy in” to a perspective taking exercise involving a hypothetical situation on an alien planet (because of their prior experience and achieved rank in their organizations). However, such an exercise would certainly serve as a valuable starting point from which to work from in order to devise a tool specifically designed for this population. Furthermore, these two workshops might be enhanced by taking into account the findings from Work Package 2 (in terms of understanding efficient means to reduce sexism).

- Page 27 of 50 -

Work Package 4 – Skills Training for Future Women Leaders

Objective

Design and deliver a seminar to increase the knowledge of sports management and the charismatic and visionary skills of potential women leaders.

Introduction

Prototypical leaders are associated agentic attributes such as being dominant, achievement-oriented, and so forth. Men also tend to be associated with such agentic attributes, while women are usually associated with communal attributes, such as being nurturing, being understanding, etc. Because of the mismatch between the prototype of a leader and the prototype of women, women are perceived as less leader-like than men. Furthermore, if women behave in agentic ways, they will be disliked because as a result of violating the gender role expectation. Given that leadership emergence depends on the extent to which individuals are perceived to match the leader prototype, teaching individuals to behave in more leader-like ways and to be more charismatic (an important component of leadership) should increase the probability that these individuals attain leadership positions, either through elections or through an employment selection process.

In addition, anecdotal evidence suggests that in a number of athletics federations, lack of training or knowledge in sports management and international contacts are used as reasons to pass over women candidates for leadership positions. Taking away these arguments against individuals by providing specific international training could also be important factors in them advancing into leadership positions.

Studies have shown that leadership and charisma can be taught and it is certainly

possible to increase the knowledge of an individual about sports management and related topics. The seminar that was the focus of this work package was designed to increase the participants’ chances of attaining leadership positions in the future by providing extensive training in a broad array of leadership and charisma skills, by providing technical and organisational information about athletics in Europe and by integrating participants in a network of future women leaders. Importantly, we wanted to test the format and key elements of the seminar to see if they could be used (and developed) to training other potential leaders in athletics and other sports.

Content and implementation

In January 2011, 44 women identified by their national athletic federation as having the potential to attain a higher leadership position in the future took part in a week-long seminar in Evian, France. Most were nominated by the partner federations of the project but two additional participants were added at European Athletics’ expense as a part of an existing leadership training programme. Demographic information about the participants is presented in Figure 9 and Figure 10.

- Page 28 of 50 -

Figure 9 – Distribution of participants by country

Figure 10 – Distribution of participants by age

1

6

5

1

5

5

1

4

5

5

2 Turkey

Sweden

Slovenia

Norway

Netherlands

Malta

Iceland

France

Estonia

Denmark

Austria

- Page 29 of 50 -

The seminar was built as follows:

two and a half days were dedicated to leadership skills training (described below);

three evenings were dedicated to networking and to the exchange of leadership experiences;

half a day served as a team building exercise during which participants engaged in activities with young athletes from the Evian area;

a day and a half was dedicated to meeting top-level leaders in European Athletics (e.g., Dr Sylvia Barlag – ambassador for this project, Christian Milz – Director General of European Athletics and experts from other sports) who shared their experiences and knowledge about leadership and practical issues with the participants;

one half day was dedicated to group exercises, which served the participants to apply the training they had received, and for us, to test the effectiveness of this training.

See Appendix G for further details on the programme and timetable.

Upon arriving to the seminar, participants were randomly assigned (within country) to one of two groups. As part of a three-day program dedicated to leadership skills training, participants from both groups received training in:

intrapreuneuship (2h)—i.e., the capability to foster entrepreneurial activity within an organization;

gender awareness and leadership (2h);

leader judgment and decision making (4h);

transformational leadership (8h)—i.e., a highly effective form of leadership based on vision and ethics.

Additionally, one group (we will refer to this group as to the charismatic group) received one hour of training in charismatic leadership during the sessions dedicated to transformational leadership. The training on charismatic leadership emphasized an understanding of what is charisma, on presenting charismatic rhetorical techniques associated with higher attributions of charisma, and on how to communicate a vision. The professor in charge of the transformational and charisma leadership module based this training on some previous work he had published in a top scientific journal17. It is noteworthy that transformation leadership is the best available intervention in terms of leadership training18. Transformational leadership training served here as our benchmark training and we investigated whether we could further improve transformation leadership training with charismatic training. This set-up ensured that both groups received valuable leadership training (rather than testing one training against a control group receiving no training).

17 Antonakis, J., Fenley, M., & Liechti, S. (in press). Can charisma be taught? Tests of two interventions. Academy of Management Learning and Education. 18 Avolio, B. J., Reichard, R. J., Hannah, S. T., Walumbwa, F. O., Chan, A. (2009). A meta-analytic review of leadership impact research: Experimental and quasi-experimental studies, The Leadership Quarterly, 20(5), 764-784.

- Page 30 of 50 -

The effectiveness of the training was assessed during the half day dedicated to group exercises. Groups of four participants were made by randomly selecting two participants from the transformational leadership group and two participants from the charismatic group. All groups engaged in a series of three coordination and information sharing exercises. As mentioned before and as we will discuss in the next section, these group exercises served to establish whether the participants in the charismatic group would come out as being more leader-like than participants in the other group.

Participants completed a number of questionnaires (Work Package 1) before and after receiving the leadership skills training and after completing the group exercises. The measures collected served different purposes as explained below.

Prior to the leadership skills training, we collected from participants measures of traits that are known to be associated with leadership emergence and leadership effectiveness. These are: the big-five dimensions of personality19, fluid intelligence20, and self-efficacy21. Accounting for the effect of these stable individual differences allows to us to better detect the training effects.

After the leadership skills training and before the group exercises, participants took a quiz on transformational and charismatic leadership, completed a scale measuring core self-evaluations (i.e., self-esteem, self-efficacy, locus of control, and neuroticism), and evaluated the trainers. Following previous scientific research on charismatic leadership training22, we employed the evaluations of trainers and the scores on the quiz to test whether any post-training measured differences between the charismatic group and the other group could be attributed to the charismatic leadership training. Indeed, the quiz served to test whether participants in the charismatic group had a greater understanding of charismatic leadership than did participants in the other group. The evaluations of the trainers allowed us to verify whether the participants from both groups had perceived their training in a similar fashion.

After the group exercises, participants rated their group members in terms of leader prototypicality using a scientifically validated instrument23. Our interest was to see if the within group ranking of leader prototypicality could be predicted by the knowing from which groups participants came, and whether any effect could be attributed to the additional charismatic training.

19 Donnellan, M. B., Oswald, F. L., Baird, B. M., & Lucas, R. E. (2006). The Mini-IPIP Scales: Tiny-yet-effective measures of the Big Five Factors of personality. Psychological Assessment, 18(2), 192-203. 20 Vernon, P. A. (1993). Der-Zahlen-Verbindungs-test and other trail-making correlates of general intelligence. Personality and Individual Differences, 14, 35-40. 21 Chen, G., Gully, S. M., and Eden, D. (2001). Validation of a New General Self-Efficacy Scale. Organizational Research Methods, 4, 62-83. 22 Towler, A. J. (2003). Effects of charismatic influence training on attitudes, behavior, and performance. Personnel psychology, 56(2): 363 - 381. 23 Cronshaw, S. F., & Lord, R. G. (1987). Effects of categorization, attribution, and encoding processes on leadership perceptions. Journal of Applied Psychology, 72, 97–106.

- Page 31 of 50 -

Results

Participants from both groups rated all four trainers of the leadership skills training on a seven-point scale. The trainer responsible for the transformational and charismatic training was equally well evaluated by both groups indicating that participants from both groups found the training of value. There was a significant difference in the mean evaluation of the three other trainers by both groups. These results are depicted in Figure 11. This difference was such as that the charismatic group evaluated the other trainers less positively than did the other group. This difference is likely to result from an ordering effect between the different session. Indeed, the charismatic group began the seminar with the transformational leadership session which was the most appreciated. Thus, the following sessions probably suffered from the comparison.

Figure 11 – Evaluation of trainers by group

As figure 12 indicates participants in the charismatic group scored higher on the charisma related questions indicating that they had a better knowledge of charismatic leadership.

0

1

2

3

4

5

6

7

Group receiving transformational

leadership training

Group receiving transformational and

charismatic leadership training

Ratings of trainer for the transformation and charismatic leadership session

Ratings of the trainers for the other sessions

- Page 32 of 50 -

Figure 12 – Participants’ score to questions on charismatic leadership

Note: Error bars represent the standard error of the mean.

We used leader prototypicality ratings to establish a within group ranking of the individuals perceived as being the most leader-like. We tested whether this ranking could be attributed to the knowledge about charismatic leadership (as measured by the score on the charisma quiz).

Finally, we tested whether the differences in scores on the charisma could be attributed to the training in charismatic leadership, and this beyond the effects of personality, intelligence, self-efficacy, and age. Table 6 below presents the results of these analyses.

These results show a strong effect of the charismatic leadership training on leader prototypicality ranking (model 1), an effect that is fully mediated by the knowledge of charismatic leadership (model 2). The effect of the charismatic leadership training on leader prototypicality ranking holds when not controlled for the knowledge of charismatic leadership (model 3).

The results also lend support to the effectiveness of charismatic leadership training above and beyond that of transformational leadership training. In this intervention, the participants who received both the transformational and charismatic leadership training were perceived as more leader-like by their group members (across all three group exercises).

- Page 33 of 50 -

Table 6 – Model predicting the effects of charisma training on leader prototypicality ranking through charisma quiz

Variable

(1) Leader prototypicality ranking

Charisma quiz 0.383*** (0.115) R2 31.1%

Variable (2)

Charisma quiz

Charisma training 2.045*** (0.552) Extraversion 0.456 (0.251) Neuroticism -0.268 (0.324) Agreeableness 0.646 (0.445) Conscientiousness -0.094 (0.304) Openness 0.428 (0.491) Fluid intelligence -0.010 (0.032) Self-efficacy -0.754 (0.545) Age -0.035 (0.027) Intercept 7.926 (2.786) R2 40.9%

Note: Robust standard errors in parenthesis. The model predicting Leader prototypicality ranking is an ordered probit model. The model predicting Charisma quiz is an OLS model. Both models were estimated simultaneously.

The error terms of both models do not significantly correlate ( 2(9) = 7.67, p = 0.57). *** p < 0.001

- Page 34 of 50 -

Table 7 – Model predicting the effects of charisma training on leader prototypicality ranking without charisma quiz

Variable7= (3)

Leader prototypicality ranking

Charisma training 0.774* (-0.394) Extraversion 0.623* (-0.284) Neuroticism -0.161 (-0.228) Agreeableness 0.522 (-0.439) Conscientiousness -0.016 (-0.131) Openness 0.787* (-0.319) Fluid intelligence 0.0406** (-0.016) Self-efficacy -0.319 (-0.394) Age -0.0207 (-0.018) Intercept 0.774* (-0.394) pseudo R2 13.24%

Note: Robust standard errors in parenthesis. This model is an ordered probit model. ** p < 0.05, ** p < 0.01

We have been able to demonstrate that the leadership training we provided to the participants had significant effects and that it can be conducted over in a relatively short period of time. Thus, scientifically validated leadership training could be proposed to women having the potential to attain leadership positions as a way of helping them advance more effectively.

Interestingly, participants assigned to the charismatic leadership group scored lower on the post-training core self-evaluation questionnaire in comparison to participants from the other group. This result can most probably be attributed to what can been referred to as “the burden of expertise”24. Indeed, while individuals have a tendency to overly evaluate their abilities in areas of incompetence, in areas of competence, individuals have the opposite tendency, namely that of underestimating their own skills.

Follow-up

The University of Lausanne will be conducting two follow-up studies. The first will be to assess whether the training will impact participants’ propensity to be elected or appointed to key positions in the sport. The second will be to assess the change in participants’ leadership behaviour over time, by obtaining ratings of the participants’ leadership behaviours from peers, subordinates, and superiors (i.e., using a 360° degree leadership evaluation). This later study was not planned in the initial proposal and is thus an addition to the project.

24 Kruger, J., & Dunning, D. (1999). Unskilled and unaware of it: How difficulties in recognizing one’s own incompetence lead to inflated self-assessments. Journal of Personality and Social Psychology, 77, 1121-1134.

- Page 35 of 50 -

Following the seminar, European Athletics established a Facebook page entitled “European Athletics Women’s Leadership Network” to promote discussion on issues related to women in the sport and to strengthen and expand the relationships formed during the seminar. The network is open to anyone but the seminar participants have been specifically invited to join, as have the 24 winners of the 2009 European Athletics Women’s Leadership Awards. The network discussions are currently driven and monitored by a staff member of European Athletics with input from Dr Sylvia Barlag, but it is hoped that in the future a critical mass of members will be reached and the discussions will be self-sustaining.

European Athletics has created a working group that will be reviewing its strategy for promoting women leaders in the sport based on the results of the seminar and the overall project as well as ideas that emerge from the European Athletics Women’s Leadership Network. Particular consideration will be given to the possibility of regular training measures for potential women leaders.

Recommendations

12. European Athletics should offer similar leadership training on a regular basis for women with the potential to attain leadership positions in the sport, including coaches, club leaders and federation leaders.

13. Transformational leadership training—the best intervention available at this time (see footnote 12)—be supplemented with charismatic leadership training. Although the specific mechanisms through which charisma influences followers are by and large unknown, there are some elements that are understood and help to explain the influencing effect of charismatic rhetoric on followers.

- Page 36 of 50 -

Concluding Remarks

On behalf of all the partners, we would like to thank the European Commission for its interest and confidence in our project and for the support provided by the Sports Unit over the course of its implementation.

In summary, we are satisfied with the overall outcome of this project. Indeed, at the end of the project, three of the four work packages have had significant results:

Work Package 1—Survey: We obtained a statistical picture of the levels of sexism in Europe;

Work Package 2—Online Awareness Training: We devised an online tool which was successful at reducing prejudice towards women;

Work Package 4—Skills Training for Future Women Leaders: We delivered a targeted intervention to potential future women leaders which provided them with leadership skills. Additionally, we enhanced the best available intervention in terms of leadership training and observed a difference in terms of leadership emergence.

As for the remaining work package (i.e., Work Package 3—Awareness Training for Federation CEOs and Presidents), we can only speculate as to its effectiveness, but as discussed earlier, we have good reasons to expect a positive effect. The following summarises our recommendations:

Part I - Survey:

1. Conduct surveys similar to our general athletics population survey over time to document sexism in European athletics, in order to i) observe the evolution of attitudes so as to decide when and where to conduct gender awareness interventions, ii) allow decision makers to optimally decide when and where to implement interventions to reduce sexism and iii) assess the effectiveness of such interventions.

2. Conduct similar surveys in other sports.

3. Ensure the availability of the specialized IT support required to implement a survey running custom programmed interfaces.

4. Ensure that the aim and value of gender issue surveys to the sport as a whole are

well communicated.

Part II - Gender awareness interventions:

5. Offer an enhanced version of the awareness training for the athletics population intervention comprising both quizzes and of the perspective taking exercise to young athletes, both boys and girls, coaches, club leaders and others through an online platform including animations, videos, etc.

- Page 37 of 50 -

6. Develop a custom survey platform so as to be able to automatically run the various

surveys that form the awareness training for the athletics population intervention and randomly assign participants to experimental conditions.

7. Account for the expectations and tolerance of the participants, particularly in terms of the time necessary to complete an online awareness training task, when planning such interventions.

8. Survey federations to test to what extent changes in the attitudes of federation

leaders carry over to their followers as a result of the present intervention or of future projects.

9. Stage a second leadership workshop for federation CEOs and presidents to i) fully

assess the effectiveness of the intervention and ii) investigate to extent to which changes in the attitudes of federation CEOs and presidents carry over to their followers.

10. Offer additional transformational and charismatic leadership trainings to male

leaders who have been identified as non-effective and/or problematic with regard to gender equality issues in the sport.

11. Ensure future leadership workshops and gender awareness interventions for

federation leaders are tailored to the participants in order to be well received.

12. Offer leadership training on a regular basis for young women with the potential to attain leadership positions in the sport, including coaches, club leaders and federation leaders.

13. Supplement transformational leadership training—the best intervention available at

this time (see footnote 12)—with charismatic leadership training for women with the potential to attain leadership positions in the sport.

- Page 38 of 50 -

Appendix A – Survey homepage

- Page 39 of 50 -

Appendix B – Sample press coverage of the survey

- Page 40 of 50 -

Appendix C – Text of an e-mail from a Member Federation Communications Manager (female)

Trust you are well. [The federation CEO] has passed this onto me to take a look at in terms of both filling in and publicising and I have to say I am very uncomfortable with the tone of the survey and is not something I would want to push through our comms channels. I appreciate the sentiment of the Women’s Leadership programme and imagine there is probably some sort of educated psychological reasoning as to why these questions are being asked, but in my humble opinion the very fact that you’re even having to ask them is firstly a sad reflection on our sport and will probably only tell us what we already know – one look at both your and our councils is enough to demonstrate the problem. Surely you’re better off engaging with those of us in leadership roles and work to ensure our organisations reflect society by providing the right opportunities instead of confirming that there are lots of old men in our sport with outdated views. Sorry to be controversial, but a number of our senior team in the office here were wholly uncomfortable with the tone of questions.

- Page 41 of 50 -

Appendix D – Perspective taking exercise from the awareness training for the general athletics population.25

Imagine that, when you were only a young child, you crash-landed on an alien planet, Vigg-O-tron, along with 3000 other humans. Unfortunately, your spaceship was damaged beyond repair and all hope of escaping from the planet is gone. The inhabitants of this new planet, Vigg-O-tron, appear remarkably Earth-like and friendly. They speak the same language as you, and it is impossible to distinguish them physically from humans from a distance. The Vigg-O-tronians express a desire to welcome you to their planet. They provide each of you with an apartment, a small car, and an entry-level job.

Treatment 1—situational constraints Treatment 2—no situational constraints

Over the years, you learn that Vigg-O-tronians view humans as overly affectionate and caring, and consider humans as too sensitive to be of equal value with them. Despite this view, humans and Vigg-O-tronians live together in harmony. However since you arrived on Vigg-O-tron, not much has changed: Vigg-O-tronians have control over nearly all the major institutions including politics, the universities, the judiciary, businesses, the arts, the professions, etc.

You have just begun your first job and you are determined to make it to the top! During your first weeks on the job, you hear that one of your human colleagues who has been one of the top performers in your department over the past years has recently been refused a promotion on the grounds that he was not suited for the job because he wasn’t sufficiently forceful and aggressive. It is now clear to you, if you want to be respected by the Vigg-O-trons on your job, you will have to behave more like them. In the following weeks, you decide to give it all you have: you are assertive and dominant, and you display confidence and ambition. With this kind of attitude you WILL make it. Alas, this morning during the team meeting while you are forcefully discussing how to solve the problem at hand, your superior cuts you short and throws at you: “Hey human, don’t try to be who you’re not. It doesn’t look good on you. Why don’t you just keep cute and let the Vigg-O-trons lead this meeting?” Everyone, Vigg-O-trons and humans, roar with laughter. When you leave the conference room, one of your colleagues says quietly in your ear: “If you’re smart, you’ll forget about that kind of behaviour”.

Over the years, you learn that Vigg-O-tronians view humans as affectionate and caring, and consider humans to be of equal value with them. Because of this view, humans and Vigg-O-tronians live together in harmony. Since you arrived on Vigg-O-tron, a lot has changed: Vigg-O-tronians and humans now share the control of nearly all the major institutions including politics, the universities, the judiciary, businesses, the arts, the professions, etc.

You have just begun your first job and you are determined to make it to the top! During your first weeks on the job, you hear that one of your human colleagues who has been one of the top performers in your department over the past years has recently been promoted in part because of his forceful and aggressive attitude. It is now clear to you, if you want to be respected on your job, you will have to behave more like this colleague. In the following weeks, you decide to give it all you have: you are assertive and dominant, and you display confidence and ambition. With this kind of attitude you WILL make it. Indeed, this morning during the team meeting while you are forcefully discussing how to solve the problem at hand, your superior interrupts: “Hey human, I like how you think. Good job.” Everyone, Vigg-O-trons and humans, nods in approval. When you leave the conference room, one of your colleagues says quietly in your ear: “If you’re smart, you’ll keep showing that kind of behaviour”.

25 Adapted from Hillman, J., & Martin, R. A. (2002). Lessons about gay and lesbian lives: A spaceship exercise. Teaching of Psychology, 29 (4), 308-11.

- Page 42 of 50 -

Appendix E – Gender quiz from the awareness training for the general athletics population.

Women in positions of power in Europe. Athletics Q: In 2008, what percentage of registered athletes in Europe were female? A: In 2008, women represented nearly half (48.45%) of the registered athletes in

Europe. (Source: IAAF). Politics Q: In 2009, out of the 27 member countries of the European Union, how many had a

female head of state? A: Only 5 of the 27 member countries of the European Union had a female head of state

in 2009. (Source: Catalyst.org) Athletics Q: In 2008, what proportion of the various committees in the 50 European Athletics

Member Federations were chaired by a woman (in percentage)? A: In 2008, women chaired 15.9% of committees in the sport of athletics in Europe. One

quarter of those chaired a women’s committee. (Source IAAF) Corporations Q: How many women are currently CEOs of the 500 largest companies in Europe? A: Only nine women are currently CEOs of companies listed in the Financial Times

Europe 500. (footnote: http://www.catalyst.org/publication/285/women-in-europe) The situation of women in positions of power is such because… TRUE OR FALSE The reason why there are so few women in leadership positions in Europe is because…

1. Women are more emotional than men are and emotions get in the way of efficient leadership.

FALSE: The difference between men and women in terms of emotional stability is

negligible and cannot explain the lack of women in leadership. (footnote reference: Costa, Terracciano, & McCrae, 2001)

- Page 43 of 50 -

2. Women are in general less educated than men. Therefore it is only natural that there are more men than women who are recruited for positions of power.

FALSE: In 2008 in Europe, 81.3% of females aged 20 to 24 had completed at least

upper secondary education compared with 75.6% of males. In 2006, 55.1% of students enrolled in tertiary schools were women. (footnote source: Eurostat)

3. Leaders should have tough-guy characteristics (that is, achievement-oriented,

assertive, ambitious, and dominant). Women who behave in tough-guy ways are disliked because they don’t behave in a way we usually think women should behave. As a consequence, women are not promoted as often as men are.

TRUE: Women who act in tough-guy ways challenge the commonly shared

conception of how women should behave and we don’t like when something goes against our preconceived ideas. So, a woman who is not acting like a (typical) woman will usually be disliked which penalizes them when they seek to occupy a leadership position.

- Page 44 of 50 -

Appendix F – Diversity quiz from the awareness training for the general athletics population.

Diversity in Europe Athletics Q: How many Member Federations does European Athletics have? A: European Athletics has 50 Member Federations (Source: http://www.european-

athletics.org/european-athletics-member-federations-new.html) Politics Q: How many ethnic groups live in Europe? A: Researchers have reported that 87 distinct "peoples of Europe" cohabit together

across European countries (Source: http://en.wikipedia.org/wiki/Ethnic_groups_in_Europe)

Athletics Q: How old is the oldest man to have taken part in an official athletics competition? A: John Whittemore threw the javelin and discus at the age of 104! (Source:

http://en.wikipedia.org/wiki/John_Whittemore) Corporations Q: Which percentage of Europe's leading companies mention "diversity" in their annual

reports? A: An astounding 76 percent of Europe's leading blue-chip companies mention

"diversity" in their annual reports A new survey, notes that "diversity is now a strategic success factor" that "leading companies can no longer overlook." (Source: http://www.diversityinc.com/article/5829/Are-Top-Companies-in-Europe-Communicating-Diversity)

The effect of diversity is… TRUE OR FALSE Diversity can have an important effect on organizations of all sorts. Indeed …

1. A team composed of homogeneous members (i.e., similar in terms of background, age, race, and so on) will usually make better decisions than a team composed of very diverse members.

- Page 45 of 50 -

FALSE: Heterogeneous groups benefit from a greater variety of opinions and are therefore more likely to take into account different facets of a problem and, ultimately, make better decisions. (footnote reference: Janis, 1972)

2. Teams whose members are diverse in terms of personality traits, values, and

attitudes are more effective than teams whose members are diverse in terms of race or age (or any other characteristics which can be perceived upfront). FALSE: As individuals work together and get to know each other, the impact of visible

differences between team members (such as, e.g., race or age) decreases. However, differences in job-related beliefs, attitudes, and values will continue to be a source of friction and conflict. (footnote reference: Harrison, Price, Gavin, & Florey, 2002)

3. In the end, every one has the same chances. It’s just up to you to try to make it

happen.

FALSE: The expectations we have of others influences our judgement. If we like someone, we will immediately be more helpful with this person than if we don’t like him or her—even if you just met this person a second ago! If a group of people is associated with not so positive expectations, members of this group will have to work harder to overcome these expectations.

- Page 46 of 50 -

Appendix G – Volcano forces change to CEO Conference plans

- Page 47 of 50 -

Appendix H – Future Women Leaders Seminar Programme

Closing the Leadership Gender Gap in European Athletics

Future Women Leaders Seminar Evian/FRA, 16 - 21 January 2011

Programme

Sunday – 16 January All Day: Arrival and Check-in 12:00-13:00 Lunch Evening: Welcome Dinner Monday – 17 January 09:00-09:05 Welcome (Marcel Ferrari, Vice President, Fédération Française d’Athlétisme) 09:05-10:00 Introduction (Bill Glad, Development Manager, European Athletics, and Dr Philippe

Jacquart, Closing the Leadership Gender Gap Project Manager) 10:00-10:30 Coffee Break 10:30-12:00 Developing Women Leaders in Athletics – The Rhône-Alpes Programme (Anne Roger,

Council Member, Ligue d'Athlétisme Rhône-Alpes and Member of the Center of Research and Innovation in Sport - University of Lyon 1)

12:00-13:30 Lunch 13:30-15:15 Workshop Session I Group A: Leadership & Influence) – Part 1 (John Antonakis, Professor, University of

Lausanne) Group B: Intrapreneurship (Dr Jane Khayesi, University of Lausanne) 15:15-15:30 Coffee Break 15:30-17:30 Workshop Session II Group A: Leadership & Influence – Part 2 (Antonakis) Group B: Leader Decision Making – Part 1 (Jacquart)

- Page 48 of 50 -

19:00-20:00 Dinner 20:30-21:00 Case Study Presentation I Tuesday – 18 January 8:30-10:15 Workshop Session III Group A: Leadership & Influence – Part 3 (Antonakis) Group B: Leader Decision Making – Part 2 (Jacquart) 10:15-10:30 Coffee Break 10:30-12:30 Workshop Session IV Group A: Leadership & Influence – Part 4 (Antonakis) Group B: Gender Awareness (Marika Fenley, University of Lausanne) 12:30-13:30 Lunch 13:30-15:15 Workshop Session V Group A: Gender Awareness (Fenley) Group B: Leadership & Influence – Part 1 (Antonakis) 15:15-15:30 Coffee Break 15:30-17:30 Workshop Session VI Group A: Leader Decision Making – Part 1 (Jacquart) Group B: Leadership & Influence – Part 2 (Antonakis) 19:00-20:00 Dinner 20:30-21:00 Case Study Presentation II Wednesday – 19 January 8:30-10:15 Workshop Session VII Group A: Leader Decision Making – Part 2 (Jacquart) Group B: Leadership & Influence – Part 3 (Antonakis) 10:15-10:30 Coffee Break 10:30-12:30 Workshop Session VIII Group A: Intrapreneurship (Khayesi) Group B: Leadership & Influence – Part 4 (Antonakis) 12:30-14:00 Lunch 14:00-16:00 Sports Activity 19:00-20:00 Dinner 20:30-21:00 Case Study Presentation III

- Page 49 of 50 -

Thursday – 20 January (Day at Olympic Museum) 08:20-09:15 Ferry from Evian to Lausanne

Walk to Olympic Museum 9:15-10:00 European Athletics: Structure, Strategy and Programmes (Christian Milz, Director

General, European Athletics) 10:00-10:30 European Athletics’ Strategy for Developing Women Leaders (Dr Sylvia Barlag, Vice

President, Atletikunie and Council Member, European Athletics) 10:30-10:45 Coffee Break 10:45-11:15 My Life in Athletics and the European Athletics Women’s Leadership Awards (Thordís

Gísladóttir, Assistant Professor, Reykjavik University and Winner of 2009 European Athletics Women’s Leadership Award)

11:15-11:45 The Barcelona 2010 Experience (Monica Barra, General Coordinator, 2010 European

Championships in Athletics) 11:45-12:15 Q&A’s (Barlag, Gísladóttir, and Barra) 12:30-13:30 Lunch 13:30-14:15 Women’s Leadership Project in Cycling (Andrea Marcellini Mendonça, Coordinator of

Cycling for All, Masters and Sustainable Development, Union Cycliste Internationale The International Working Group on Women and Sport (IWG) Nicole Gruber, Commercial Manager, Union Cycliste Internationale 15:00-15:45 Q&A’s (Marcellini Mendonça and Gruber) and Open Discussion 15:45-16:30 Coffee Break 16:00-18:00 Olympic Museum Visit 18:30-21:30 Return to Evian and Dinner Friday – 21 January 9:00-10:00 Group Exercises – Part I (Fenley, Jacquart and Khayesi) 10:00-10:30 Coffee Break 10:30-12:30 Group Exercises – Part II (Fenley, Jacquart and Khayesi) 12:30-14:00 Lunch 14:00-15:00 Discussion: What do we do from here? (Glad and Barlag) 15:00-15:30 Coffee Break 15:30-17:00 Final Review & Wrap-Up 18:00-19:00 Cocktail

- Page 50 of 50 -

19:00 Bus leaves for dinner Evening: Group Dinner Saturday – 22 January All day: Departures

January 2011