Literature Review Women on boards and in TMTS and firm performance64 total In another case I identify a 7455 sha re of female employees as a faulty answer Third when respondents state to have zero members on the executive board and or supervisory board in one case although claiming to be share and bond issuer I delete the entries relating to the share of female board members and leave the fields wi thout an entry in order to avoid bias In the end 93 evaluable data sets remain A quantitative content analysis is used for evaluation of collected data predominantly as univariate or bivariate frequency analysis 3 4 2 Limitations of the survey method The advantage of the present survey is clearly the exclusive access to a narrowly defined group of participants through leading professional associa tions Almost 90 percent of the respondents are members of the investor relations department of a listed company I believe that an IRO who is usually highly specialized in finance and corporate governance is in a daily dialogue with internal de cision makers and external stakeholders Owing to this function as interface between the issuer and capital markets the survey method provides direct insight in the attitudes on both sides towards the issue of gender diversity in top echelons However the chosen survey method is subject to at least three limitations First the IRO himself could not have an appropriate insight into or un derstanding of top management s attitude towards the topic of women on executive levels The same could be true for capital markets actors such as investors or research analysts Second sample selection bias could be present Sample selection bias can occur from several fac tors To start with the chosen survey period from mid January to mid February is traditionally a time of heavy workload as IROs contribute to the crea tion of the annual financial statements Even so I decided to conduct the survey at this particular time as the women s quota was a hotly debated issue in the public exactly then The rather low response rate of 6 9 percent however may indicate that a significant number of addressees did not participate in the survey due to time constraints This could also mean that actual respondents are less involved in certain important tasks such as creation of financial reports and thus represent a certain subgroup of all IROs The non participating IROs responses could have been quite different On the other hand representativeness does not automatically increase in line with the response rate Surveys with very low response rates may even provide more precise results than surveys with high response rates Krosnick 1999 Nonetheless nonresponse bias could affect my results validity Furthermore self selection is a possibility A total of 41 terminations occurred at page 1 It is concei vable that only participants with a strong personal interest in the topic of gender diversity decide to continue A disproportionate participation of wo men does not seem to be the case as the gender ratio is fairly balanced Yet respondents individual involvement or experience could influence the like lihood of continued participation It is also possible that participants mistrust the given guaranty of an onymity and thus decline or interrupt participation Several missing entries regarding characteristics of the relevant company such as industry classificati on or total number of employees may be interpre ted as evidence for distrust For both reasons the surveyed data could not be representative for the population Third social desirability bias could be present due to the sensitive topic Social desirability could im pact responsiveness in both cases when reflecting Gender diversity on corporate boards and in TMTs in practice

Vorschau DIRK-Forschungsreihe Band 21 Workforce diversity and personal policies Seite 64
Hinweis: Dies ist eine maschinenlesbare No-Flash Ansicht.
Klicken Sie hier um zur Online-Version zu gelangen.