Mixing it up with technology

AI is the most used technology in Human Resources Management. Since Machine Learning is a subset of AI and is often connected to AI for technological applications within Human Resources Management, it is concluded within this section.

Within the employee lifecycle, Artificial Intelligence and Machine Learning are deployed in multiple areas within the DEI context:

Using Virtual Reality to change attitudes towards a different group is becoming increasingly relevant in the literature on DEI technology.

Computer vision is useful to automate the analysis of visual data. The

JanetBot

Analytics tools are often referred to as People Analytics in an HR context. Organisations can support their decision-making through analytics based on an objective, data-driven approach. Analytics can determine pay gaps of diversity representation and create valuable insights for the development of DEI initiatives.

Within the data sample, it was found that the analytics solutions either target a specific problem area (mostly pay gaps, see

PayAnalytics

Visier

Joonko

Kanarys

AllVoices

Culture Amp

Umbrella

Employee attraction is the first stage of the employee lifecycle where potential candidates become aware of a company as employer. There can be differences in how people see the external image of an employer.

For example, while men rate a high salary as more important when considering an employer, women are attracted more towards employee diversity, work-family balance, and similar colleagues. Similarly, underrepresented identities emphasise diversity within the workplace (Backhaus et al., 2002; Thomas & Wise, 1999) [5][6].

Another important aspect is the wording within job advertisements. For example, stereotypically male language in job ads can decrease the number of women that will apply for an occupation (Gaucher et al., 2011; Hentschel et al., 2020) [7][8].

An adjustment of how a recruiting message is conveyed can positively influence whether an applicant feels threatened in their identity by the advertisement – so whether a so-called Stereotype threat is triggered – and therefore if the applicant is attracted towards applying or not (Liu et al., 2021) [9].

Recruiting is the area where the tech space is most saturated and also where most issues for underrepresented identities are listed in the scientific literature.

In a study performed in the Netherlands, Derous and Ryan (2018) [10] found that people with Arab names received fewer callbacks than those with Dutch names. The same adverse effects have been found for African American names on applications compared to typically white-sounding names in the US (Bertrand & Mullainathan, 2004; Dovidio & Gaertner, 2000) [11][12].

Further, women’s qualifications are often rated lower than men’s, even if they are the same or strongly comparable (Campbell & Hahl, 2020) [13].

Within interviewing, mental processes and unconscious biases can impact the result of the interview negatively for underrepresented identities (Gaucher et al., 2011).

Factors outside of interpersonal biases that can keep diverse applicants from being recruited can also result from institutional or structural discrimination like the recruitment preference from prestigious universities (Bogen & Rieke, 2018) [14].

Literature has found significant differences in the way women and minority employees are generally rated in performance reviews and assessments of how well they are performing in their job. Performance reviews are a relevant factor in determining someone’s compensation and steering career trajectories.

In an earlier study, McKay and McDaniel found that white employees were rated significantly higher than their black counterparts within performance ratings (2006) [15].

A recent study examined language differences between performance reviews for men and women found that women are often described as more communal in their personality and communication style (Correll et al., 2020) [16]. They have less favourable future-oriented reviews. Furthermore, they are less likely to be related with ‘standout words’ like ‘visionary’ or ‘brilliance than men .

Racial minorities and women receive less mentoring on higher levels within a company. Therefore, they are appointed to corporate boards in smaller numbers than White men (McDonald & Westphal, 2013) [17].

Women and people of colour are also at a disadvantage in their assignments at work - receiving less prestigious tasks than white men (Williams & Multhaupt, 2018) [18].

A Canadian study also found that racial minority employees are less likely to be promoted than their White counterparts, particularly at lower organizational levels (Yap & Konrad, 2009) [19].

Engaging and retaining diverse employees is deeply connected to inclusion and how integrated they feel into the workplace (Brown, 2018) [20]. An essential way of fostering the feeling of inclusion is employee voice. Voice is understood as how employees can influence decision-making and other aspects like activities at work.

Voice mechanisms like anonymous feedback channels or dedicated networks have shown to be very important for underrepresented identities like members of the LGBTQIA+ community (Bell et al., 2011) [21].

Unconscious bias trainings are commonly used to improve DEI at the workplace. In 2017, companies in the US have spent 8 billion dollars on diversity training (Kirkland & Bohnet, 2017) [22].

However, except for initiatives that were scoped for an extended period (Rudman et al., 2001) [23], few positive results have been reported for these types of training (Forscher et al., 2019) [24]. On the contrary, it has been shown that these initiatives can even lead to increased discriminatory behavior (Kalev et al., 2006) [25].