Bechtel – reducing bias in recruitment

Bechtel is an engineering, construction and project management organisation with approximately 1000 staff in the UK and 53,000 globally. Most staff are based in the London area, but it also has a presence in Bristol, Reading, Swindon, Birmingham, Warrington and Cumbria. Bechtel works with UK clients on major projects in the rail, aviation, nuclear and process market sectors and has been in the UK for nearly 70 years.

What did they do?

Bechtel has worked with external vendors to develop two online assessment programmes with the aim of minimising the opportunity for bias to creep into the selection of candidates during recruitment. The target population for both programmes is initially early career hires (ECHs). Once it has been established there, the programme will be extended to include the recruitment of experienced hires.

In the first of these programmes, Bechtel have worked with CAPP[1] to develop a series of work-based scenarios that will be used in situational judgment testing. Based on input from a broad range of stakeholders and  focus groups, all the scenarios incorporate Bechtel’s core values and are entirely bespoke to reflect its requirements.  In the completed scenario, candidates are presented with a typical scenario and asked to rank possible responses in order of preference.

Bechtel engaged with a broad range of stakeholders for their views and insights of what it believes ‘great’ looks like. Future proofing was another area Bechtel wished to consider, so the stakeholders had a good view of the organisation’s future requirements of the workforce.

As the target candidates are ECHs, it is “potential” that is being tested. In particular, Bechtel is looking for the following behavioural characteristics:

  • Collaboration
  • Learn it, do it, share it (growth)
  • Resilient adaptability
  • Deliver
  • Learn it, do it, share it (learning agility)
  • Clarity (communication)

The scenarios have been tested with a diverse internal audience of 100 individuals from the current ECH population. The diversity characteristics currently in focus are gender and ethnicity. This has allowed the validation of the preferred ranking of responses to each scenario. Those responses are now being normalised with an external population of ECH candidates to set acceptable response levels. External candidates were used they would not have the same insight into Bechtel as the current high performing ECHs.

Candidates applying for a position who meet the minimum requirement level are sent a log-on, which allows them to view video that sets out the scenario (around one minute in length) and rank the available responses in order of preference.

The second assessment, which would be applied to candidates who pass the first assessment, is an online interview tool, developed in partnership with ‘HireVue’. In this programme, questions are pre-selected based on the role in question. For the ECH roles that are the initial focus the questions relate to the same behavioural characteristics mentioned above. The system allows the style and format of the questions to be varied according to the requirements of the situation. Candidates are then sent a link that allows them to log in at a time of their choosing within the following seven days. They are also able to practice using the tool ahead of time.

When the interview takes place, candidates respond to the pre-set questions by talking to their screen and the whole interview is recorded. That recording can then be shared with those managers that play a role in selection. In this way Bechtel can ensure that the tone and content of questions is always the same and share the interview with a much wider group of assessors than would be the case through a face-to-face structured interview. The recording can be stored and referred to at a later stage if necessary when making selection decisions.

Both assessment tools eliminate the possibility of influencing the candidate or assessor through direct contact and significantly reduce the possibility of differences in interpretation of the candidate’s responses. Bechtel have also developed a framework to mark responses against to ensure greater consistency and further reduce bias.

What was the impact?

At the end of the process, the aim is to be able to show that the diversity of the candidate pools through each stage of the selection process has not been adversely impacted, including the final selection for hire. In addition, Bechtel are looking to be able to show that performance of the ECH population has also increased, however, the data to show that will take more time to gather.

Bechtel have seen the process work effectively screening out candidates who do not align as closely with the benchmark. In previous employee surveys, some employees had fed back an impression that referrals receive preferential treatment and this approach builds on other actions to reduce the potential for bias. The  consistent and objective manner of these assessments have meant that Bechtel is hiring the best students for its organisation.

It has analysed the applicant data for the UK and US populations, where the assessment has taken place. Bechtel noticed no adverse impact in both gender and minority candidates.

Where there any challenges?

It took time and the right opportunity for this initiative to come into fruition. The time it took was effective in producing a good product. There were and are some challenges in gaining buy-in at every level, even though it may be endorsed at the most senior levels there may be pockets of resistance whose voices are often louder than those who really embrace the assessment and the change it brings.

Hints and tips

Think about reasonable adjustments that need to be made and how candidates can get in touch to request this. Bechtel have added a general email address online and with all online communication so candidates can get in touch for further assistance. If this is your first step in introducing a formal assessment in selection then think about the people impact, how you can bring your stakeholders in the journey with you and what questions may come out from the doing this activity.

Think about GDPR, and how vendors can support in protecting candidate data.