Trust Issues: Obstacles to Using People Analytics for DEI Efforts
How can people analytics play a vital role in DEI efforts? It all starts with “trust.”
Algorithms and analytics are often portrayed as inherently objective and neutral, but there are several examples of discrimination manifesting in technical models through biased data and development. It is therefore important to always be wary of unintended bias when leveraging analytics. However, this weariness should not prevent progress; rather, industry leaders and data scientists can utilize analytics to understand and solve the issue. A recent Knowledge at Wharton blog post, entitled How Data Analytics Can Help Advance DEI, explores how people analytics can work to improve diversity, equity, and inclusion (DEI) efforts in the workplace. The article highlights different obstacles to using data analytics for DEI purposes, including resistance from company leadership and data collection. Yet, there is one word that doesn’t appear in the article that is essential to these efforts: trust. This five-letter word, which has operated in the background of this entire course, is crucial to unlocking the power of people analytics in DEI initiatives and must be present at different stages.
Trust in Data Collection
Today, corporations can take advantage of passive data about their teams and employees, as well as solicit active data from surveys and employee submissions. In either case, companies must consider the how trust impacts their ability to collect the information they need. When implementing surveys, for example, companies need to engender trust with their employees to ensure that people feel comfortable providing honest responses. This is especially true when seeking data for DEI purposes as those survey questions may ask individuals to share personal information that can be viewed as intrusive. When deciding how to leverage more passive data like employee emails and calendars, corporations must also decide what to share about the analytics they conduct using employee data. Because data scientists need robust and accurate datasets to draw quality conclusions from the analysis, companies planning to utilize data analytics in furtherance of DEI goals must take time to build trust amongst employees, particularly marginalized and minority populations.
Trust in Implementation
Beyond collecting data and employee information, companies must also consider how to build trust once they have results from the analysis. In the first instance, the data scientists responsible must be able to convey the results to leadership in a way that allows everyone to trust the findings and takeaways. This is essential to the process, particularly as the models used become increasingly complex and much more “black box” in nature. Without trust in the results, companies may hesitate to implement vital DEI initiatives due to lack of buy-in. In the second instance, in addition to being able to trust the output, employees need to believe that company leadership will act upon the data they collect and analyze. When a corporation releases a survey seeking to gather data for DEI efforts, employees that complete the survey will likely expect management to take certain actions afterwards. Thus, analysts and research should communicate to employees what actions they intend to take (and perhaps paths they choose not to pursue) in the aftermath of DEI data analytics and why.
Ultimately, as the Knowledge at Wharton article demonstrates, people analytics can play a vital role in addressing issues of bias and inequity in the workplace. In order to do this, however, companies must invest in building trust among employees at various stages of the analytics process.
Hi @user_62261c668aab2,
Thank you so much for your article. You touch on a lot of important points with regards to the ethics in ‘big data’ usage and highlight some of the key areas that it can be used for in DEI. Trust I think is a crucial aspect in approaching any sort of data analysis that may involve making decisions upon other people. As we know, biased data will only lead to biased outcomes if it is not managed appropriately. Even if the data is unbiased, if it is sourced unethically, that is also a problem. It made me think of the Cambridge Analytica saga whereby a whole range of data points were used to sway voters, which I think goes completely against ‘trust’. Even when some participants requested the data Cambridge Analytica had on them, they refused. To build ‘trust’ is also to reduce the temptation to cross ethical boundaries and harness the power of data for certain causes. This is very tricky and involves a lot of complex power dynamics.
Hi @user_62261c668aab2,
Thanks for sharing—you picked a very interesting article! Trust is an essential characteristic of any high-functioning organization, and I agree that it will be key to realizing the full potential of people analytics. As you have mentioned, achieving trust in the data collection process is difficult. Oftentimes employees do not feel comfortable sharing personal information with the company through surveys or their digital behavior through surveillance mechanisms. During this step, companies must be sure to adhere to strict data privacy standards, as any accidental breach of sensitive employee data would result in a major loss of trust. As you have also mentioned, achieving trust in the implementation stage is challenging. We’ve discussed the importance of transparency a good bit in class—companies must be forthcoming about the technical details behind the people analytics models they are using and the end goals that these models will serve. If companies can achieve trust in these two important aspects of people analytics, I too feel that it can be used to further promote crucial DEI efforts in the workplace. One specific example that comes to mind in terms of using people analytics to improve DEI that we touched on briefly in class is organizational network analysis, a technique that would enable companies to better understand how interactions between employees differ across gender and race.
This is a topic that is personally very interesting to me, as someone who wants to be a DEI practitioner. You have made great points about the moments in the people analytics that require employee trust, and I completely agree that trust must be part of the people analytics playbook. However, this made me wonder about the tactics of building trust. Francis Frei would posit that trust comes from having all 3 sides of the “trust triangle”–logos (logic), ethos (empathy), and pathos (authenticity). It seems easiest for people analytics teams to show their employee logic–that they have a sound and reasonable plan for how to collect, use, and analyze data. But I’m left wondering what organizations at an institutional level can do to show employees that they understand their concerns and that they care about their success and wellbeing? How can employees be assured that data will be used to improve their experience and not in a way that will negatively impact them?
I find it really interesting to consider the role of people analytics within DEI efforts, so thanks for sharing this article! I think a crucial part of building that trust is having representation at each stage of the process. This also addresses BRolan’s previous comment. The way an organization can demonstrate that they truly empathize with the various populations within their company is to have those voices incorporated along the pipeline. I feel that having BIPOC, women, LGBTQ+, etc. input across the life cycle of a DEI project is vital to garnering employee trust. Including diverse voices from the start can ensure projects are being structured in ways that don’t have blindspots that could affect the results. Discovering the “why” of the findings within the data will also be easier and more accurate if you have representation during the analysis phase as well. Finally, the lived perspective of various groups can make the communicating of results to leadership all the more impactful. The deeper consideration here though is to not automatically burden all underrepresented groups with this type of work where they feel obligated to be the “mouthpiece” for all who have a similar identity to them. This type of work definitely requires a thoughtful approach to collaboration.
I think this trust element is so essential in thinking about data in a DEI context. What makes it especially tricky is that we are at an inflection point on DEI with lots of thoughts on where we should be going but little agreement on how exactly we get there. Without an agreed means to an end, most uses of data will alienate/offend some. I think that for something as sensitive as DEI, for now, we should focus on using passive data to explore and expose inequities rather than active collection unless there is a clear use case. I worry that trying to over-use data to solve such a sensitive and subjective problem could frustrate and dis-empower minority voices.
Hi Kendall,
Thank you for your great post and astute identification of trust as the bedrock of people analytics within DEI efforts in particular, and of people analytics in general. In reviewing your post and previous comments, it seems as though essential components of an organization cultivating trust within the workplace—and thereby creating an environment conducive to the implementation of people analytics— are (1) operating with transparency, so that employees understand the goals of the DEI efforts and how the organization intends to achieve said goals, and (2) demonstrating respect for employee input and wellbeing by incorporating the individuals whom the DEI efforts are ostensibly aimed to serve into the effort development process, so that the DEI efforts ultimately realized are indeed catered to the needs and desires of their targets. All the while, in implementing people analytics into DEI efforts, it remains important to avoid overburdening those very same employees and to be sensitive to their data, which leads me to believe that, though potentially counterintuitive or counterproductive, providing an opportunity for employees to opt out of data being used in DEI efforts—in addition to transparency and inclusion—can serve to engender greater trust. An opt-out mechanism recognizes the data autonomy of employees and aids an organization’s ability to assess genuine buy-in for people analytics within DEI.