Diversity and Inclusion are each meaty topics on their own - building a survey to measure these two constructs and help organizations move forward was quite the task! In this article, we want to explain our intended purpose for the survey, how we created the questions and demographics for it, and why we chose to format it as an unattributed survey.
Purpose: Why you should (and shouldn't) run the Diversity and Inclusion Survey
Running a single engagement survey won’t solve all of your organization’s issues. It will, however, start conversations and get you on the path forward. Similarly, a Diversity and Inclusion Survey will bring data to the table; highlighting if, where, and how different groups of employees experience your company culture differently.
This template combines the measurement of evidence-based and research-driven constructs, inclusive demographics, global benchmarking, text analytics, and an action framework driven by collective intelligence. By collecting, understanding, and acting on feedback related to diversity and inclusion, you can expect to learn how employees with intersectional identities experience your organization's culture.
We do not recommend using the survey for other intended purposes - for example, we do not recommend using this survey to solicit feedback on sexual harassment. If you intend to use the Diversity and Inclusion Survey for a non-obvious purpose (or find yourself using lots of unique survey questions that are different from what is in the template), please check in with firstname.lastname@example.org to see if we can assist you.
Questions: How we identified the survey's D&I constructs
For a full list of the Diversity and Inclusion Survey template questions, developed in partnership with Paradigm, please see the Template Library inside your account.
We identified 7 constructs of Diversity and Inclusion – areas where we know negative stereotypes impugn minority groups and can prevent all employees from achieving their full potential; but also where optimism in these areas can lead to new pathways for innovation and business success. These constructs are Belonging, Fairness, Opportunities and Resource, Decision Making, Diversity, Voice, and Contribution to a Broader Purpose. The questions in each of these constructs originate from applied experience or modern research.
1. Applied Experience
Some questions come directly (or very closely) from our engagement template. They might not even seem like they are a “diversity and inclusion” question. For example:
"ACME enables me to balance work and personal life"
- Culture Amp has surveyed over 2 million people, and we can see that men score around 10 points higher than women and non-binary employees.
- We know from applied experience that minority groups may have a harder time adapting to rigorous work obligations.
- Sure enough, we’re seeing single parents score lower than married/partnered parents – solidifying this question on our Diversity and Inclusion template.
2. Modern Research
Some questions were adapted from modern research topics, such as Carol Dweck’s work around the Growth Mindset, or Geoff Cohen and Greg Walton’s research around belonging. Consider this question:
"Even when something bad happens (e.g., when I get critical feedback from my manager, I have a negative social interaction with a peer, etc.), I don’t question whether or not I belong at ACME"
- The line of research around "Belonging Uncertainty" shows that majority group members do not question their belonging when they encounter difficulties or setbacks. Minority group members, often the subject of negative stereotypes, are vigilant for cues that they belong.
- When their belonging is doubted, rather than assumed, critical feedback and other business interactions present specific challenges for employees that are underrepresented in a setting.
- Questions pulled from modern research were tested extensively with Amazon mTurk data and subsequently with a pilot group of companies.
Demographics: How we created the survey's demographics and the messages they send
Keep in mind that, similar to all employee surveys, the questions you ask in a survey are bi-directional; while you are collecting feedback from employees, you are also communicating to employees with the questions and demographics you include (and those you leave out). We put a lot of time into our demographics because they send a strong signal to your employees about your diversity ideology. With this thinking in mind, many practical and pragmatic considerations factor into how we iterate on our template demographics.
1. More Inclusive Options
If you decide to list “Man” and “Woman” as the only options for gender identity, you are communicating to your employees that you believe that gender is binary. The Diversity and Inclusion Survey template provides 5 gender identities to choose from, 7 sexual orientations, and more race/ethnicity options than what is provided by the EEOC. When choosing the demographics for your own Diversity and Inclusion Survey, we suggest administrators consult with Employee Resource Groups, your legal team, or any other resources you have available to you if you want to make any changes to our template's demographics.
2. Statistical Validity
Ultimately, our unique identities can be complex and checking a box doesn’t always feel inclusive to everyone. We could have provided even more options in some areas, however, we had to consider creating statistically valid samples for reporting. We provided demographics at the level where we believed we would have enough responses for statistically robust reporting.
Consider your own employee base and where you may need to condense options or be able to expand them. In the technology industry, where Asians are well represented, we suggest using our demographics that split East Asian, Southeast Asian, and South Asian. Our template lists “Black/African-American” as a single choice; if you have a large population and want to understand the nuance behind how this population is feeling, consider splitting out Caribbean or other identities from the African Diaspora.
3. More than visible factors
The Diversity and Inclusion Survey template promotes the thinking that diversity goes beyond visible traits. By including structural factors, organizations can foster a culture that includes a broad and encompassing definition of diversity. We included Veteran Status, Family Status, Disability Status, and Socio-economic Status (measured by the degree of education obtained by parents/legal guardians).
You may decide to add more factors, such as religion/faith or political affiliation, as we have seen this in some Diversity and Inclusion surveys that have been run on the Culture Amp platform. While these factors are not listed in our template, you may decide to include them in your version after consulting with your team.
Format: Why the survey is set up as an unattributed survey
Unlike other survey templates on our platform, this survey’s default setting is “unattributed”, also known as the completely anonymous format. You can read more about this format here – it does not operate like a traditional engagement survey.
Given the sensitive nature of the topic, we believe that running the Diversity and Inclusion Survey in the unattributed format gives even more assurance to survey takers that there is no identifiable information is tracked to survey responses or accessible to administrators.
Some organizations intend to use the Diversity and Inclusion Survey to test the impact of interventions and nudges related to diversity and inclusion. For example, an organization may want to survey as a baseline, expose half of the workforce to unconscious bias training, and survey again (seeing if the unconscious bias training had positive impacts on the 50% of employees that were granted access to the training). This sort of efficacy analysis is not possible under the unattributed/completely anonymous format because responses are not tied to individuals in any capacity. If you intend to test interventions, please reach out to email@example.com to change the template to an attributed survey, and see this article for more information about running it attributed.