Two of my current projects are drafting an AI Policy for a human services organization and collecting data for a community membership organization.
Coincidentally, both the organizations came to face the same challenge in the past few days: neither of them are confident in answering questions like:
Who lives in your data?
What are your community's strengths from your data?
Do you understand the overlaps, gaps, and barriers in your existing data tools and processes?
How do you build inclusion in data?
How we collect, interpret, and use data has profound implications for our communities. While data can enhance services and inform policy, it can alienate and exclude significant portions of the population if not managed thoughtfully.
Ensuring that our data practices are inclusive and equitable is essential to fostering a sense of belonging and representation for all our community members.
In this post, I want to share with you seven questions that everyone at your table must be able to answer:
Is Your Data Representative of Your Community?
One significant risk in data collection is the potential for non-representative sampling. If our sample does not accurately reflect the diversity of our community, the insights and decisions derived from the data will be skewed. For example, imagine a city planning survey distributed primarily through online platforms. This approach might inadvertently exclude older adults who are less comfortable with digital technology, leading to city plans that do not adequately address their needs.
2. Are Your Data Collection Methods Accessible?
Accessibility is another critical factor in inclusive data practices. If data collection tools fail to consider language barriers, disabilities, and varying literacy levels, large portions of the population may be excluded. For instance, if a health survey is only available in English in a multilingual community, non-English speakers cannot participate, skewing the data and potentially leading to health initiatives that serve only some of the community.
3. Is Your Data Interpretation Free from Bias?
Even with representative data, biases in interpretation can lead to exclusion. Analysts' preconceptions and the algorithms used (if AI is involved) can introduce biases, causing misinterpretations that reinforce stereotypes or neglect the needs of marginalized groups. For example, if crime data analysis primarily focuses on low-income neighborhoods in the data, it may perpetuate the stereotype that these areas are inherently more prone to crime, without considering underlying factors like economic inequality and systemic discrimination.
4. Do You Understand the Context of Your Data?
Data points do not exist in a vacuum - understanding the broader context is crucial. Without understanding the social, cultural, and historical contexts, analysts (and others) might draw conclusions that do not reflect the lived experiences of all community members. For example, interpreting higher unemployment rates in specific communities without considering systemic barriers such as access to quality education and discrimination in hiring practices can lead to ineffective or harmful policy recommendations.
5. Are You Balancing Quantitative and Qualitative Data?
While quantitative data provides valuable insights, over-reliance can exclude nuanced, qualitative perspectives. For example, statistics on school performance might show overall improvement but miss out on the specific challenges minority students face. Incorporating qualitative data, such as interviews and focus groups, can capture these nuanced experiences, leading to more comprehensive and inclusive insights.
6. Are You Communicating Results Transparently?
How we communicate data results also plays a crucial role in inclusivity. Using clear, jargon-free language and sharing findings in accessible formats ensures all community members can understand and engage with the data.
7. Are Your Data Collection Methods Flexible?
Flexibility in data collection methods can significantly enhance inclusivity. For instance, allowing respondents to choose between online, phone, or in-person survey options can accommodate different preferences and capabilities. Moreover, regularly reviewing and updating data collection tools to address emerging accessibility needs can help maintain inclusivity over time.
Join the 3-hour LIVE workshop on Advancing Equitable Data Collection on Wednesday, July 31st!
Here is a simple infographic to remind you that de-weaponizing data means actively working towards placing systems, structures, and tools that do not create harm, barriers, and divides.
Comments