AISP Toolkit Feb25 2025 - Flipbook - Page 52
Positive and Problematic Practices:
Racial Equity in Data Analysis
CENTERING RACIAL EQUITY THROUGHOUT THE DATA LIFE CYCLE
POSITIVE PRACTICE
48
PROBLEMATIC PRACTICE
Engaging a range of expertise (e.g., subject matter,
methods, lived experience) to ensure that the data
analysis approach is appropriate for the research
questions and local context.
Not building in time and resources for data
discovery and exploratory data analysis before
diving into formal analyses.
Making sure the people responsible for data
wrangling understand the datasets, variables, and
analytic plan.
Failing to document potential data quality issues,
processes used to remedy issues, and how these
may impact analyses.
Designing mixed methods analytic plans that
purposefully seek out and combine data sources
to better understand social problems through
“multiple ways of knowing.”
Creating analytic approaches that are
indecipherable to nonexperts without explaining
them clearly for a general audience.
Disaggregating data and analyzing intersectional
experiences without compromising data privacy.
Assuming that data representing small populations
are not meaningful to analyze because of
statistical insigni昀椀cance.
Training analytic staff on best practices for
analyzing RELD and SOGIE data (see RELD & SOGIE
Data Standards Framework)
Failing to recognize the distinctiveness of
identities and intersectional experiences (e.g.,
assuming gender queer youth, youth of color,
and gender queer youth of color all have similar
reasons for program nonparticipation).
Carefully considering how subgroups are de昀椀ned,
analyzed, and reported, with an emphasis on
asset-based framing.
Not using appropriate comparison groups to
contextualize 昀椀ndings (e.g., assuming White
outcomes are normative).
Highlighting structural factors within analyses
(e.g., overlaying redlining data to correlate place to
outcomes).
Using one-dimensional data to propel an agenda
(e.g., use of student test scores in isolation from
contextual factors such as teacher turnover or
school-level demographics).
Drawing on community member expertise when
interpreting analyses and identifying root causes
of 昀椀ndings.
Interpreting results without examining larger
systems, policies, and social conditions that
contribute to disparities in experiences and
outcomes (e.g., poverty, housing segregation,
access to education).
Empowering professionals and community
members to use the results of analyses to improve
their work and their communities.
Analyzing data with no intent to drive action or
change that bene昀椀ts those represented in the
data.
