Select All | Deselect All

Methodology
Disproportionality & Disparity Indices

Notes on Source Data

IMPORTANT NOTE: The following methodology may include references to report features not available on the public site version. On the public site, masking is performed to protect the privacy of individuals served by CDSS and comply with CDSS data de-identification guidelines. Values of 1 to 10 and calculations based on values of 1 to 10 are masked ('M' or '*'). In stratified views of the data, additional values (the lowest available) are masked to prevent calculation of values 1 to 10.

These reports compute ethnic Disproportionality and Disparity Indices based on population projections from the California Department of Finance. Please refer to the Population Data Index methodology for information on these data.
https://ccwip.berkeley.edu/cwscmsreports/methodologies/default.aspx?report=Population

Disproportionality is the degree to which groups of children are present in the child welfare system at rates that are higher or lower than their presence in the general population. Although we compute a Disproportionality Metric (DM) for descriptive purposes, we do not recommend that it be used as a measure for comparing racial representation. The Disproportionality Metric is problematic due to its mathematical construction (which imposes a theoretical maximum based on the size of the minority group population). (For additional details, please see Shaw, T. V., Putnam-Hornstein, E., Magruder, J., & Needell, B. (2008). Measuring racial disparity in child welfare. Child Welfare, 87(2), 23-36.)

Instead, we recommend using a relative risk or relative rate ratio, which we refer to as a Disparity Index (DI). Disparity is used to refer to the lack of equity between groups and derived in one of two ways: either by computing a ratio based on the rates per 1,000 for two groups or by computing a ratio based on the disproportionality metrics for two groups.

The Disparity Index (DI) can in be interpreted in the following ways:

  • A DI of 1.00 means that the risk of the event is identical in two groups.
  • A DI that is less than 1.00 means that the risk is lower in the selected group versus a comparison.
    • For example, a DI of 0.30 indicates that the risk of the outcome is reduced to 30% for the selected group versus the comparison. A DI of 0.30 may also be interpreted as indicating that the risk is reduced by 70% for the selected group versus the comparison, and stated in a way the expresses the difference. For example, a DI of 0.30 indicates that the selected group is 70% less likely than the comparison group to experience the outcome.
  • A DI that is greater than 1.00 means that the risk is greater in the selected group versus a comparison.
    • When the DI is greater than 1.00 but less than 2.00, the index may be interpreted as indicating the greater likelihood as a percentage. For example, a DI of 1.34 indicates that the selected group is 34% more likely than the comparison group to experience the outcome.
    • When the DI is greater than 2.00, the index should be interpreted as indicating how many times as likely the selected group is to experience the outcome. For example, a DI of 2.34 indicates that the selected group is more than two times likely to experience the outcome than the comparison group.

These reports are based on the population of children who had child welfare system contact during a given year while between the ages of 0 and 17. Indices are stratified at the level of child welfare contact: Allegations, Substantiated Allegations, Entries, and In Care. Allegations are unduplicated counts of children for whom a child maltreatment allegation was received during the analysis year. Substantiated Allegations are unduplicated counts of children with a substantiated case during the analysis year. Entries to care are based on the count of unique children who entered care (both entries and re-entries) without restriction on the days spent in care. Entries and In Care Rates are restricted to cases supervised by a Child Welfare Agency.

Children with missing county assignment are included in the statewide calculation. Given the methods outlined above, county values may not sum to statewide total.

Disparity is computed for each ethnicity as it compares with every other ethnicity (i.e., Black vs. White, Black vs. Hispanic, etc.). Please note that these Indices should be interpreted cautiously, especially at the county level. As is the case whenever a rate is computed based on a small population, large fluctuations and margins of error are common.

Cells containing a period (".") represent a value of zero. In cells representing quotients, a period may also indicate the indeterminate form 0/0.