Appendix C Technical Notes
In this section, notes are provided on the data validation and analytic procedures used in this report.
Validation Steps
Several steps occur prior to the analysis of the Summary Data Component (SDC) data. The 15 data items in each State submission are validated. Each submission is checked to ensure internal consistency and to gauge whether data are within an expected range. In addition, data are compared to the previous year's submission on the same item. State comments from the current reporting year and previous years are reviewed and updated to further clarify the data submission.
Analytical Methods
The 1995 National Findings are based on data submitted to the NCANDS by all 50 States, the District of Columbia, and the Armed Services. Data from Montana and West Virginia are from 1994. Guam, Puerto Rico, and the Virgin Islands were unable to provide data for 1995.
In the data tables, data from the Armed Services are provided below the national totals, since Armed Services personnel also report to State agencies.
Unless otherwise noted, rates were calculated using population averages for the reported population and were based on the number per 1,000 children younger than age 18 in the reporting States. The population of children younger than 18 years of age for each State was derived from data for 1995 supplied by the U.S. Bureau of the Census.
Data Items
Clarification of the analyses of selected data items is provided below. The item numbers correspond to the item numbers in the SDC data tables (section 3).
Item 1: Reports
States count reports alleging child maltreatment by child or by incident. Screened-out reports were not included in the number of reports provided. Only those reports that were referred for investigation were counted.
Item 2: Children Subject of a Report
Item 2 presents duplicated counts of children who were the subject of a report. The number of children was estimated for those States unable to provide these data. Using the data from States that reported incident-based data and the duplicated number of children in these reports, a mean ratio of incident-based reports to duplicated counts of children was computed. The Tukey biweight estimator was computed on the resultant mean rate. A multiplier of 1.62 was used to multiply the incident-based reports for those States that did not provide duplicated child data. The resulting products were used as estimates. The rate of duplicated children reported as alleged victims of maltreatment was calculated using the State child population. The rate is somewhat higher than a rate of unduplicated children per 1,000 children.
Item 3: Reports by Source
The percentages discussed in the report are based on the distribution of report sources from 42 States (1,486,605 reports). Most States collect data on only the primary source of the report, but some States record multiple sources of the same report. The same categories of report sources are not used by all States. For example, many States do not classify substitute care providers and perpetrators as report sources. Some States classify parents with other relatives.
Item 4: Investigations by Disposition
The percentages discussed in the report are based on the total number of dispositions (1,675,372) in 48 States. The percentage of intentionally false reports is based on the number of unsubstantiated reports that were considered to be "intentionally false" by five States.
Item 5: Children and Families Subject of a Completed Investigation
Data on children subject of a completed investigation are analyzed using item 6. Item 5 provides an indication of the ratio of children to families in the States that are able to provide both children subject of an investigation and families subject of an investigation in item 5.
Item 6: Children by Disposition
Percentages of children by disposition have been calculated using only those States that reported on both substantiated and unsubstantiated dispositions. The percentages are not significantly different from data submitted for 1994 if this subset of States is used.
The rates of victimization discussed in the report are based upon the 1,000,502 substantiated or indicated child victims in 49 States. The child population in these States (67,046,118) was used as a denominator and the resulting quotient was multiplied by 1,000 to express the rate in terms of every 1,000 children. The rate for the two-tier States is based upon 688,752 substantiated children among a child population of 53,235,427 children in these 38 States. The three-tier rate is based on 311,750 substantiated and indicated children among the 13,810,691 children in these 11 States. The calculations of these overall rates included submissions of duplicated or unduplicated numbers from the States and should be considered as best estimates.
In order to compare child reporting rates to child substantiation rates, only data from the 37 States that provided duplicated counts for both items and provided the number of unsubstantiated children were used in the analysis.
Item 7: Types of Maltreatment
Percentages discussed in this report are based on the number of maltreatments suffered by the confirmed victims (1,000,502 substantiated and indicated children) in 49 States. Rates for maltreatment types for every 1,000 children are based on the child population (67,046,118) in these States. Most States were able to collect and report data on the multiple types of maltreatment suffered by a child. However, many States do not record "unknown" as a type of maltreatment, and some States do not use the "other" category.
Item 8: Age of Victims
Most States provided age data in 1-year increments, but some States reported data by age group (e.g., ages 0-2, ages 3-5). For these States, the age of victims was estimated for each year by dividing the grouped figure by the number of intervals included in that grouping. Percentages in the report are based on the total number of victims (833,115) for whom age data were provided.
Item 9: Sex of Victims
Percentages in the report are based on the total number of victims (834,174) for whom age was reported.
Item 10: Race/Ethnicity of Victim
Percentages in the report are based on the total number of victims (822,609) for whom race/ethnicity was reported.
Item 11: Victims Removed From the Home
The percentage of victims removed from the home is based on the responses for the total number of victims (852,540) in the States that reported this item. For some States, the number of victims removed from the home included the siblings of these victims.
Item 12: Court Action Initiated
The percentage of victims for whom court action was initiated is based upon the responses for the total number of victims (487,233) in the States that reported this item.
Item 13: Receiving Additional Services
For most States, the numbers do not include information on services that are not captured on the State child protective services information system. These counts are from States that provided unduplicated numbers and from States that provided duplicated numbers.
Item 14: Died from Abuse or Neglect
The fatality figures refer primarily to those victims who were known to the child protective services agency. States are continuing to develop interagency child death investigation teams that might identify instances of death due to maltreatment more completely. Montana and West Virginia did not submit 1995 data. Therefore, 1994 data were used for those two states.
The rate of fatalities is based on the child population of those States that submitted fatality data. For this calculation, a denominator of 61,241,428 children is used. The resulting quotient is multiplied by 100,000 to obtain a rate expressed in terms of every 100,000 children.
Item 15: Relationship of Perpetrator to Victim
States that count perpetrators by incident may count fewer perpetrator relationships than States that use victim-based reporting, since an incident may include more than one child victim. Percentages are based on a total of 653,736 perpetrators.
