Data Analysis
Prevalence and Agreement Rates
Simple frequencies are used to describe the prevalence and concordance rates. We categorize as concordant those cases where both the official investigated records and youth self-reports indicate that a particular form of maltreatment occurred before or after a youth's entry into out-of-home care. Where there is concordance between sources, we distinguish between cases where there is agreement between data sources that maltreatment did or did not occur. We distinguish between two distinct types of disagreement between the official records and youths’ reports. Cases where the official investigated maltreatment records indicate an allegation occurred and a youth reported that it did not are referred to as ‘youth false negatives.’ In other words, we assume that the youth in question did indeed experience a particular form of maltreatment but that for one or more reasons the young person did not respond affirmatively to the survey questions used to identify that form of maltreatment. For instance, survey item wording may not have captured a variety of maltreatment that is identified in official reports. Cases where official records indicate that maltreatment did not occur and a youth report indicates that it did occur are referred to in this study as ‘system false negatives.’ In these cases, we assume that the youth in question did indeed experience a particular form of maltreatment but that for one or more reasons the official maltreatment records do not reflect this. Frequencies are used to describe the rate of agreement (i.e. occurred and/or did not occur), rate of youth false positives, and the rate of system false negatives. Finally, we combine sources to report rates of maltreatment across both sources of information.