12.07.2015 Views

Fourth National Incidence Study of Child Abuse and Neglect (NIS–4)

Fourth National Incidence Study of Child Abuse and Neglect (NIS–4)

Fourth National Incidence Study of Child Abuse and Neglect (NIS–4)

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

the population. The estimated totals are the weighted totals <strong>of</strong> the children’s records inthe final analysis file, subset to those children who experienced the maltreatment inquestion. <strong>Incidence</strong> rates are computed by dividing the estimated total number <strong>of</strong>children who experience the maltreatment in question by the number <strong>of</strong> children in theU.S. population 16 <strong>and</strong> by 1,000. This provides the rate per 1,000 children in the generalpopulation. The ensuing chapters present estimates for the incidence maltreatment asdefined by the Harm St<strong>and</strong>ard <strong>and</strong> by the Endangerment St<strong>and</strong>ard, giving the estimatednumbers <strong>of</strong> children who experienced the different categories <strong>of</strong> maltreatment as well asthe corresponding rates per 1,000 children in the population. Analyses also producedestimated totals <strong>and</strong> rates for subgroups <strong>of</strong> children, defined by various child <strong>and</strong> familycharacteristics. In each case, the denominator used to compute the rate reflected thenumber <strong>of</strong> children in the general population with the corresponding characteristic. 17The national estimates in this report derive from a sample <strong>of</strong> maltreatedchildren. The final NIS–4 analysis file includes a sample <strong>of</strong> 12,408 children whoexperienced Endangerment St<strong>and</strong>ard maltreatment, 5,226 <strong>of</strong> whom also experiencedHarm St<strong>and</strong>ard maltreatment. Any estimate based on a sample has some degree <strong>of</strong>associated uncertainty. Sample-based studies index the amount <strong>of</strong> uncertainty onestimates with the “st<strong>and</strong>ard error <strong>of</strong> estimate,” or S.E. The square <strong>of</strong> the st<strong>and</strong>ard erroris the variance. The st<strong>and</strong>ard error or variance indicates the precision <strong>of</strong> the estimate. Itis critical in underst<strong>and</strong>ing how reliable an estimate is <strong>and</strong> in deciding whether anobserved difference in estimated rates is “statistically significant,” or whether it couldsimply be due to sampling error (due to chance fluctuations). 18 Statisticians recommend16 In order to accommodate any changes in the size <strong>of</strong> the overall child population in the U.S. between theNIS–4 reference periods (fall 2005 <strong>and</strong> spring 2006), NIS–4 incidence rates used the average <strong>of</strong> the July2005 <strong>and</strong> July 2006 U.S. Census Bureau annual estimates as the population denominator (U.S. CensusBureau, 2008a).17 To provide estimates for the broad range <strong>of</strong> characteristics <strong>of</strong> interest, the NIS–4 used populationstatistics from various sources; subsequent chapters provide those details. In all cases, those sourcesprovided the information about the distribution <strong>of</strong> the characteristics in the child population. To establishthe specific denominators for computing rates, that distribution was applied to the NIS–4 overall childpopulation denominator (based on the average <strong>of</strong> July 2005 <strong>and</strong> July 2006 child population estimates, asdescribed in an earlier footnote). This ensured that the subgroup denominators summed to the NIS–4child population denominator.18 Statistically significant differences have a very low probability <strong>of</strong> emerging by chance (probability <strong>of</strong> 5%or less, or pp>.05).2–18

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!