Skip to main content



Child-Maltreatment-Research-L (CMRL) List Serve

Database of Past CMRL Messages

Welcome to the database of past Child-Maltreatment-Research-L (CMRL) list serve messages. The table below contains all past CMRL messages (text only, no attachments) from Nov. 20, 1996 - December 22, 2017 and is updated quarterly.

Instructions: Postings are listed for browsing with the newest messages first. Click on the linked ID number to see a message. You can search the author, subject, message ID, and message content fields by entering your criteria into this search box:

Message ID: 9592
Date: 2014-04-03

Author:Joseph A. Mienko

Subject:Re: Lifetime incidence of child maltreatment

Sheri and Emily, You are both correct that this is an overestimate of the lifetime incidence of child maltreatment (sorry for confusion there as this is what Jessica had originally asked for). My calculations were for the related measurement of the total maltreatment rate. The total maltreatment rate would report the expected total number of maltreatment incidents for a child over their lifetime. When you use the CB reports (which only give unique counts per year unless you dig into the actual NCANDS data like Emily has done), you get an underestimate of the total maltreatment rate but an overestimate of lifetime incidence. On Wed, Apr 2, 2014 at 8:19 PM, D F MCMAHON > wrote: Clarification please: You state " I should also note that this number is an underestimate as the CB numbers are unique counts of maltreatment for individual children within a given year. Some of the children will have repeated instances of maltreatment within a given year." I take your statement to mean that if, say, child X experiences maltreatment twice in the year that would result in a count of 2 (possibly assigned to different age groups depending on when the events occurred relative to the child's birthday). If the question is how many children experience maltreatment then wouldn't this be an over-, not underestimate? (Leaving out stuff like population changes due to mortality, children moving into or out of Washington state, and what have you, the calculations you do would result in an actual count of --which? maltreatment episodes or children maltreated?) Sheri McMahon North Dakota ________________________________ Date: Wed, 2 Apr 2014 12:26:24 -0700 Subject: Re: Lifetime incidence of child maltreatment From: mienkoja@uw.edu To: child-maltreatment-research-l@list.cornell.edu You can also tackle this with a synthetic cohort approach. That is, you can take exposure information for children in a given geographic region for a particular point in time (e.g. census data) and counts of maltreatment for the same time period to calculate what could be called a total maltreatment rate (similar to a total fertility rate for demographers in the room). I had to do a similar calculation recently for Washington State and will share the basic process below. I will note that Washington tends to have a relatively low substantiation rate. So, I would not treat my calculations as representative of the country as a whole. You would want to repeat these calculations for several states or for the entire country. The computer code shown is for R, but this could easily be accomplished in a spreadsheet. Washington State total maltreatment rate calculation process Get age-specific population data We can get age-specific 2010 Washington population data from the following site . The relevant values for 2010 are the population ages from 0 through 17. We can assign these values to a person-years variable called pyWA2010. pyWA2010 <- c(87016, 87607, 89399, 89097, 86538, 86550, 85890, 84916, 85058, 87463, 88129, 87626, 87497, 87399, 87582, 89006, 90813, 93768) If we had multiple years of data, we would multiply these values by the count of observation years (or just sum population estimates for each individual year). However, since we are only dealing with a single year (2010), our person-years are just the age-specific population estimates. Get age-specific maltreatment data We can get Washington state maltreatment information from the CB reporting site . The relevant values for 2010 are given for each age range (0 through 17). We can assign those to another variable (malWA2010) as shown below. malWA2010 <- c(11.1, 8.6, 7.9, 7.6, 6.3, 6, 6.2, 5.5, 4.9, 4.6, 4.4, 4.1, 4, 4.6, 4.3, 3.8, 3.6, 2.1) Before we can use this CB data, however, we need to convert it into a count. The CB’s report displays information in percentages, but it also gives a total count of maltreatment for 2010 (6,593). In order for us to make use of this data, we can convert our percentages to a proportion (divide by 100) and multiply by the 6,593 value. This calculation is shown below. malWA2010_tot <- 6593 malWA2010_cnt <- (malWA2010/100) * malWA2010_tot Make the calculations If we divide the age-specific maltreatment variable (malWA2010_cnt) by the person-years variable (pyWA2010), we will get the age-specific maltreatment rates for Washington State in 2010 (asWA2010). asWA2010 <- malWA2010_cnt/pyWA2010 asWA2010 ## [1] 0.008410 0.006472 0.005826 0.005624 0.004800 0.004571 0.004759 ## [8] 0.004270 0.003798 0.003468 0.003292 0.003085 0.003014 0.003470 ## [15] 0.003237 0.002815 0.002614 0.001477 This information is useful in its own right as a measure of the age pattern of maltreatment - the relative frequency of maltreatment among children of different age. We can also use this information to calculate a total maltreatment rate for our synthetic cohort by summing the values. tmWA2010 <- sum(asWA2010) tmWA2010 ## [1] 0.075 The resulting value is the total maltreatment rate per child in Washington State in 2010. If we multiply by 1000, we will get a total maltreatment rate per 1000 children (i.e. 75). tmWA2010 * 1000 ## [1] 75 This number can be interpreted as the average rate of maltreatment for a child from birth until age 18. To be clear, this is a hypothetical measurement. It represents the expected rate of childhood maltreatment that would occur if the child experiences the exact current age-specific maltreatment rates throughout their childhood and they survive from birth through the end of their childhood. I should also note that this number is an underestimate as the CB numbers are unique counts of maltreatment for individual children within a given year. Some of the children will have repeated instances of maltreatment within a given year. For an accessible introduction to this and related techniques, I highly recommend The Tools and Methods of Demography . On Mon, Mar 31, 2014 at 1:56 PM, Chris Lloyd > wrote: Katherine M. Keyes et al. have a paper in the British Journal of Psychiatry titled Childhood Maltreatment and the Structure of Common Psychiatric Disorders (2012, vol. 200, pp. 107-115) that estimates lifetime incidence of maltreatment among what they claim is a nationally representative sample of US adults (it probably is, but I haven't read that methodology myself). It is retrospective by self-report but uses an instrument with good psychometric properties. Also the NIS-4 (and presumably it's predecessors) has an estimate of that as well, I believe. Here's the reference: Sedlak, A.J., Mettenburg, J., Basena, M., Petta, I., McPherson, K., Greene, A., and Li, S. (2010). Fourth National Incidence Study of Child Abuse and Neglect (NIS–4): Report to Congress. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families. and here's the URL to it: http://www.acf.hhs.gov/programs/opre/research/project/national-incidence-study-of-child-abuse-and-neglect-nis-4-2004-2009 There are others, but these are the two I can put my hands on right now. Chris On Thu, Mar 20, 2014 at 12:19 PM, Jessica Bartlett > wrote: Hello everyone, Does anyone know of any national studies that report on lifetime incidence of child maltreatment (birth through adolescence, preferably substantiated reports)? We have determined a rate in the Early Head Start randomized sample (birth to approximately age 15) but we are interested in samples for comparison and can only find information on annual incidence. Thank you! Jessica --------------------------------------------- Jessica Dym Bartlett M.S.W., Ph.D. Project Director, Research and Evaluation Brazelton Touchpoints Center Boston Children's Hospital 1295 Boylston St., Suite 320 Boston, MA 02215 (857) 218-4375 jessica.bartlett@childrens.harvard.edu -- E. Christopher Lloyd, PhD, LCSW Associate Professor School of Social Work University of Arkansas at Little Rock 2801 South University Avenue Little Rock, AR 72204 501.569.8464 -- Joe Mienko Policy Research Scientist Partners for Our Children | University of Washington School of Social Work Box 359476 | Seattle, WA 98195-9476 T: 206-221-8725 | F: 206-221-3155 www.partnersforourchildren.org -- Joe Mienko Policy Research Scientist Partners for Our Children | University of Washington School of Social Work Box 359476 | Seattle, WA 98195-9476 T: 206-221-8725 | F: 206-221-3155 www.partnersforourchildren.org

Sheri and Emily, You are both correct that this is an overestimate of the lifetime incidence of child maltreatment (sorry for confusion there as this is what Jessica had originally asked for). My calculations were for the related measurement of the total maltreatment rate. The total maltreatment rate would report the expected total number of maltreatment incidents for a child over their lifetime. When you use the CB reports (which only give unique counts per year unless you dig into the actual NCANDS data like Emily has done), you get an underestimate of the total maltreatment rate but an overestimate of lifetime incidence. On Wed, Apr 2, 2014 at 8:19 PM, D F MCMAHON > wrote: Clarification please: You state " I should also note that this number is an underestimate as the CB numbers are unique counts of maltreatment for individual children within a given year. Some of the children will have repeated instances of maltreatment within a given year." I take your statement to mean that if, say, child X experiences maltreatment twice in the year that would result in a count of 2 (possibly assigned to different age groups depending on when the events occurred relative to the child's birthday). If the question is how many children experience maltreatment then wouldn't this be an over-, not underestimate? (Leaving out stuff like population changes due to mortality, children moving into or out of Washington state, and what have you, the calculations you do would result in an actual count of --which? maltreatment episodes or children maltreated?) Sheri McMahon North Dakota ________________________________ Date: Wed, 2 Apr 2014 12:26:24 -0700 Subject: Re: Lifetime incidence of child maltreatment From: mienkojauw.edu To: child-maltreatment-research-llist.cornell.edu You can also tackle this with a synthetic cohort approach. That is, you can take exposure information for children in a given geographic region for a particular point in time (e.g. census data) and counts of maltreatment for the same time period to calculate what could be called a total maltreatment rate (similar to a total fertility rate for demographers in the room). I had to do a similar calculation recently for Washington State and will share the basic process below. I will note that Washington tends to have a relatively low substantiation rate. So, I would not treat my calculations as representative of the country as a whole. You would want to repeat these calculations for several states or for the entire country. The computer code shown is for R, but this could easily be accomplished in a spreadsheet. Washington State total maltreatment rate calculation process Get age-specific population data We can get age-specific 2010 Washington population data from the following site . The relevant values for 2010 are the population ages from 0 through 17. We can assign these values to a person-years variable called pyWA2010. pyWA2010 <- c(87016, 87607, 89399, 89097, 86538, 86550, 85890, 84916, 85058, 87463, 88129, 87626, 87497, 87399, 87582, 89006, 90813, 93768) If we had multiple years of data, we would multiply these values by the count of observation years (or just sum population estimates for each individual year). However, since we are only dealing with a single year (2010), our person-years are just the age-specific population estimates. Get age-specific maltreatment data We can get Washington state maltreatment information from the CB reporting site . The relevant values for 2010 are given for each age range (0 through 17). We can assign those to another variable (malWA2010) as shown below. malWA2010 <- c(11.1, 8.6, 7.9, 7.6, 6.3, 6, 6.2, 5.5, 4.9, 4.6, 4.4, 4.1, 4, 4.6, 4.3, 3.8, 3.6, 2.1) Before we can use this CB data, however, we need to convert it into a count. The CB’s report displays information in percentages, but it also gives a total count of maltreatment for 2010 (6,593). In order for us to make use of this data, we can convert our percentages to a proportion (divide by 100) and multiply by the 6,593 value. This calculation is shown below. malWA2010_tot <- 6593 malWA2010_cnt <- (malWA2010/100) * malWA2010_tot Make the calculations If we divide the age-specific maltreatment variable (malWA2010_cnt) by the person-years variable (pyWA2010), we will get the age-specific maltreatment rates for Washington State in 2010 (asWA2010). asWA2010 <- malWA2010_cnt/pyWA2010 asWA2010 ## [1] 0.008410 0.006472 0.005826 0.005624 0.004800 0.004571 0.004759 ## [8] 0.004270 0.003798 0.003468 0.003292 0.003085 0.003014 0.003470 ## [15] 0.003237 0.002815 0.002614 0.001477 This information is useful in its own right as a measure of the age pattern of maltreatment - the relative frequency of maltreatment among children of different age. We can also use this information to calculate a total maltreatment rate for our synthetic cohort by summing the values. tmWA2010 <- sum(asWA2010) tmWA2010 ## [1] 0.075 The resulting value is the total maltreatment rate per child in Washington State in 2010. If we multiply by 1000, we will get a total maltreatment rate per 1000 children (i.e. 75). tmWA2010 * 1000 ## [1] 75 This number can be interpreted as the average rate of maltreatment for a child from birth until age 18. To be clear, this is a hypothetical measurement. It represents the expected rate of childhood maltreatment that would occur if the child experiences the exact current age-specific maltreatment rates throughout their childhood and they survive from birth through the end of their childhood. I should also note that this number is an underestimate as the CB numbers are unique counts of maltreatment for individual children within a given year. Some of the children will have repeated instances of maltreatment within a given year. For an accessible introduction to this and related techniques, I highly recommend The Tools and Methods of Demography . On Mon, Mar 31, 2014 at 1:56 PM, Chris Lloyd > wrote: Katherine M. Keyes et al. have a paper in the British Journal of Psychiatry titled Childhood Maltreatment and the Structure of Common Psychiatric Disorders (2012, vol. 200, pp. 107-115) that estimates lifetime incidence of maltreatment among what they claim is a nationally representative sample of US adults (it probably is, but I haven't read that methodology myself). It is retrospective by self-report but uses an instrument with good psychometric properties. Also the NIS-4 (and presumably it's predecessors) has an estimate of that as well, I believe. Here's the reference: Sedlak, A.J., Mettenburg, J., Basena, M., Petta, I., McPherson, K., Greene, A., and Li, S. (2010). Fourth National Incidence Study of Child Abuse and Neglect (NIS–4): Report to Congress. Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families. and here's the URL to it: http://www.acf.hhs.gov/programs/opre/research/project/national-incidence-study-of-child-abuse-and-neglect-nis-4-2004-2009 There are others, but these are the two I can put my hands on right now. Chris On Thu, Mar 20, 2014 at 12:19 PM, Jessica Bartlett > wrote: Hello everyone, Does anyone know of any national studies that report on lifetime incidence of child maltreatment (birth through adolescence, preferably substantiated reports)? We have determined a rate in the Early Head Start randomized sample (birth to approximately age 15) but we are interested in samples for comparison and can only find information on annual incidence. Thank you! Jessica --------------------------------------------- Jessica Dym Bartlett M.S.W., Ph.D. Project Director, Research and Evaluation Brazelton Touchpoints Center Boston Children's Hospital 1295 Boylston St., Suite 320 Boston, MA 02215 (857) 218-4375 jessica.bartlettchildrens.harvard.edu -- E. Christopher Lloyd, PhD, LCSW Associate Professor School of Social Work University of Arkansas at Little Rock 2801 South University Avenue Little Rock, AR 72204 501.569.8464 -- Joe Mienko Policy Research Scientist Partners for Our Children | University of Washington School of Social Work Box 359476 | Seattle, WA 98195-9476 T: 206-221-8725 | F: 206-221-3155 www.partnersforourchildren.org -- Joe Mienko Policy Research Scientist Partners for Our Children | University of Washington School of Social Work Box 359476 | Seattle, WA 98195-9476 T: 206-221-8725 | F: 206-221-3155 www.partnersforourchildren.org