Child Victimization and polyvictimization in China: Prevalence and health correlates Dr. Edward Chan The University of Hong Kong

Optimus Study in China A Study of Child Victimization & Sexual Abuse • Funded by the UBS Optimus Foundation, Switzerland. • Epidemiological study based on representative samples from households and schools in Hong Kong and 5 Mainland cities: Shanghai, Shenzhen, Tianjin, Wuhan and Xi’an. • Period of data collection: Nov., 2009 – July, 2010 2

Household survey • The household-based approach, with both children aged 15 – 17 and parents with children aged below 15 were interviewed • Viable option for interviewing parents and to collect information on child victimization experienced by children under age 15. • The primary objective is to collect data to better estimate of prevalence rates of CSA and child victimization experienced by children under age 15, and to test a comprehensive profile of individual and family risk factors correlating to CSA and child victimization. 3

Sampling in household survey • Multiple stages stratified probability sampling procedure • 1st stage: two urban administrative districts and one rural/suburban administrative district were randomly sampled from each of the five mainland cities. • 2nd stage: three neighborhoods, towns, or villages were randomly selected from each administrative district. Communes or smaller districts were then further sampled from each selected neighborhood, town, or village. • 3rd stage: the official registers of households and temporary migrants as the sample frame for eligible families. • 4th stage: Within each household, one of the parents or caregivers of any child aged 17 years or under was randomly sampled by selecting the parent/caregiver with the most recent birthday.

School survey • The school-based approach is more appropriate to gather information from children aged 15 -17 to estimate the prevalence of child victimization experienced by them. • less subject to underreporting than that collected in the household setting. • The primary objective of this approach is to collect data to better estimate of prevalence rates of CSA and child victimization experienced by children aged 15-17. The secondary objective is to test risk factors of CSA. 5

Sampling in school survey • used a 2-stage stratified probability sampling procedure to identify eligible students in the 6 cities • 1st stage: randomly chose 150 schools from the urban and rural areas of all the sampled cities with the probability of selection proportional to school enrollment size • 2nd stage: randomly sampled 1-2 classes from each of grades 9-12 in the selected schools. All students aged 15-17 years in the sampled classes were invited to participate.

Sample size and response rates 1. Household-Based Survey (Parents) : 8,945 Parents or guardians of children aged 0 – 17 Response rate: 76.8%

2. Household-based Survey (Children): 3 321 children aged 15-17 years 3. School survey: 18,341 students, aged 15-17 Response rate (school level): 76.5% Response rate (individual level): 95.8%

7

Ethics approval • The study was conducted in compliance with the principles enunciated in the 1996 or later version of the Declaration of Helsinki and the Hospital Authority of Hong Kong Investigator’s Code of Practice in Undertaking Clinical Research. • The IRB of the University of Hong Kong and the Hospital Authority Hong Kong West Cluster provided ethical approval for the whole study. • The local IRB in each mainland city provided approval for the study at that study site.

School Data (Children of 15-17 years)

Publications on this project • Chan, K. L. (Ed.) (2012). Child Protection in Chinese Societies: Challenges and Policies. Hauppauge, NY: Nova Science Publishers. • Chan, K.L. (2013). Victimization and Poly-Victimization Among School-Aged Chinese Adolescents: Prevalence and Associations with Health, Preventive Medicine, 56(3-4): 207-210. • Chan, K.L., Yan, Elsie, Brownridge, D.A. & Ip, P. (2013). Associating Child Sexual Abuse with Child Victimization in China. The Journal of Pediatrics, 162(5): 1028-1034 .

Lifetime Prevalence of CSA (N = 18,341) All (N = 18,341)

Boys (n = 9,773)

Girls (n = 8,568)

Contact sexual abuse

5.2%

6.1%

4.1%

Penetration

4.4%

5.5%

3.2%

Non-penetrative touching

4.2%

5.0%

3.4%

Non-contact sexual abuse

6.8%

8.1%

5.4%

Any type of sexual abuse

8.0%

9.3%

6.6%

Abuse

Major issues

Issue 1: Lower rates of CSA in China • Existing studies: – 2% and 59% (Alikasifoglu et al., 2006 ; Campbell, Greeson, Bybee, & Raja, 2008; Catania et al., 2008; Figueiredo et al., 2004 ; Finkelhor, Moore, Hamby, & Straus, 1997; May-Chahal & Cawson, 2005)

– 2% to 14% among the Chinese populations (e.g. Chen et al., 2004; Hong Kong Medical Coordinators on Child Abuse, 2003; Luo, Parish, & Laumann, 2008; Tang, 2002; Yen et al., 2008)

– 4% to 7% among the other Asian populations (Ashcroft, Daniels, & Hart, 2003; Jirapramukpitak, Prince, & Harpham, 2005; Kim & Kim, 2005).

International Comparison • Stoltenborgh, M., van IJzendoorn, M.H., Euser, E.M., & BakermansKranenburg, M.J. (2011). A global perspective on child abuse: Metaanalysis of prevalence around the world. Child Maltreatment, 16(2), 79101.

• The global prevalence of CSA was estimated to be 11.8% • CSA rates for Asian women 11.3% vs. 18.0% for all countries. • Asian men 4.1% vs. 7.6% for all countries.

• Ji, K., et al. Child sexual abuse in China: A meta-analysis of 27 studies. Child Abuse & Neglect (2013), http://dx.doi.org/10.1016/j.chiabu.2013.03.008

• Comparing Chinese to International: – Total CSA for Chinese females, insignificant – Contact & penetrative CSA for females, lower – Total CSA for Chinese males, more – Contact for males, similar – penetrative CSA for males, lower

Stoltenborgh et al. (2011) explanation 1 • Under-reporting: The fairly low CSA rates for both genders in Asia seem to be consistent with the idea that abuse experiences are less often disclosed in a collectivist culture than in individualistic cultures. • The highest prevalence for girls found in more individualistic countries like Australia and New Zealand might partially stem from culturally based willingness to disclose their sexual experiences and the ease with which they talk about sexuality

Critique: • Over simplistic polarization of countries into collectivist and individualistic cultures • Under-reporting is possible universally but no evidence showing the extent of underreporting by countries.

Stoltenborgh et al. (2011) explanation 2 • Real difference in rates: socioeconomic and cultural differences • E.g. Male dominant society in South Africa, India • rapid social changes in Africa along with increases in urbanization and individualism have led to greater isolation of families.

Critique: • Handy-pick of cultural elements • Explaining one culture but inconsistent with others • Very few studies were using scientific measures of cultural elements e.g. Chan, K.L. (2012). The Role of Chinese Face in the Perpetration of Dating Partner Violence. Journal of Interpersonal Violence, 27(4): 793-811.

Possible explanations • Under-reporting: – Culture-specific face concern – Shame – Victim blaming, emphasis of virginity

• If real difference exists: – Strict parenting style and close parental supervision – Larger proportion living in extended families

Issue 2: Gender difference in CSA victimization • Girls are more likely to be victimized by CSA (e.g. Cawson, Wattam, Brooker, & Kelly, 2000; Fergusson, Horwood, & Lynskey, 1996; Finkelhor, Ormrod, Turner, & Hamby, 2005 ; Finkelhor, Turner, Ormrod, & Hamby, 2009; Jirapramukpitak et al., 2005; May-Chahal & Cawson, 2005; Priebe & Svedin, 2008)

• Boys are more likely to be victimized by CSA in studies conducted in settings high in anonymity such as schools (Madu & Peltzer, 2000; Yen et al., 2008; Choo et al, 2011) and computer-assisted settings (Luo et al., 2008).

• Among Chinese studies, the findings of this study were consistent with a Chinese national study (Luo et al., 2008) as well as a recent study in Shandong (Don et al, 2013)

Possible explanations • Survey methods • Gender role expectations: – Chinese girls leaving toddlerhood are often discouraged for boyish behaviours such as hanging around or playing outside. – more likely to be expected to stay at home, under parents’ supervision (Zhao, 2003) – high value placed on female virginity in Chinese societies may encourage parents to be more protective to their daughters from sexual abuse (Zhou, 1989). – Strict parental supervision might reduce the likelihood for Chinese girls to experience sexual victimization (especially by non-family members)

• Gender Differences in Child Victimization : – Children who experience one kind of victimization are at greater risk of experiencing other forms of victimization. – victimization of any one type left substantial vulnerability even for different types of subsequent revictimization (Finkelhor et al, 2007). – In this study, boys are more likely than girls experiencing child victimization. It may be explaining why more boys than girls reported an experience of CSA victimization.

Table II. Independent Effects of Socio-demographic Variables and Other Types of Child Victimization on Child Sexual Abuse (CSA) (n = 11,150)

• Disclosure: female virginity may act as a barrier for girls to disclose their experience of CSA victimization to others: face and shame

Issue 3: Family violence and CSA

FV & CSA

Issue 4: Risk factors and Impacts of CSA

Risk Factors of Child Sexual Victimization (CSV) Risk factors increase the odds of both lifetime and preceding-year prevalence of CSV Parents’ Characteristics: • Father having secondary 3 or below • Father’s unemployment status • Marital status such as being widowed and divorced or separated • Family with no and low income • Parental negative life styles such as gambling, smoking, alcoholic abuse and substance abuse Children’s Characteristics: • Being boys • Age • Children having siblings

Violence Characteristics within the family • In-law conflict • Physical parental IPV • Psychological parental IPV • Different forms of elder abuse such as elderly physical abuse, elderly verbal abuse and elderly neglect

Impact of CSV Both lifetime and preceding-year CSV was associated with • Poorer physical health • Poorer mental health • Specially, mental health problems such as depression and PTSD

Health Correlates and Child Sexual Abuse (CSA)

Conclusion • Prevalence rates of child victimization: 35% 72% children had been victimized in their lifetime • Poly-victims: 7% - 29% • Children who experience one kind of victimization are at greater risk of experiencing other forms of victimization and subsequent re-victimization (Finkelhor et al, 2007).

Major issues • CSA associated with child victimization and family violence • Psychological impacts • Insufficient evidence to conclude if rates of CSA in China are lower than Western countries or rates for boys are lower than girls • Need more scientific studies

High prevalence rates of child victimization!