Sibling Coercion & Interaction Quality Among Youth in Foster Care

Background and purpose:

Sibling coercion is the most common form of family violence (Button & Gealt, 2009) and is related to child externalizing and internalizing symptoms (Kim et al., 2007; Stocker et al., 2002). Siblings in foster care are at risk of being highly coercive given their exposure to abusive homes (Bank & Burraston, 2001) yet studies of this population are rare (Linares, 2006). This is the first study to describe coercion levels among siblings in foster care using rich observational data in relation to child reported sibling interaction quality.

Methods:

Data derive from the Supporting Siblings in Foster Care Study, an experimental test of a sibling relational enhancement intervention (Kothari et al., 2017). At baseline, children were 7-15 years old; 60% youth of color; 72% placed with at least one sibling.

Observational videos of structured sibling interactions were coded to measure coercion.  A count of coercive behaviors was divided by the total minutes of observed time to produce child coercion rates per minute.

Children completed a 13-item sibling interaction quality (SIQ) questionnaire measuring how easy or difficult it is to do certain activities with their sibling (e.g. Solve problems that keep us from getting along…Tell my sibling what I like about him or her).  Items were rated on a 4-point scale (1=Very Hard, 4=Very Easy). A weighted sum score was created if at least 10 of 13 items were completed.

Department of Human Services administrative records provided demographics and sibling placement.

Descriptive statistics and bivariate tests of coercion rates, demographics, and sibling placement were conducted.  Negative binomial regression analysis adjusted for observed time and standard errors clustered by family examined whether SIQ was related to coercion rates, controlling for age, gender, and sibling placement.

Results:

Coercion rates per minute varied widely (0 – 2.32). The average was .38 behaviors per minute; above the average .27 rate observed in the general population but below the average .75 rate for a known clinical sample (Arnold et al., 1975). Older children displayed lower coercion rates (r=-.13, p<.05); siblings placed together had a higher average coercion rate (z=2.48, p<.01).  

The overall regression model was significant (x²(4)=11.23, p<.01). SIQ was a significant predictor; The incidence rate ratio showed for a one-point increase in SIQ the expected coercion rate decreased by a factor of .98 while holding all other variables constant (IRR=.98, 95% CI:[.96, .99], p<.01).  Or, the expected coercion rate was reduced by 2.4% for a one-point increase in SIQ. The expected coercion rate was reduced by 16.0% for one standard deviation increase in SIQ.

Conclusions and implications:

Though abusive home environments may increase sibling coercion, our findings show not all children in foster care meet this expectation.  Importantly, perceived sibling interaction quality may protect against coercive behavior. If children learn to engage with siblings in a pro-social manner, it could also positively impact other relationships (e.g. peers and foster parents) and in turn improve well-being outcomes for youth in foster care.  Next steps are to complete longitudinal analyses.