Bellamy et al aim to evaluate the effectiveness of outpatient mental health
services on the externalizing and internalizing behavior problems of children
in long-term foster care. The sample of children were selected using a
two-stage combined stratification and cluster design, which was used to control
for potential confounders, which in turn minimizes the risk of internal
validity. For outcomes, they engaged in externalizing and internalizing
behavior problems using the CBCL checklist. The CBCL has been used frequently
throughout research in foster care and mental health studies, making it an
adequate tool to use in this study. They also used multiple imputation, MI, to
address missing data, which again seemed reasonable because they stated that MI
generally outperforms other approaches because it often does not lead to bias
and false identification significant differences.
The findings of this study suggest that
outpatient mental health services provided to children who have experienced
long-term foster care in the U.S. do not result in any improvement in
children’s behavioral health. These findings can be generalizable across the
larger child and adolescent mental health system, which means this study
contains external validity. The researchers’ biggest limitation was that there
was a balance problem on the type of maltreatment. One limitation that they did
not address but that I believe is important was that they do not define
maltreatment. There are many different views of maltreatment. Therefore,
defining it could have been more helpful to readers.
Burns et al aim to identify factors
related to the need for and the use of mental health services among youths with
the child welfare system. The data presented in this study is from the National
Survey of Child and Adolescent Well-Being, which is a nationally representative
sample of youth who were subjects of maltreatment investigated by child welfare
agencies. This makes the data and survey significant to the study’s aim because
it is a nationally represented sample. The study’s design consisted of a
stratified two-stage sample which is appropriate for this study because it is a
large sample size which makes it more statistically significant because it more
reliably reflects the population mean. Continuous t-tests were used to control
for multiple comparisons, which I believe strengthened their internal validity
because it is important to know if there was significant difference between the
means of the two groups. Their main findings were that one half or more of
youths who come into child welfare have manifest emotional and/or behavioral
problems. These findings are statistically significant because they are
consistent with existing literature on mental health need for maltreated youth.
The authors identify five
limitations. A couple of them are measurement issues and reliability and
validity. To add on to their already mentioned limitations, when parents were reporting
their children’s behavior, they could have been using social desirability. A
different way researchers could have approached this was maybe observing the
children themselves so they avoid social desirability.
Minnis et al aim to assess the
prevalence of mental health problems in children in foster care. The children
were included in either medium to long-term foster care or permanent foster
care, which is good because it makes the results more generalizable to children
in long-term or permanent foster care, emphasizing its external validity. The
study has a strong randomized control trial. For measures, they used the Strengths and
Difficulties Questionnaire which focused on both the positives and the
negatives. Some can argue that it is a strong tool to use for this specific
study because it is well-validated. To control for internal consistency, they
used the Reactive Attachment Disorder Questionnaire. To control for potential
confounders, they entered the confounders into the regression model one by one.
Data were analyzed for female foster care mothers only, which may lead to
potential bias because male caretakers were not included except if they were
single parents. The main findings supported the authors’ hypothesis that almost
all of the children had been abused or neglected and over half of them had
mental health problems.
The study’s biggest limitation was
that the sample may be biased. While I do agree with the limitations, I do
believe that there are more that were not mentioned. The biggest one was the
fact that the SDQ could have been filled out by the parents, which can call for
potential bias and social desirability.
Tarren-Sweeney aims to look at rates
of meaningful change in the mental health of children in long-term foster care.
Researchers used the Children in Care Study epidemiological study of mental
health children in long-term foster care. This was a good tool to use because
it consisted of a baseline cross-sectional survey which included a follow up
survey which is always good to have because it checks for consistency. They
used the CBCL to measure child problem behavior. I agree with this checklist in
this study because it considers valid and reliable mental health aspects. To
control for internalizing and externalizing factors, they measured social,
attention and thought problems. Results confirmed the research question:
substantial proportions of children in care experience very different mental
health trajectories. The authors mention two limitations: high sample attrition
and collection of data at only two points in time. A limitation I found was the
survey procedure section. It seemed unclear to me how and who conducted the surveys.
Also, how they went about follow-up surveys remained unclear to me as well.