xOfficial report about CWS services to families: Eight out of ten are satisfied?
Children and parents are mostly satisfied with the child protection services. This is the result of a survey recently published by the Norwegian Bufdir (The Norwegian Directorate for Children, Youth and Familiy Affairs), according to the Norwegian Government:https://www.regjeringen.no/no/aktuelt/a ... id2454049/
The report is available athttp://www.bufdir.no/bibliotek/Dokument ... UF00003223
Family Minister Solveig Horne has no scientific background, and has to rely on her co-workers when interpreting the results of the survey. Nevertheless, there are reasons to raise questions concerning statements that most families involved are happy with the Norwegian CWS.DISCUSSION
Let us take a look at the possible sources of error in the report.Enrolment
. Children placed outside their homes are not included in the survey. If all measures prescribed by the CWS had been investigated collectively, the share of satisfied users would most likely be significantly reduced.Selection
. The researchers have interviewed children, parents and the caseworkers/supervisors in ongoing cases. The cases are selected, and not random as they should be to produce reliable results. The CWS, who themselves are participating and being evaluated in the survey, have been engaged to organize important parts of it. In fact, the CWS itself has chosen and approached the families participating.Relationship research/casework
. The survey was carried out in close cooperation with the CWS. The researchers acknowledge that the participants because of this may have had problems distinguishing between casework and research, something which considerably reduces the reliability of the study.Participation
. The answers percentage in this kind of study is always essential to how to interpret the results. In this survey, we do not know how many families were asked to participate, and we therefore do not know the share of families included in the survey. Nor do we know how many families in the group of interest were actually asked to contribute. All we know is that the researchers wanted the CWS to invite all families with ongoing cases. The researchers admit that the families picked can hardly be regarded as representing the population of families receiving measures from the CWS.Data collection and anonymity
. In addition, the means of data collection is not fully described in the survey. Most participants have answered the question forms in writing. From the information given, one can not conclude as to whether the answers given have been available to the CWS. With the very close co-work between researchers and the CWS, one cannot exclude the possibility of participants anticipating that the CWS will read the answers.
For minority/refugee families, cultural aspects can lead to an expectation that the CWS itself will have access to the results, and that the responders will not be anonymous. Children will probably have an even greater expectation that the caseworkers will read their answers. If the participants do not know if their discretion is being respected, this will influence both the answers given and the number of responders. Negative families will not respond, while the number of positive answers might increase, because the participants may expect their answers to influence on their own ongoing cases. The researchers acknowledge this problem.SUMMARY
The sources of error are many. Children taken out of their homes are not included, which increases the percentage of positive answers. Answers percentage is unknown, because the number of families invited for interviews is unknown. One can not exclude the possibility that families not coping well with the CWS do not want to participate. We do not know, however, for how many this might be the case. Selection of families has not been random, and the participants have been chosen by the CWS itself. This increases the risk of a bias towards positive families. Uncertainty concerning anonymity can increase the number of positive answers, and likewise reduce the number of negative participants.CONCLUSION
Regarding the design of the study, one might conclude that it to some extent is an investigation of the CWS carried out by the CWS itself.
Within the design of the study and its methods lie considerable sources of error. In my view, it is impossible to make an estimation of to what extent the families responding in the survey mirror the population of families receiving measures from the CWS. Sympathetically, the researchers have themselves listed possible sources of error in the report.
Noteworthy is also the researchers' statement that the effects of CWS measures are largely unknown, and likewise that possible negative effects of CWS are hard to estimate. They also conclude that the overall cost/effect balance of CWS is unclear.
Even though the researchers in my opinion understimate the influence of the sources of error in the survey, they seem to have a more careful interpretation of their own figures than the Bufdir and the Government.
The conclusion that eight out of ten families are happy with CWS measures, is hard to extract from the report considering the above mentioned sources of error.
(For references, see )