The main finding from this study was that the mental health of second-generation Vietnamese in Norway, assessed by the children themselves, is better than that of their Norwegian compatriots. Norwegian-Vietnamese children and their parents reported greater levels of low-risk or normal functioning, although the parents reported that their children had more total problems and problems with peers than did parents in a Norwegian comparison study.
Contradictory results from studies of the mental health of Vietnamese children in exile suggest that our study belongs to a research field with many controversies.
Studies of immigrant mental health have been criticized for their lack of information on the mental health of the inhabitants of the country of origin [1, 16]. Analyses of the Achenbach Child Behavior Checklist (CBCL) data in a population-based survey of mental health problems in Vietnamese children in Hanoi  showed that the Vietnamese children had lower scores than the US norms for this test, with only half as many scoring in the clinical range. Their result is consistent with our findings.
The discovery of better mental health in our study may have three different interpretations.
First, the results may indicate a true difference between Norwegian-Vietnamese and Norwegian children, as the lower prevalence of mental problems in Norwegian-Vietnamese children concurs with the results of other studies of South-East Asian immigrant children who have been assessed by the CBCL [17, 18] or by the Rutter Parent Questionnaire , a predecessor of the SDQ. The CBCL and the SDQ are both designed to obtain ratings of children's problems and can be used to identify high-risk children .
The distributions of SDQ scores are found to be similar across the Nordic countries , including Norway.
Beiser et.al.  report better mental health in children of immigrants; this is partly attributed to Canada's selection - policy, "helping to ensure selection of healthy, resilient ...families and children". A "healthy immigrant effect" has been described, e.g. in studies from Canada [21, 22], which has a large contingent of immigrants and an immigration selection-policy. After arriving as apparently healthy immigrants , the health of immigrants subsequently declines and converges towards the native- born population. Contrary to this, the unselected Vietnamese parents of the study sample arrived in Norway with higher levels of psychological distress than in the host-populations , 1/4 scoring as "cases". Norway had no pre-existing South East Asian cultural community and none of the refugees had any knowledge of the Norwegian language prior to their escape from Vietnam. As a group, they were relatively unprepared for migration, and the changes represented large-scale acculturative stress. Consequently, the finding of good mental health in the refugees' children in our study cannot be explained by the "healthy immigrant effect".
Some aspects of the Vietnamese children's family life may account for a lower prevalence of mental illness. Possible protective factors include a family structure firmly rooted in a tradition and value system [17, 24], as well as parental supervision . Cross-cultural differences in socialization practices and expectations for children's behaviour [5, 26] may cause Vietnamese parents to discourage externalizing behaviours more forcefully in their children. Even so, the children in our study had levels of self-rated emotional problems comparable to their Norwegian counterparts. Thus, our findings may indicate an immigrant advantage in terms of emotional and well-being outcomes.
Other factors that should be considered include genetic factors, temperamental differences  and the parents' health . The relationships between the parents' and the children's health will be reported in a forthcoming paper.
Second, the reports of good mental health may be biased. As a consequence of the high expectations concerning their behaviour and performances, and the upbringing in a culture in which mental illness is highly stigmatized , immigrant adolescents may feel less comfortable reporting behaviours that might be perceived as deviant. Such social desirability may be seen as a bias, as well as an adaptation to Vietnamese cultural and parental values.
Surprisingly, we found that the parents reported as much disruptive behaviour as the Norwegian community sample, and some scores were even higher, especially the number of peer problems (Table 2). Being less acculturated than their children, immigrant parents may be mostly at a loss when evaluating peer relationships in the Norwegian cultural context. Parents worry that their children are not working hard enough to achieve academic success . This may explain the parents' reports of high levels of problems in their children as possible instances of over-reporting.
Third, the Norwegian-Vietnamese children, but especially their parents, may understand the statements in the SDQ differently from Norwegians, parallel to the conclusion in a Chinese study . Assumptions about development, normality and psychopathology are culturally embedded [31, 32], and there are cultural differences in definitions of psychopathology . In his studies, McKelvey [3, 34] mentions that, despite the CBCL's impressive performance in several cross-cultural settings , symptoms that are possibly related to child mental illness may have a different meaning within the Vietnamese cultural context. The higher parent-rated problem scores in our study may reflect the parents' critical or anxious monitoring of their children's school performances, more so than reflecting any symptoms of psychopathology.
Stevens et al.  discussed the validity of cross-cultural assessment. Although several studies indicated that their instrument showed sufficient validity for their populations, as comparable factor structures and high reliabilities for both the migrant and the native populations were revealed , the instruments used may be less valid for assessing migrant samples. This explanation of the differences in problem behaviour between migrant and native youth has been supported by others utilizing the SDQ [37, 38].
Strengths and limitations
This research formed part of a prospective longitudinal follow-up study. The personal follow-up design of the study was strengthened by a culturally relevant approach enacted by the Vietnamese co-researcher. As he was responsible for making contact with the families, his efforts contributed to the high inclusion rate of children (91%), which is considered a major strength of the study. The longitudinal prospective design, with information on the parents' mental health, is another strength.
Additional strength is the use of two informants. The discussion on what type of informant carries the highest weight is ongoing . Montgomery  wondered whether the Youth Self Report (YSR) and the CBCL might be considered as measuring two qualitatively different constructs, with the difference between informants not just resulting from cross-informant disagreement. This difference is found to a higher degree in refugee- and immigrant populations , as in our study (to be reported elsewhere). A similar question may be posed for the reports from the SDQ, as from the CBCL/YSR. As a group, children of Vietnamese refugees are higher acculturated than their parents . Consequently, comparison of self-reports may be considered as more culturally relevant than a comparison of parents' reports for the two samples.
One important limitation of the study is its small sample size that requires a cautious interpretation of the findings. It made it difficult to adjust for the number of children in some families, as siblings' reports cannot be considered as independent. However, the small sample is from an unselected group of refugee parents. Countries with a large immigrant population, as Canada, have immigrant selection policies probably resulting in a different composition of immigrants, also in terms of mental health. The results from a non-selected group of refugee-families, although small, can therefore also be considered as strength of the study.
A major advantage as well as a challenge of the study was the comparison of the Norwegian-Vietnamese and the Norwegian community samples. Some basic information available in both samples on the families, including the parents' income and perceived economy, made possible a sensitivity analysis of the comparison of two samples, using somewhat broader information than just gender and grade. On one hand, the Norwegian - Vietnamese children were to a higher degree than their peers living together with both parents, a fact expected to explain a better mental health in the children . On the other hand, the lower level of education as well as economy in the Vietnamese families would expectedly result in worse mental health in the children . Still, basing our comparison on all these variables, the pattern of better mental health in the Norwegian-Vietnamese sample persisted.
A limitation of the study is the lack of comparison groups for the whole age range included in the study, that is, for both the self - reports and the parent reports.
A possible a limitation of the study is that the question whether the differences in mental health in the two samples can be explained by the cultural differences is still unanswered. The three different aspects described in the discussion-section are all, to some extent, related to the issue of "culture", and the role of migration and culture are difficult to disentangle from each other.
The lack of cultural validation of the assessment tools is a general problem that is not limited to this study and represents a major challenge in trans-cultural research.
The refugees studied at T3 were considered to be a representative sample of the third wave of boat refugees who arrived in Norway in 1982. The major characteristics of the parents included in the study were the same as those in the group who did not have children born in Norway. Consequently, and in spite of the reported limitations of the study, the children may be considered a representative sample of second-generation Vietnamese in Norway, who belonged to this group of Vietnamese refugees.