Parenting and Adolescent's Psychological Adjustment: Toward a Systemic Approach in Parenting Research

Abstract

This article's objectives are twofold: (a) to disclose the possible distortion of the associations found in the reductionist research that prevails in many areas, in order to ensure greater caution and better understanding of such research. (b) To study the associations between family and parental factors and adolescent psychological disorders (PD) according to a systemic model that analyses eight familial factors and eleven parental factors in addition to two nominal ones: culture and the adolescents' sex. The study is based on a data collected from nine countries (1358 male and 1526 female adolescents), regarding two categories of family factors (socio-economic and connectedness) and three categories of parenting factors (control, inconsistency, and rejection) and adolescent psychological disorders (PD). To compare different levels of reductionism, four analyses of the same data were carried out, ranging from an analysis of the associations between each factor and adolescent PD (reductionist), to analysis of the associations between all the factors taken together (systemic) and adolescent PD. In addition, the systemic analysis was carried out among different groups of adolescents according to two nominal variables: culture (western and eastern) and the adolescents' sex (male female). Our results show that in a reductionist analysis most of the family and parental factors have significant associations with adolescents PD, and altogether explain 37.2% of adolescents' PD. Most of these associations were diminished or changed in the systemic analysis and explained only 13.5% of the PD variance. The associations of the more systemic analysis changed again when two nominal factors (culture and sex) were taken into consideration. These findings indicate that reductionist analyses may lead to illusionary associations and that mixed results are an inevitable or even inherent byproduct of reductionist research. Phenomena take place in nature and society through interactions between variables that constitute a dynamic system. In order to study a phenomenon, researchers define variables relevant to the phenomenon and develop instruments to measure these variables. Most of the studies in social science and medicine tend to be reductionist; they deal with only a few relevant variables at a time and exclude many other relevant variables. Therefore these fields of research are flooded with mixed or inconsistent results. For example, studies examining the influence of personality factors on health [1], of thinking on pain [2], debriefing on persons exposed to traumas [3, 4], viewing of violent films on violent behavior [5-7], and emotional expression on blood pressure [8] arrived at mixed and inconsistent outcomes. SYSTEMIC EXPLANATION FOR THE INCONSISTENT RESULTS Inconsistent results are usually attributed to differences in methodology, such as differences in the samples, tools, and data analysis. This explanation ignores a more crucial attribution-the reductionist approach in research. Inconsistent results are not necessarily caused by methodological problems; they are inherent in the reductionist approach itself: Once researchers select certain variables to study at a time and exclude other relevant variables, based on their theory or for practical reasons, they in fact create an artificial phenomenon. Such phenomenon is substantially different from some other artificial phenomenon, created by other researchers through their selection of a different complex of variables. We may expect both phenomena to be different from the real phenomenon in nature or society. In my article on causality and mixed results [9], I demonstrated that the association between A and B depends on the presence or absence of other relevant C variables. Thus, when reductionist researchers find a significant association between A and B, systemic research that integrates more relevant variables may change this association, because both A and B are never independent factors and have other associations with additional relevant variables (Cs) in the system. Therefore, when the association between A and B is studied in the presence of Cs, then the variance of B, for instance, may be explained better by these additional factors (Cs), rather than by A, which had originally been perceived as having a significant association with (or explained by) these C factors. The difference between reductionist and systemic research can be expected to be dramatic, when some of the C factors have a negative correlation with A or B. In that case the positive correlation found between A and B may turn into a negative one, when Cs are added. Therefore, regression analyses on different groups of variables may convey to a different outcomes. To clarify the effect of adding one variable to another, here is a simple illustration from chemistry: Simple chemistry teaches us that combining oxygen and hydrogen