Suraporn Ounsuan. 1996.  Synthesis of Research Results on Moral Reasoning: A Meta-Analysis. 
Dissertation, M.S. (Behavioral Science )Behavioral Science Research Institute. 


            This study synthesized, the research results concerning the relationships between moral reasoning
(Kohlberg' s moral development approach) and some variables, using a meta-analysis. The objectives of this study
were to find the average of the weighted effect sizes of the three groups of independent variables : cognitive,
affective and child-rearing. 
                The data used in this study were from 26 research reports during 1963-1994  from Srinakrinwirot
University, Kasetsart University, Chulalongkorn University and National Youth Bureau. The cognitive group
included intelligence, academic achievement, thinking ability and language ability. The affective group included
ego-identity, mental health and future orientation trait. And the child-rearing included love oriened type, control type,
reasoning oriented type, physical punishment type and strict oriented type. The indices of the relationship used in
these studies were from the following statistics : r, t, F, Z or  c2 
                           The relationship indices were transformed to effect sizes by Rosenthal's formula, and the
homogeneity of the effect sizes were tested by using Hedges' formula. Finally, The averages of the weighted effect
sizes were calculated by using Rosenthal' s formula. 
                The results of the study were as follows : 
                1.  The cognitive variables. 
                    1)  The effect sizes of intelligence, academic achievement and thinking ability were heterogeneous. The
variations of the effect sizes could be explained by the subjects' age. The highest weighted effect  sizes averages
were found for the 2-9 year old subjects, and the lowest averages were from the group of subjects who were 14 years
and older. 
                    2)  The effect sizes of intelligence and language ability were homogeneous and their averages were
small. 
                    3)  The effect sizes of academic achievement were heterogeneous, but their averages could not be
detemined. 
                    4)  The effect sizes of thinking ability were heterogeneous. The variations of the effect sizes could be
explained by the subjects' age. The 2-9  and the 10-13  year old subjects had a positive effect sizes, with the higher
value for the 2-9 year olds. The 14  year and older subjects had a negative effect size. 
                2.  The affection variables. 
                    1)  The effect sizes of ego-identity and mental health were heterogeneous, but their averages could not
be determined. 
                    2)  The effect sizes of ego-identity were homogeneous and their averages were moderately. 
                    3)  The effect sizes of mental health and future orientation trait were heterogeneous, but their averages
could not be detemined. 
                3.  The child-rearing. 
                    1)  The effect sizes of child-rearing were heterogeneous. Their variations could be explained by the type
of child-rearing. The reasoning orientation type and the lover orientation type had a positive effect  sizes. And the
physical punishment type, the control type and the strict orientation type had a negative effect sizes. 
                    2)  The effect sizes of the positive child-rearing were heterogencous. The variations of the effect sizes
could be explained by the subjects' reliability of moral reasoning tests. The heighten reliability subjects had
moderately weighted effect sizes averages. 
                    3)  The effect sizes of the negative child-rearing were heterogeneous, but their averages could not be
determined. 
                    4)  The effect sizes of the lover orientation type, the control type, the physical punishment type were
heterogeneous, but their averages could not be determined. 
                    5)  The effect sizes of the reasoning orientation type was heterogeneous. The variations of the effect
sizes could be explained by the subjects' age. The 10-13  year old subjects had moderately weighted effect sizes
averages, and the 14  year and older subjects had small, but the 2-9  year old subjects found not effect sizes. 

                                         | SWU | | BSRI |