Family Background and the Process of Ability Formation


  • Original Article: Katei kankyo to noryoku keisei no katei. 2008. Shakaigaku hyoron 59: 514–531.

  • Translation by Philip Flavin.


This paper takes a sociometric approach to the process of skills formation in children as they mature into adults. Honda suggests that meaningful communication with family as a child is the determining factor that raises those abilities (ability β1) that fall outside of scholastic aptitude – motivation, personal relations skills, personality, and emotions. Along with considering the results of Honda's thesis, I have constructed an operating hypothetic model that includes two additional factors: the financial state of the family during childhood, and the transition of the communication target from within the family to a target outside the family. The data for this investigation were then verified. From the results, the direct effect of the observed family communication, when isolated from other influencing variables in the formation of abilities β, is not as decisive a factor as Honda emphasized. It was clear that the financial status of the family played a role, as did the indirect results of the intermediary shift from communication within the family to outside the family (the process of children becoming socially independent). The results of this analysis show that we should not necessarily be looking for the decisive factors of abilities β within the family; if anything, it reflects the existence of a formative route for abilities β that capitalizes on the resources outside the family. Based on the above, the results also show the potential for the introduction of social intervention to address the disparity and inequality in skills formation, and the possible anticipation of more rationality in the overblown discourse surrounding the role of the family in facilitating educational ability.


There is a large amount of international research on the formation of human skills in fields such as sociology, economics, psychology and more (Jencks et al., 1973; Herrnstein and Murray, 1994; McCutcheon, 2006). Recently, it has become clear from the construction of new economic theoretic models that reflect reviews of this research and the data from numerous tests, that a family's initial investment has a deep impact on the formation of children's skills and the later disparity and inequality in their occupation life as adults (Cunha and Heckman, 2007). Till this point, in economics, there was great interest in the formation of human capital primarily through workplace training and in high school and university education, and it is significant that the formation of skills as a child would later have significant influence after reaching adulthood (Toda, 2007). Compared with the amount of academic research on skills formation produced overseas, research by Japanese scholars remains relatively small, and thus results in Japan have yet to be fully validated.

Honda's (2005) study is one of the few that alludes to various aspects of skills formation during childhood and adult work life.2 Turning to several research projects, Honda clarified the importance of “family communication” during childhood and the formation of skills that fall outside of scholastic ability – motivation, abilities with personal relations, personality and emotion. Of great interest is the connection of Honda's research with Cunha and Heckman's (2007), which insisted that the decisive factors for those skills seen as necessary in actual society depended upon the childhood family environment.

Nevertheless, much of Honda's analysis never moves beyond a multiple regression analysis conducted on pre-existing cross-sectional data. Therefore, neither the statistical correlations that are conjectured to exist between explanatory variables in the model's interior, nor the theoretic and logic structural correlations have been scrutinized. Thus, it is difficult to claim that the mechanism used to examine the multiple layers of skills formation during the lengthy period of childhood to adulthood has been fully delineated. This paper, therefore, will first construct an operating hypothetical model for the process of adult skills development based on the family environment during childhood. The verification of the research data and the results of the analysis will present a deeper understanding of this research topic.

The Data and the Structure of the Paper

For the purpose of constructing an operating hypothetical model that shows the skills formation process from childhood to adulthood, and for the verification of data, it is necessary to have long-term data that includes information for each subject on: family environment, childhood experiences, academic history (including marks), and an evaluation of present workplace skills. The research object is therefore adults who have already graduated high school or university, and I investigate their present work and life conditions, as well as their family environment as children along with their childhood experiences. Because of the nature of the research, it was necessary to rely upon the subjects' recollections.3 Despite the fact that this is a paper on sociological investigation that encompasses childhood conditions and the adult work conditions for the purpose of analysis, there is, unfortunately, a dearth of earlier data sampling. The data were collected in a research project, Wakamonono shigoto seikatsu jittai chōsa (The Present Working and Life Conditions for Young Adults), conducted in January 2006 by the Benesse Educational Research and Development Centre. The target subjects were young adults from the ages of 25 through 35. The Internet was used to acquire the data.4 Sample size was 2500, with a ratio of 1:1, men to women. The analysis in the fifth section used 1078 data samples that corresponded to a list of employed men.

The structure of this paper is as follows. The next section is a review of Honda's (2005) thesis, which serves as the foundation for the argument of this essay. I will construct an operating hypothetical model based on domestic and international research on skills formation. Then, I will verify the model by using data from the earlier-mentioned research project. The last section presents the new insights gained from the results of the analysis, and suggests future topics for research.

Review of Earlier Work

There are two central points I wish to review from Honda's (2005) work as it forms the base of this paper. The first point is the analyses results, and the second point is what is suggested by the results. As mentioned above, from the results of several research projects, Honda argues that the most important aspect of those skills for children and adults5 believed to be formed from other elements that fall outside of those that are quantifiable through grades or paper tests (abilities α) – such as motivation, personal relationships, personality and emotion (abilities β) – is the home environment. The first step was to use data collected from primary, junior high and senior high schools, and clarify through techniques, such as multiple regression analysis and other statistic analyses, whether or not there is a correlation between what is thought to conform to the concept of “abilities β”: Honda's analyses being based on the responses to the different questions and home environment. In concrete terms, these are the answers6 to questions that were thought to indicate “abilities β”, which were the following for primary and junior high school students: I am capable of entertaining those around me; I can express my opinion even in front of others. The questions for senior high school students were: I can convey my opinion clearly to others; I have the ability to lead others. Those topics, thought to indicate the status of the family environment in which the students were raised, were divided into two target groups. The first target group of primary and junior high school students were presented with seven topics believed to represent family communication, which included grades/marks or classes, and future employment. The frequency with which a total of seven topics were discussed with parents was transformed into a score, which was then added. The eight topics for senior high school students included what they did at school and personal concerns or problems. The frequency with which these were discussed with a parent was also transformed into a score and added, this providing a statistical confirmation for the meaningful existence of a connection between the two fields. Representing this on a figure would doubtless take the form given in Figure 1. The arrow represents “cause and effect”, while the + represents “the richer the communication is in the family, the higher the ‘abilities β’.”

Figure 1.

The correlation between family communication and abilities β

Based on the results of this analysis, Honda develops the following argument:

All that can be said is that, when compared to “modern skills”(abilities α), “postmodern skills”(abilities β) are all the more the congenital nature of the individual, or they are largely determined by the requisite continued ordinary environment during the process of development. The reason for this is that “postmodern abilities”(abilities β) are physically integrated into the personality, emotions and physical body. This, I believe, is due to the important component of family environment in the formation ofpostmodern skills(abilities β). In this instance, however, it is impossible to attribute family environment to family economic abundance or to the social status of the parents. Rather, qualitative informal interaction with the family on a daily basis is the aspect that occupies an important role in the formation ofpostmodern skills. (Honda, 2005: 23–24. Emphasis and parenthetical inserts by this author.)

Here, Honda clearly suggests that “family environment” is of extreme importance, and that quality interaction between members of the family – family communication – is more important than the factor of the family's financial background. To rephrase this, Honda is positing that rather than financial capital, what counts most in the formation of skills is cultural capital. This point will be pursued in the next section.

Following this, Honda continues with an analysis of “life skills” (“communication skills” and “positive intention”), which is believed to correspond to “abilities β”, the data for this analysis being taken from research on working adults who have graduated from school. From this analysis, Honda then posits that life skills, which corresponds to “abilities β”, is what decides social, economic and psychological positioning upon becoming an adult. Honda has named this social condition a “hyper-meritocracy”, and provides the following scenario.

The problem of social inequality reflects the relations between members of a family, and is the most difficult region for political intervention, the problem being that the significance in deciding an individual's future is greatly increased. The quality of family environment controls “postmodern skills” (abilities β), and should these “postmodern skills” (abilities β) be important in the life of the individual, then it can be assumed that the family into which one is born will play a significant role in determining the qualitative advantages and the disadvantages in the survival of the individual. (Honda, 2005: 32. Emphasis and parenthetical inserts by this author.)

This suggestion – that quality interaction between family members as communication is a decisive factor in forming “abilities β” as cause and effect – is the fundamental basis for Figure 1. Honda suggests atop this that there is a theoretic base for the difficulty of political intervention in family communication. The suggestion is then that the family into which one is born, something that is impossible to control, is then connected to what disparity there may be in the formation of skills, which in turn gives rise to social inequality throughout life (Suggestion 1).

Honda (2005) further develops her argument by suggesting that under “hyper-meritocracy,” there is an increasing pressure to recognize the figure of the parents during the period of child-rearing, particularly the mother, as she trains and educates the child, which is a concept that relies upon Hirota's (1999) concept of the “perfect mother.”

How to have them do homework begins with teaching them how to wash slippers, how to converse, how things are positioned in the living room – the consideration of any number of things must be instilled. It is necessary to construct a delicate parent-child relationship in which the child is not spoiled, but neither can the child be given excess freedom. But with this comes the image of the “perfect mother” who takes the lead in the formation of the “child's personality”. This has been suggested as the ideal model for the mother, and the pressure to achieve this ideal has started to include all women with children in present-day Japan. (Honda, 2005: 210–211. Emphasis by this author.)

The suggestion is that the mother is then forced into a condition under this hyper-meritocracy in which there is no limit to the investment she must make in the formation of the child's skills. This is a causality model fundamentally based on Figure 1, which shows how family communication controls “abilities β.” The parent becomes the sole controlling agent in family communication, and the suggestion is that the mother in particular is subjected to excessive pressure (Suggestion 2).

A reduction of Honda's (2005) argument is as follows. First: the results of the analysis can be reduced to the cause and effect model shown in Figure 1 that shows the influence of family communication on the formation of “abilities β.” Second: the suggestions deduced from this are: (i) the problem of inequality and disparity in the development of skill is due to the family environment into which one is born (Suggestion 1); and (ii) the pressure of the perfect mother belief seen during the period of child-rearing is influential (Suggestion 2). The following two points – the decisive factor in the formation of “abilities β” is family communication, and the impossibility of exterior social intervention into family communication – are the basis for the thinking behind the two suggestions. The two suggestions that have been deduced from Figure 1 are logically correct. At the same time, however, there is the question of whether or not this model properly represents the actual process of skill formation, and, as will be seen in the next section, there is room for further investigation.

The Construction of an Operating Hypothetical Model

Two Primary Factors in the Family Environment: Cultural Capital and Economic Capital

The primary factors in family environment that influence skills formation can be divided into the two large divisions of cultural capital and economic capital. Honda attaches great importance to family communication, which corresponds to the former division of cultural capital. Needless to say, this argument is based on Bourdieu's and Bernstein's theory of cultural reproduction. Japanese research in this area has been carried out by researchers such as Kariya and Shimizu (2004). In this research, a number of questions thought to reflect the cultural level of the family were presented to primary and junior high school students as part of a paper survey, an exemplary question being whether or not “your family had ever taken you to a museum or art gallery.” From the answers given, they constructed a uniform scale that reflected the cultural level of the family, which was then used to analyze the correlation between family cultural level with scholastic ability and learning behavior. The results clearly showed that the higher the cultural level of the family, the higher the students scored on paper tests. It also showed these students were more positive towards research-based learning and group learning. While it is difficult to claim that there is a large amount of academic research in Japan, such as Honda's, that focuses on “family communication” as a factor of cultural capital, there is nevertheless a large amount of international research, both theoretical and empirical, on “family communication” (Adachi, 2006). Basing his work on previous “family communication” studies, Le Poire (2006), for example, concludes that with the positive interaction between parents and children, there is a strong and unmistakable correspondence with the children's performance at school, and furthermore, that the more children were aware of the parent's expectations, the greater the children's aspirations towards education and career. Honda's (2005) research and its focus on family communication clearly fall into this area of scholarship.

In contrast with the above-mentioned sociological research that focuses on skills formation and family cultural levels, there is, however, economic research that indicates a strong connection between the family's economic capital and the formation of skills in children. There is a large amount of international research, from Becker and Tomes (1979) to Cunha and Heckman (2007). In looking at the recent direction of research in Japan, Mimizuka (2007), for example, has clarified that the indicators of family financial ability, family income and educational expenses outside of school have a strong influence; this conclusion comes from the results of an analysis of the decisive factors for the scores from an investigation into the scholastic abilities of year-six primary school students. Just as with Kariya and Shimizu's (2004) work, there is the problem of analyses that make use of the results of investigation in which children provide answers on the cultural environment and education of the parents. In this research, the earlier-mentioned problem was avoided by the inclusion of a solution in which the investigation into children's scholastic abilities was matched with the results of answers to questions directly addressed to the guardians. With this, it was clear that, rather than the cultural environment and educational level of the parents, financial ability had greater influence. Based on the discussion above, this paper aims to introduce the indicators of a family's financial ability to the family communication model believed to represent the family's cultural environment as this is the most appropriate method for pursuing an understanding of the decisive factors in skills formation.

Should the premise behind Figure 1 be adhered to, the new model will then resemble Figure 2. With the correlation between “financial ability” and “family communication”, however, it is extremely difficult to determine which precedes the other.

Figure 2.

The correlation of family financial ability, family communication and abilities β (revised model)

Therefore, rather than a relationship of cause and effect, it is appropriate to view their relationship as interrelated or correlating. The single-headed arrows in Figure 2 indicate a cause-and-effect relationship, and the double-headed arrows represent a corresponding relationship.

Family Communication and Extra-family Communication

In seeing a mother excessively concerned over communication with the child, Kindaichi (2006) offers the following:

When children become adults, they reduce their communication with their parents. This is necessary for establishing independence. The parents, however, are unaware of this and continue to treat their children as children and bemoan the lack of communication.

Writing something that is so abundantly clear becomes idiotic; however, it is surprising how many people there are that remain unaware of this basic fact.

(Kindaichi, 2006: 76. Emphasis by this author.)

If, for example, the causality model for skills formation shown in Figure 1 is correct, then the child's establishing independence (the reduction of family communication) creates a dilemma as it causes the formation of “ability β” to retreat. As Honda's (2005) research on the importance of family communication only covers primary, junior high and senior high school students, its stands to reason that the formation of “abilities β” will be accelerated by not allowing the child to establish independence. Nevertheless, in actuality, the child's relationship with the parents changes as the child grows and develops, makes new friends, forms relationships with upper and lower classmates, and passes through the rebellious age. The reduction in communication is hardly a rare thing. Also, Honda presents the definition of the “formation of postmodern skills,” which corresponds to “abilities β,” as “the mutual formation of a flexible network between differing individuals, and the occasional use of others as a response to need” (Honda, 2005: 22). These relationships therefore are not limited to family, and should rather been seen as the construction of new human relations composed of those with different backgrounds and values from those of the immediate family. Therefore, it is difficult to imagine that the frequency of communication that is limited by the nature of inter-familial relationships directly influences skills formation as a linear mechanism. At the very least, it is difficult to say that this model theoretically describes the process of human growth and development. Therefore, this research includes the indicator of “extra-familial communication” with the indicator of “family communication” in order to pursue an understanding of what influence the shift or move from “family communication” to “extra-familial” communication may have on skills formation.

Up till this point, the amount of research that has focused on extra-familial communication has been relatively small. Reliable research produced in the field of family sociology, for example, Axinn and Barber (1997), has clearly shown that the awareness of the sexual division of labor, which develops after leaving the home through life experiences with extra-familial people, erodes the representative traditional family view. Ghimire et al. (2006) have shown that non-family experience plays an important role when individuals select their spouses. With this, it becomes clear that the experiences of interaction with people outside the family provide children and young adults with the opportunity to access resources not held by members of the family, which then illustrates the process of socialization. In this fashion, extra-familial communication can be understood as part of the transition process in the construction of new social relations after leaving home; it can be seen as an indicator of the level of social independence. Based on the above, the variable of “extra-familial communication” can be introduced to “family communication”, which in turn redefines the process of skills formation. In general, it is difficult to imagine that “extra-familial communication” precedes “family communication.” Accordingly, it is therefore possible to suggest the model shown in Figure 3, as opposed to the Figure 1 model. The arrow in Figure 3 represents the cause-and-effect relationship.

Figure 3.

The correlation of family communication, extra-familial communication and abilities β (revised model)

The Process of Skills Formation and an Operating Hypothetical Model

Based on the above discussion, I wish to construct an operating hypothetical model for the process of skills formation. Before constructing this model, however, I shall confirm the definition and concept of “skills,” and also confirm the establishment of a corresponding variable.

As discussed in the third section, in this paper, “abilities α” are defined as “those skills that are quantifiable by paper tests, such as scholastic ability, grades or marks,” and “abilities β” as “those skills that include the emotions, personality, personal relationships, motivation, which are formed of elements that fall outside of ‘abilities α’.” Based on this definition, “abilities α” are represented by the two observable variables of “school mark/grades”7 and “level of education,”8 in which common elements are dormant. From this, it is possible to suggest that these can then be considered latent variables reflecting the subordinate concepts of both variables. As for “abilities β,” three of the twelve questions that comprised the research heading of “skills normally used at work” were focused on. The questions concerned: the degree to which the subject felt they succeeded in speaking before people (henceforth: presentation); demonstration of leadership (henceforth: leadership); and the ability to clearly explain one's thoughts (henceforth: self-expression).9

The method and reasons for this are as follows. In order to fully grasp the structure of adult abilities, all of the twelve items mentioned above were analyzed using factor analysis. I extracted two factors. For Factor 1, I extracted items that are more of the type of skill elements formed within the individual, such as: “the understanding of one's abilities and talents,”“construct an overall view of work and progress according to a plan,”“construct a system and think logically” etc. Items that contain dynamic elements of encouragement towards others, such as “the ability to speak well in front of people,”“demonstrating leadership” and “clearly explaining one's thoughts,” made up Factor 2. It is then possible to state that Factor 2 conforms to those elements that are deemed important in “abilities β”: “relations with others” and “motivation and emotion.” The three above items showed a strong factor contribution of over 0.600 for the second factor group. The same as “abilities α,” the latent elements of the three items above correspond to the concept of “abilities β,” which means that “abilities β” can been interpreted as a latent variable reflecting a subordinate concept of variable.

While “abilities α” can be defined at the point of graduating school, “abilities β” appear after graduation as skills in the workplace. It is therefore reasonable to assume that “abilities α” play a role in the formation of “abilities β” in the context of time. Consequently, the path from “abilities α” to “abilities β” is established.

A figure reflecting the operating hypothetical model of the skills formation process based on the discussion above in “Two Primary Factors in the Family Environment” and “Family Communication and Extra-family Communication” of “family environment” would resemble Figure 4. A perfect sequential model for a path from the three variables comprising “family environment” (“family financial abilities,”“family communication” and “extra-familial communication”) to the two variables (“abilities α” and “abilities β”) comprising “skills” has been hypothesized.

Figure 4.

Operating hypothetical model for skills formation and family background


The Variables Used in the Analysis

Before conducting the verification, the fundamental statistics for the variables used in the analysis will be provided and their correlation confirmed. First, the variables used in the analysis have all been represented in Table 1. Of these, those indicators of “family communication,” the target of this paper's argument, are points that must be kept in mind. As discussed above, Honda (2005) makes use of a composite set of variables derived from a set of scores indicating the frequency with which children discuss with their parents a number of topics that indicate “family communication,” representative topics being their studies, their future and personal concerns. This research, however, makes use of the single item of “speaking with one's parents about the future” as an indicator of “family communication.” In this investigation, however, this question was posed to young adults, between the ages of 25 and 35, about the frequency with which these topics were discussed with their parents. Because of this, it has been assumed that it is extremely difficult to pursue the contents of specific conversations, which in turn lowers the reliability of the answers; the measurements were therefore based on the one item above. Nonetheless, it was assumed that most responses to the question of whether or not they spoke with the parents during childhood would be an affirmation that they had, and the question was modified to specifically inquire about the frequency with which they spoke about their future with their parents. According to earlier research on the contents of conversations (Benesse Kyōiku kenkyū kaihatsu senta, 2005), the correlation between the frequency of conversation between parents with primary and junior high school children with skills (school marks) showed a correlation between the greater frequency with which children interacted with their parents, regardless of topic, and better school marks. Nevertheless, the possibility that this influenced the results of the analysis was left untouched, and I wish to state here that this remains a topic for future research.

Table 1. Variables in analyses
VariablesQueryScoreMeansStandard deviation
  1. Frequency for all variables is: n = 1078.

Family environment    
 Family financial abilityLooking back on your childhood during primary and junior high school, how were your family's financial circumstances?Comfortable: 5, Moderately Comfortable: 4, Average: 3, Moderately Strained: 2, Strained:
 Family communicationLooking back on your childhood during primary and junior high school, how often did you discuss the following: your futureOften: 4, Occasionally: 3, Seldom: 2, Never:
 Extra-familial communicationLooking back on your childhood during primary and junior high school, how often did you communicate with people outside of your parents and your teachers?Often: 4, Occasionally: 3, Seldom: 2, Never: 1.2.430.81
Abilities α (Latent Variables)    
 School marksWhat were your marks/grades during primary and junior high schools? Please provide answers for both.Top Level: 5, Above Average: 4, Average: 3, Below Average: 2, Poor: 1(Average for both)3.851.02
 Years of educationWhich of the following education institutions have you graduated from?The last school I graduated from was: junior high: 9, senior high school: 12, junior college, college, trade school, university: 16, graduate school:
Abilities β (latent variables)    
 PresentationHow well do you feel you accomplish the following as part of your normal work? (Presentation)Above average: 4, average: 3, below average: 2, not at all:
 LeadershipHow well do you feel you accomplish the following as part of your normal work? (Demonstrate leadership)Above average: 4, average: 3, below average: 2, not at all: 1.2.310.77
 Self-expressionHow well do you feel you accomplish the following as part of your normal work? (Clearly express your views)Above average: 4, average: 3, below average: 2, not at all: 1.2.590.69

Table 2 represents the correlation between the respective variables. I wish to point out three characteristics. The first is that it is possible to confirm a strong positive correlation exceeding 0.400 between the variables that transform “abilities α” and “abilities β” into latent causes and effects (“primary and junior high school mark” and “educational level”: “presentation,”“leadership” and “self-expression”). The second characteristic is that the comparatively strong correlation of 0.372 can be seen with “family communication” and “extra-familial communication.” The third characteristic is that, in viewing the coefficient for the group of variables that reflect family environment and the group of variables that transform “abilities β” into latent cause and effect, it becomes clear that “extra-familial communication,” rather than “family communication,” has a strong correlation with the variables that transform “abilities β” into latent causes and effects.

Table 2. Family background and the process of skills formation (structural equation modeling)
  1. Frequency for all variables is: n = 1078. All values have a significance level of 5%.

Family environment        
 1 Family financial ability0.0780.1000.1360.1770.1230.1370.140
 2 Family communication 0.3720.1420.1400.2390.1890.245
 3 Extra-familial communication  0.0720.0380.2710.2630.253
Abilities α (latent variables)        
 4 Primary and junior high school marks   0.4140.1880.1750.188
 5 Years of education    0.1450.0980.091
Abilities β (latent variables)        
 6 Presentation     0.6350.575
 7 Leadership      0.527
 8 Self-expression       

Structural Equation Modeling

With the above in mind, I shall progress to an empirical analysis. Figure 5 represents the results of a structural equation modeling of the operating hypothetical model, shown in Figure 4, into which the variables shown in Table 1 have been introduced. The numerical values are a standard regressive coefficient. The solid line arrows represent significant paths while the broken line arrows represent those without significance.

Figure 5.

Family background and the process of skills formation (structural equation modeling)

χ2/d.f. = 65/28 = 2.32, Goodness-of-fit index = 0.991, Adjusted-goodness-of-fit index = 0.976, Root mean square error of approximation = 0.028

1 Analysis only for employed men (n = 1078)

2 The numerical value is a standard regressive coefficient

3 Significance levels: *** 0.1%, ** 1%, *5%

4 The broken line represents a Pass Analysis of a 5% level, not significant

First, in order to judge the degree of conformity for the overall model, χ2/d.f.,10 goodness-of-fit index (GFI), adjusted-goodness-of-fit index (AGFI) and root mean square error of approximation (RMSEA) were calculated, the results for each being: χ2/d.f. = 2.32, GFI = 0.991, AFGI = 0.976, and RMSEA = 0.028. For a model without the slightest configuration for error correlation, the data application is perfectly acceptable.

From the two established latent variables “abilities α” and “abilities β”, if the paths subtracted from two observed variables are examined, both of them have a large value of over 0.650, which signifies an average of 0.01%. Therefore, it can be said that “abilities α” and “abilities β” have been shown according to an assumed observed variable.

If the variables outside of the observed variables that compose the two latent variables – in order from the upper left of Figure 5: “family financial ability,”“family communication,”“extra-familial communication,”“abilities α” and “abilities β”– are examined, I would then like to focus on the standard regressive coefficients subtracted from the respective variables. They can be reduced to the five characteristics given below.

  • 1In examining the model as a whole, in all of the paths, the highest value was “family communication”→“extra-familial communication” at 0.372***, followed by “extra-familial communication”→“abilities β” at 0.262***. From this, it can be understood that, as an intermediary, “family communication” to “extra-familial communication”, the indirect results of being connected with “abilities β” has a decisive influence, while “extra-familial communication”→“abilities α” has next to no significance for those skills that are quantifiable through paper tests, such as scholastic aptitude and school marks.
  • 2In the model, “abilities α”→“abilities β” is the third highest standard regressive coefficient at 0.236***. Those skills that are measurable through paper tests (scholastic aptitude and school marks) do have a positive role, on those skills that include motivation, interpersonal skills, personality and emotions.
  • 3On “abilities α”, “family financial ability” (0.225***) and “family communication” (0.205***) had an unmistakable influence, and while the difference is small, the family's financial abilities had a more significant influence. It is clear that the family's financial capital as a primary factor was linked to the measurable skills (scholastic aptitude, school marks).
  • 4On “abilities β,” the direct results of “family financial ability” (0.076*) and “family communication” (0.135***), when compared to the direct results of “extra-familial communication” (0.236***), are considerably lower. From this, it is impossible to claim that the primary cause of the financial capital and the primary cause of the cultural capital of the family have a decisive influence in the development of skills in adulthood.
  • 5In viewing Honda's model (Figure 1), which corresponds with “family communication”→“abilities β,” the influence of other variables has been excluded, and observable direct results have been given a significant value (0.135***); however, these direct results are intermediated through other variables derived from “family communication,” and do not even amount to half of the total results (0.135 + 0.145 = 0.280), which include the intermediate results (0.145) connected with “abilities β”: 0.135/0.280 × 100 = 48.2%. Conjecturing from the above, there is the possibility in Honda's (2005) analysis of a large apparent correlation from an essential influence between “family communication” and “abilities β,” which arose due to an inflation caused by the influence of the indirect results of “family communication” to “abilities α” or “extra-familial communication.”

Results and Future Inquiries

In order to clarify the process of skills formation over the extended period of time from childhood to adulthood, this research has constructed a working hypothetic model based on the results of Japanese and international theoretical and empirical research, and the research data were then verified. The first result is the significant influence during childhood of the family environment on the formation of skills during adulthood. The second clear result is that this process is not linear, but rather passes through a complex and diverse process.

The first point – the results obtained in this research that show the family environment during childhood is a primary factor in the formation of skills during adulthood – reconfirms the facts indicated in the comparatively recent research of both Cunha and Heckman (2007), and Honda (2005) discussed at the beginning of this paper.

The second point has been confirmed anew by the verification of both the original working theoretical model constructed and the results of the research data, rendering it possible to discover the following. As pointed out in the third section, Honda's (2005) research states that the decisive influence in the formation of “abilities β” is family communication (Figure 1). Then, having formed the suggestion of the impossibility of social intervention from outside the family into this family communication as her base, Honda identifies the following two suggestions: the first suggestion is that the disparity and inequality in the formation of skills is due to the family into which one is born (Suggestion 1), and the second is the belief in the “perfect mother” seen during the child-rearing process (Suggestion 2). As indicated in the fifth section, however, the direct results of family communication on the formation of “abilities β” is actually less than the indirect influences. Instead, the composite results of the direct and indirect results of the family's financial abilities, and the direct results shown in the children's process of establishing independence through the shift from family communication to extra-familial communication largely influence the formation of “abilities β.” That is to say, this indicates the unmistakable existence of the route of skills formation (both the direct and indirect results connected with “abilities β” from the family's financial abilities and the method of encouraging the direct results connected with “abilities β” from extra-familial communication) that does not exist in family communication, into which social intervention is difficult. Based on the above, it can then be asserted that social intervention is possible in Suggestion 1. With Suggestion 2, based on the data, it is necessary to urge the recognition anew, and with an awareness of the new knowledge of skill formation discussed above, it is possible to anticipate more rational discussion in the overheated discourse on the role of the family in education.

I wish to finish with a review of remaining topics, the first of which is a discussion on the benefits of the meaning and the function of “extra-familial communication,” which is thought to be an invaluable node in the formation of individual skills. In this paper, this has been posited as an analysis and interpretation of “the process of establishing independence by children.” With the findings, however, the discussion of the transition from child to adult (adolescence research, for example, is a well-known field) seen from the viewpoint of human growth and development is extremely simplistic. Anderson and Sabatelli (2007) have proposed the extremely interesting concept of “individuation,” and it is necessary to be completely aware of this type of research.

The second topic is a discussion of the political development of analyses results. In this paper, from the results of the analyses, it is clear that social intervention was possible with regards to the problem of disparity and inequality in the formation of skills. There is, however, the problem of whether or not intervention should be undertaken, and it is necessary to discuss the problem of which methodology to use if intervention should be undertaken. With the former, it is necessary to discuss, from a legal and moral stance, social psychology as well as ascertaining the financial results. According to the analysis in this paper, the influence of “family communication” on “abilities β” figures significantly in the overall results. In addition, the influence of “family communication” on “extra-familial communication” is comparatively strong. A scrupulous investigation into whether or not policy effectiveness is desirable as a countermeasure to this is necessary. With the implementation of the latter, and should the policy purpose provisionally be “the promotion of family communication,” one concrete possibility (the methodology and results of which are considered to be well-established) is, for example, “mentoring” (Dondero, 1997). According to the analyses results in this paper, however, extra-familial communication has absolutely no influence on “abilities α.” The point in question here, however, is whether or not it is possible to make a previous investment in those skills that appear during adulthood with an eye to a long-term return as opposed to extra-familial communication, which has no connection with readily recognizable skills formation. This discussion, however, is completely from the macro (labor) policy stance. If the first point is used as a basis, a policy of this sort would be a necessary and extremely fundamental form of support for the transition of child to adult. The development of a new research model that would address the conflict between the stance of support for individual development and growth viewpoint with the macro policy stance is much needed.


  • 1

    Honda (2005) provides two concepts to represent the skills of children and young adults: modern skills and postmodern skills. See footnote 5 for more details.

  • 2

    As of the moment, there is little research that provides academic critique, advocating or developing this thesis besides Takeuchi's 2006 review and Ushiogi's 2006 citation.

  • 3

    The nature of the hypothetic model is such that the most desirable investigation would essentially be panel data that covers childhood to adulthood. Given, however, the low feasibility of immediately including the early period in the investigation, the analysis was based on data taken from recollection.

  • 4

    As for topics and response trends in monitor research via the Internet, Honda's (2006) in-depth investigation showed that it was impossible to state that the monitor research was the only strain. With this research and the rapid growth in social interest in personal information, taking into account the unlikely high return of postal research or interviews with young people as the research target, personal information was kept confidential, and the research was conducted with the web monitor as the population, given the relatively high frequency with which the younger population appears.

  • 5

    Honda (2005) provides two concepts to represent the skills of children and young adults: modern skills and postmodern skills.

    “Modern Skills”“Postmodern Skills”
    1. Honda (2005: 22).

    [Basic Scholastic Aptitude][Strength to Live]
    NormDiversity and Novelty
    Knowledge Processing SpeedMotivation Creativity
    Common Measurement Comparative MeasurementIndividuality Personality
    Cooperativeness, HomogeneityNetworking Abilities Negotiation Abilities

    Presumably, the concept of postmodern skills is to represent those skills that are necessary to exist in the postmodern fluid risky labor market; however, this author should exercise care in using the terms modern and postmodern to represent the transition of one era to another. Cunha and Heckman (2007) divide human skills into two large groups of “cognitive ability” and “non-cognitive ability,” the former being defined as those skills that are primarily quantifiable through IQ tests, for example, while the latter skills are defined as being composed of those elements that fall outside of quantifiable skills, such as motivation, patience, self-esteem, self-control, risk aversion and time preference. Honda's (2005) “modern skills” and “post-modern skills” correspond to “cognitive abilities” and “non-cognitive abilities.” Research by Bowles et al. (2001) concludes that “non-cognitive abilities” are equally important with “cognitive abilities” in accurately estimating the allocation of the individual's income and social position, but as of yet, there is no research that substantiates the suggestion that “non-cognitive abilities” are becoming more important than “cognitive abilities.” Based on Kosugi's (2006) argument—Kosugi being well versed in domestic employment trends—it is not possible to state that the relative importance of each has undergone any change over the last few years. Because of this, in this paper's argument, I have, for the present, defined “those skills that are quantifiable through paper tests, such as scholastic aptitude and school marks” as “abilities α,” and “those skills that fall outside of “abilities α” that are thought to be composed of elements such as motivation, interpersonal skills, personality emotions and others” as “abilities β.”

  • Honda (2005: 22).

  • 6

    This indicator is “skills and self-evaluation” and is therefore different from the objective meaning of “skills.”Rothwell and Arnold (2007) have organized the precise points in question for “skills and self-evaluation” at the workplace and have conducted a careful analysis and pointed out its usefulness. As the basis for their research was adult white collar workers, a future research topic is the meaning of “skills and self-evaluation” with, for example, the subjects in this paper, which includes adult non-white collar workers as part of the larger group of employed men, or, for example, the children used in Honda's (2005) analysis.

  • 7

    Please see Figure 1 for details concerning the variables used in the analysis.

  • 8

    Please see Figure 1 for details concerning the variables used in the analysis.

  • 9

    Please see Figure 1 for details concerning the variables used in the analysis.

  • 10

    In general, the indicator χ2 used to determine the degree of conformance in the structural equation modeling is subtly influenced by the sample size. To avoid this phenomenon, when the sample cases were sufficiently large, the computational method was changed to χ2/ d.f., the criterion method of model usage when this is below 3.00 being known (Kline, 2005).