A Comparative Study of Filial Piety in Beijing and the UK

Cultures may display paradoxical natures when studied closely. For example, Americans are individualistic, yet are most generous with charity giving or volunteering for community events. This illustrates the challenges associated with stereotyping groups and cultures. Cultures are sometimes treated as homogenous with attendant generalisation. In the educational context, generalisations consequently influence attitudes held towards groups of learners. People from particular cultural backgrounds are often treated as a homogenous group with its attendant stereotypes irrespective of their individual contexts; an example is individuals from a Chinese cultural background who live in the West relative to those living in China. The present study used a questionnaire to measure an important value in Confucian culture (Chinese background) Filial Piety and how it relates to affective variables suggested to be important to academic achievement: self-efficacy and motivation. A sample of 9-11 year old children from the UK were compared with a similar sample from an authentic Confucian context Beijing. The results suggest a review of expectations of what is meant by “authentic confucian” and gives some insight into the potential problems that could arise with stereotyping cultures and groups. As schools reach out to parents and communities, there are implications for the stereotypic approaches adopted.


•
The relationships between FP and Received self-efficacy (RSE)/ vicarious self-efficacy (VSE) were similar to that observed in the high FP group in the UK sample;

• The relationships between FP and Subjective norms (SN) and Intention (INT) [and indeed
Attention (ATT)] were like the observation in the high FP group in the UK sample.
This will hopefully enable a comparison to be made between a sample drawn from a modernised Chinese city such as Beijing and a group high in FP/ a group low in FP in the UK. It will provide a strong test of the generalisability of stereotypes held about groups of people. This will also add to the debate about the extent of the decline or otherwise in FP in modern Chinese society, and shed light on how typical a high FP group is of an authentic Confucian community.

Measuring Filial Piety
Measurement of FP has evolved over time in line with the changing definition and nature of the construct over generations (Lum et al., 2015).
Conceptual advances led to the development of the Dual Filial Piety Scale (DFPS) by Yeh and Bedford (2003). The dual scale merged the traditional conceptualisation for FP (authority and obedience) with the subsequent emphasis on emotion and reciprocity (Lum et al., 2015).
Conceptualising of FP has therefore evolved over time to an understanding that reflects both its traditional and current views. For instance, Lum et al. (2015) outlined six traditional and contemporary dimensions of FP: (a) conditional and unconditional care for parents, (b) care for oneself, (c) affection and consideration, (d) obedience and repayment, (e) honour and respect, and (f) family continuity.

Method
Quantitative data was collected using a questionnaire that had a 7 point Likert scale ranging from "strongly disagree" to "strongly agree".

Development of FP Measure
Elements from the dimensions posited by Lum et al. (2015) and Jones, Lee and Zhang (2011)  is an individual's conviction in their ability to successfully execute the behaviour needed for a successful outcome on a task (Bandura, 1993).
• Experiential (ESE): this refers to the learner's own experience of previous attainments.
• Received (RSE): is when a learner is told by someone 'you can do it' especially from someone they respect and whose opinion matters to them. Teachers' and parents' feedback is very important.
• Modelling/Vicarious (VSE): this is when a learner sees or watches someone do it (someone just like them) through vicarious learning. Observing someone like them succeed or fail at a task contributes to shaping their own sense of SE.

Self-efficacy and Motivation
The three sources of self-efficacy and the three elements of the TPB framework used were measured using items that targeted specific target behaviours, judged to be especially relevant to filial piety: Feedback, Grades, Concentration and Time spent. These four were chosen because they are behaviours that children perform or are concerned about in relation to pleasing their parents (see Chao & Tseng, 2002;Hong & Howes, 2014;Huang & Gove, 2012;Sham & Woodrow, 1998). The target behaviours fulfilled the multiple act criterion which according to Ajzen and Fishbein (1980), gives better measures of attitudes and behaviour.

Sample
The samples were recruited from children aged 9 -11 years in UK primary schools and a primary school in Beijing. The Beijing sample was drawn from an inner city primary school. All the participants were from a lower middle-class background with a few described as working class drawn from the same school. Official Mandarin (Putonghua) is the official and only language of instruction in the school. All the children were described as native Mandarin speakers.
The breakdown of the total sample is given in Table 1a and b below. The results from comparing the means of the two groups created by the median split were in line with what was expected for most of the variables. RSE and VSE were relatively higher in the high filial piety (HFP) group and the differences were statistically significant as expected. ESE was marginally higher in the HFP group but was not statistically significant.
Furthermore, SN, INT and FP were similarly higher in the HFP group which is to be expected for SN and FP. However, ATT was also higher in the HFP group and the difference was statistically significant; this was unexpected because the expectation for ATT was for it to be higher in the group with low filial piety (LFP).
Use of FP as a distinguishing measure for the two groups was therefore deemed to be appropriate for the purposes of this study. It produced groups that were different on most of the key variables as was hypothesised, with the exception of ESE. The difference observed for ATT in the two groups was statistically significant albeit contrary to what was expected.

Correlation Analysis
With the differences between the two groups established, correlation analysis was used to check the relations between the variables in the two groups. Since a key assumption under which correlation analysis is conducted is that the variables must have a linear relationship, all the relationships to be tested were checked for linearity using scatter plots (Cohen, Cohen, West, & Aiken, 2013).

Discussion of UK Sample
A key consideration is the extent to which filial piety (FP) worked as a differentiation device. Using the level of FP as the basis to split the data worked well; the two groups created were different (statistically) on all the variables except one. Furthermore, there was a good range of scores for FP. Therefore, it can be argued that dividing the sample up using a median split produced two distinctive groups.
In the high FP group, both subjective norms (SN) and attitudes (ATT) were associated with intention Bempechat and her colleagues (2016) further reported that learning beliefs that were culturally informed and internalized were able to have a positive effect on children even from disadvantaged socio-economic backgrounds.

Beijing Compared with High FP and Low FP
The three groups (high and low FP from the UK, and Beijing) were compared by means of all the variables using a one way Anova (Tables 4a and 4b). Subscripts are markers for significant differences. below the low FP group on all the measures except in VSE and FP.

Beijing Correlations
The correlations obtained from the data (Table 5) broadly suggests a trend expected of individuals from a Confucian culture but with a strong influence of ATT as well. ATT had a stronger relationship with INT than with SN even though both were statistically significant. The relationship between RSE and SN was stronger than that between ESE and ATT, suggesting a stronger influence of group norms over the individual. It must be noted however, that the ESE and ATT relationship was also statistically significant. Also, the relationship between RSE and SN was marginally stronger than that observed between RSE and ATT.
In line with typical expectation of individuals from a Confucian background, RSE and VSE both had stronger relationships with FP than ESE; FP and VSE and FP and ESE. There was also a statistically significant relationship between FP and SN but a non-significant relationship between FP and ATT.

Discussion
The means suggest the levels of the variables were lower than both high and low FP group from the UK sample; however, this could be due to differences in calibration. It is becoming increasingly reliable to agree with other researchers in the field who have opined that the view of children in Confucian communities lacking autonomy and a personal attitude as being