The symbolic boundaries of Zoroastrianism: 1985 survey
The data generated by the 1985 survey of the Zoroastrian Diaspora has been thoroughly documented and analysed in Professor John Hinnells’ book The Zoroastrian Diaspora: religion and migration, Oxford University Press, 2005. The analyses discussed here will deal with the symbolic boundaries of Zoroastrianism which were touched upon in this work. A scale was built to measure the symbolic boundaries using the following variables from the data set: the respondents’ attitudes to:
1. Marriages between Zoroastrians and non-Zoroastrians.
2. Bi-cultural marriage ceremony before Zoroastrian Priests.
3. Children of bi-cultural marriage to be considered as Zoroastrians.
4. Conversion of non-Zoroastrians.
5. Naujote for non-Zoroastrians.
6. Community actively seek converts.
7. Non-Zoroastrians allowed in Temples in India.
8. Non-Zoroastrians allowed in Temples in Iran.
9. Non-Zoroastrians allowed in Temple in UK.
SPSS Factor Analysis was the statistical technique that was used to investigate this set of variables. The scale was constructed so that High meant support for keeping the symbolic boundaries as they are, whereas Low meant support for change to the symbolic boundaries. A full set of crosstabular analysis of the scale, Symbolic Boundaries of Zoroastrianism Boundaries, with variables from the data set, was undertaken to give an idea of which variables were, or were not associating with the scale. See Table 95.
The next step in the data analysis was to take a closer look at the variables from the data set and their relationship with the scale, UK Zoroastrian Boundaries, using the program SPSS Correlation. The correlation coefficient does not imply causation but is a measure of linear association. First there is the list of variables that had no correlation with the scale Symbolic Boundaries of Zoroastrianism. See table 96.
The results of the correlation testing between the scale, UK Zoroastrian Boundaries, and the groups of variables, at .01, or .001 significance levels, are given below in Table 97.
The groups of variables that had the strongest correlation with the scale Symbolic Boundaries of Zoroastrianism are given in Table 98.
In Table 99, the results for the combinations of variables that originally gave an explained variance in the scale of over ten per cent are listed, and the independent action of these variables within their combinations is given.
In Table 100 the results for the combinations of variables that originally gave an explained variance in the scale of less than ten per cent will be listed, and the independent actions of the variables within these combinations are given.
When the best predictors, from all the groups of variables that had a strong explained variance with the scale, UK Zoroastrian Boundaries were tested in combination with each other and the scale, there was an explained variance of (29%). Looking at Table 101 it can be seen that when zero-order relationships were considered, the best predictors of attitudes to the symbolic boundaries were Attending Zoroastrian for Religious Reasons and Self Description/ Zoroastrian, (r = .34) and (r = .34) respectively.
The independent action of the variables is given in Table 102.
There were similar results for the two diasporic groups ‘old’ Zoroastrian communities, Hong Kong, Pakistan, Kenya, and ‘new’ diasporic communities. The most important analyses for these communities are given below (Table 103).
The explained variance of the groups of variables is given, and the independent action, within each group of variables is shown in Table 104.
It is not known in advance whether a model such as linear regression is appropriate therefore it is necessary to look for violation of assumptions. The program SPSS Residuals was used to check for violation of assumptions. The scale, Zoroastrians "Old" Symbolic Boundaries, was checked and there were problems with the groups of variables when they were entered into linear regression equations in combinations of groups of variables in the Secondary Analysis. This is the reason that, as far as the major and minor characteristics of the "old" sample are concerned, only "implications" as to those areas where these characteristics might be found can be indicated.
Stepwise regression techniques were executed on the groups of variables to ascertain how much of the explained variance each group of variables contributed to the scale, Zoroastrians "New" Symbolic Boundaries. The regression equations were built up from the groups of variables and the scale where there were significant correlations between the independent variables and the scale. R Square gauges the amount of variance in the dependent variable scale, Zoroastrians "New" Symbolic Boundaries that was accounted for by the groups of independent variables. Of particular interest was R Square = .1 or over.
The amount of explained variance in the scale, and the independent action of the variables is shown in Table 105.
The linear regression equations for the combination of the groups of variables that had an explained variance in the scale, Zoroastrian "New" Symbolic Boundaries, of less than ten per cent were conducted. See Table 106. The best predictors from these groups of variables with regard to attitudes to the symbolic boundaries are given in Table 106.
When the best predictors, from the groups of variables that had a strong explained variance with the scale, Zoroastrian "New" Symbolic Boundaries, in the Secondary Analyses, were tested in combination with each other with regard to the relationship with the scale there was an explained variance of (14%). Looking at Table 108 it can be seen that when zero-order relationships are considered, the best predictor of attitudes to the symbolic boundaries is Self- Description/ Zoroastrian, (r = .24).
When the independent action of the variables is considered Self- Description/ Zoroastrian, (beta= .22), and Behdin Boys Undertaking Navar are the best predictors, (beta = -.22). See Table 8.
The low per cent of explained variance in the scales Symbolic Boundaries of Zoroastrianism was considered when the questionnaire for the 2003 survey was drawn up and new questions were asked about the Zoroastrian purity laws these were put into sophisticated statistical analyses which gave a very high explained variance of 70%. The question has then to be asked are the purity laws and the symbolic boundaries of Zoroastrian religion part of the same religious value. This has yet to be investigated.