This thesis considers the nature and extent of constraint and internal consistency within the sets of beliefs of selected groups of Australians. Because of the extensive implications of issues of consistency and constraint within the social sciences generally, a substantial review of the literature concerning attitudes, beliefs, values and ideologies is provided. A perspective is developed whereby 'belief’ becomes the basic building block of other social science concepts such as attitudes, values and ideologies.
Particular attention is directed towards methodological issues. Many previous studies involved with the measurement of consistency and constraint have made use of average coefficients of group association (such as pearson correlation and gamma coefficients). It is shown within the thesis that great care must be exercised in the use of such methodologies. An alternate method based upon the average standard deviation of standardised items within a construct is proposed and used by the author. A simulation approach is then used to calibrate a scale based upon these standard deviations.
Data was gathered from approximately 1400 respondents. These respondents included groups of bank officials, building workers, manager?, and trade union officials. Two questionnaires were used covering both social and work related belief items. The opinions of respondents were collected in the following areas:
(a) Nature and the Nature of Man
(b) The Meaning of Work
(c) Progress Change and Stability
(d) The Perfect Market
(e) Private Property
(f) Freedom Justice and Equality
(g) The Individual, the Community and Government
(h) The Distribution of Power in Australia
(i) Trade Unions
(k) Industrial Dispute Activity
In all, data was collected on some three hundred items. The collected data was analysed using basic inferential statistics and factor analysis.
Following a preliminary frequency distribution check, the items were grouped into appropriate sub sets and then factor analysed. Two random split halves were used and only factors replicated for each sample were accepted. Factor scores were created for the constructs so formed. These first order factors from each of the areas were then placed in a common pool and second order factors extracted.
The results indicate that both work and social beliefs are constrained at the group level. At the individual level much inconsistency is evident. However, it is contended that this inconsistency is often an artifact of methodologies based upon clustering or factor analytic techniques, rather than an indication of individual inconsistency. Suggestions are given as to how such undesirable consequences may be avoided.
The internal constraint present within the data generated a number of important relationships. The most important of these is, that beliefs concerning industrial relations issues are strongly related to the experienced quality of work life. Those who are most favourably disposed towards trade unions and militant industrial action experience the least amount of quality of work life.
A strong polarisation of opinion between managers and blue collar trade union leaders is also evident. These groups hold opposite beliefs concerning:
(a) Industrial life as a continuation of the class based divisions of society;
(b) Living in harmony with nature;
(c) The value of the modern free enterprise system;
(d) The value of individual property ownership;
(e) The experienced quality of working life;
(f) The extent to which industrial action should be carried.
The closing section of the thesis relates substantive and methodological findings back to the body of theoretical knowledge from which they were derived.