|dc.description.abstract||The purpose of this research was to develop a preliminary multidimensional
measurement of self concept, within the framework of
phenomenological theory and acknowledging previous work in the field.
Twelve preliminary scales were investigated in terms of reliability,
validity and cross-validation. Further work was not done to establish
normative data, accounting for demographic differences.
Self concept was defined as "an organized, fluid, but consistent
conceptual pattern of characteristics and relationships of the "I” or
the "me", together with values attached to these concepts" (Rogers,
1951, p. 498).
Two groups of participants consisting of university and college
students at all levels were used. The first group (n = 325) was used
for item analysis and cross-validation measures, and the second group
(n = 96) for retest reliability and validity measures.
The questionnaire used consisted of thirteen scales: five scales
represented an internal frame of reference, and seven scales an external
frame of reference of the self concept. The last scale was a Lie Scale,
or "Faking Good".
There were four major sfages to the research: item develofxnent
and three psychometric studies; two approaches were used in it&n development;
the development of items for the Family Self, Social Self, Religious
Self, Occupational Self, Masculine/Feminine, Body Self, Public Self,
and Past/Present/Future Scales; and the revision of items for the Lie, Enpathy, Altruism, Self Esteem, and the Purpose In Life Scales, v^ich
were modified from already developed tests.
Prior to the first administration of the questionnaire, nine judges
rated each item of the developed scales for face validity. Inter-judge
reliability was not performed on the judges' responses due to the high
agreement found between judges.
The process of item analysis was completed in Study One. The best
16 items that resulted in the highest item-total correlation within
the eight developed scales were retained.
The second study established reliability and validity data on the
scales after item analysis. Retest reliability on 24 college students
over a three month interval was unacceptable for all scales, with the
exception of the Masculine/Feminine Self Scale. Overall, coefficient
alpha results were poor, with only two scales (Past/Present/Future and
Errpathy Scale) meeting the criteria of acceptable hcmbgeneity estimates
(n = 96). Convergent and discriminant validation was estimated using
the Counselling Fom of the TSCS and 16PF (n = 35). Results of the
TSCS revealed that most predictions made, occurred. Only a few predictions
were successful vdien the revised questionnaire's scales were
correlated with the factors of the 16PF Test. Factor analysis for
individual scales revealed ambiguous findings.
Study Three, the cross-validation estimates, were found to be more
favourable. However, the sample size was small (n = 41). Overall, the studies suggested that there is a firm enough base
to warrant further development of this multidimensional self concept