4. Results and discussion
Data collection was carried out from May to August
2012, yielding a sample of 1046 valid respondents.
The majority of the respondents are women (57.5%),
married (42.7%), or single (47.9%). In addition, 65% of
the respondents have their own home, and 55.6% have
no dependents. The majority of the respondents indicate
their race to be Caucasian (80.5%). With respect to the
educational level, some respondents did not finish high
school (22.1%), others completed high school (19.6%),
and some have incomplete higher education (14%). With
respect to occupation, most respondents are private-sector
employees (33.6%), civil servants (20.9%), or self-employed
(19.7%).
Concerning financial matters, a significant proportion
of individuals have a monthly household income of one
or two times minimum wage (24.1%). Regarding the use
of credit cards, a representative percentage (56.6%) uses
credit cards, with the greatest proportion using one (29.1%)
or two credit cards (18.5%).
After profiling the respondents, we analyzed mean
scales and subsequently examined the factors. Table 2
shows all variables used in each scale, with the respective
averages of the responses. It is important to highlight that
the scales used in this study are five-point Likert scales.
Validation of the constructs was then carried out. To
this end, confirmatory factor analysis was considered.
Relationships between the observed variables and their
constructs were examined via estimation of maximum
likelihood. The results obtained from construct validation
are shown in Table 3.
Regarding the relations of demographic and cultural
variables to indebtedness, we note that there are significant
differences in debt according to age, gender, marital
status, education, religion, religious principles, occupation,
income, use of credit, dependence on credit, and expenses.
We find that people who have not formed a family or
who already have one but are living alone (widowed, divorced,
or single), tend to have a higher propensity toward
indebtedness, which may be attributed to not having an
exclusive commitment toward a family. Individuals without
literacy and who do not work tend toward a greater
propensity for indebtedness; these results are similar to
those of Gathergood (2011).
People under 30 have a higher level of debt, in accord
with the results of Gathergood (2011) and Sevim et al.
(2012). It is also shown that men are more likely to be
in debt (mean: 1.97) than women (mean: 1.85). With respect
to religion, people who have no religion and follow
no religious principle are more likely to be in debt, supporting
Davies and Lea (1995). In regard to income, it is noted
that those with salaries in the lowest (up to minimum
wage) and highest (more than 20×minimum wage) ranges
have the greatest propensity toward indebtedness; this
outcome matches that demonstrated in Katona (1975), revealing
that there are two main reasons for indebtedness:
low and high income. In matters of credit, respondents who
use and depend on credit cards are most likely to be in debt.
For construction and validation of the constructs, confirmatory
factor analysis (CFA) was used, which is adequate
for all factors addressed in this study: financial literacy, risk
perception, risk behavior, emotion, materialism, indebtedness,
and value of money. Therefore, this study seeks to
build an integrated model that combines the measurement
model and the structural model. The theoretical model is
evaluated based on the fit indices and the statistical significance
of the estimated regression coefficients.
The initial model was estimated; due to problems of
adjustment, some changes were made. The final model
obtained after modification is shown in Fig. 2. Table 4
illustrates the equation model fitting process. Tables 5 and
6 illustrate the standardized coefficients and the model
fitting.
Fig. 2 shows the final model, with the validated factors
and their formative variables. However, it is clear that
from the factors cited, financial literacy is not in the final
model. The hypothesis for this issue is not confirmed,
which requires exclusion of this factor.
The indices shown in Table 6 reach the appropriate
limits. The standardized coefficients of the final model are
significant.
Five relations are negative: materialism and emotion;
materialism and risk perception; value of money and indebtedness;
risk perception and indebtedness; and indebtedness
and emotion. By verifying these relations, we note
that people with higher levels of materialism are accustomed
to a high level of consumption, which hinders their
perception of risk and consequently increases their risk
behavior. This behavior reduces the probability of experiencing
negative feelings toward indebtedness because
purchasing usually gives such people positive emotions.
The discussed emotions-based approach in this work is
negative; therefore, the relation between materialism and
emotions is negative.
In general, addiction to material goods (high materialism)
at first brings about a sense of wealth and quality of
life. It is because of this feeling that humans by nature conclude
that the more possessions one obtains, the higher
6 S.A.M. Flores, K.M. Vieira / Journal of Behavioral and Exp