3. Methodology
This study represents exploratory research. According
to Hair et al.(2003), the exploratory approach to research is
geared toward discovery and aims to test specific research
hypotheses. A theoretical model is developed to investigate
the influence of behavioral factors in debt situations.
Furthermore, to evaluate demographic and cultural variables,
10 relations are developed for testing. In total, 10 hypotheses
are considered relations, as shown in Table 1.
The 10 hypotheses refer to the above-described model,
which describes the relations among the constructs considered.
The 10 relations refer to demographic and
cultural variables and are analyzed from parametric hypothesis
tests; the T test is used for up to two groups, and
analysis of variance is used for more than two groups.
Regarding the theoretical model, it appears that the
first hypothesis establishes a relation between materialism
and propensity toward indebtedness. This finding is based
on Watson (1998) and Ponchio (2006), who show that
more materialistic individuals are exposed to higher levels
of propensity toward indebtedness. In Ponchio (2006), for
example, more materialistic individuals are more likely
to engage in credit for consumption purposes, showing a
more positive attitude toward debt, i.e., the higher the level
of materialism, the more likely the individual will be in
debt.
The second hypothesis refers to financial literacy. Based
on Chen and Volpe (1998) and Disney and Gathergood
(2011), we attempt to determine whether financial literacy
inversely impacts propensity toward indebtedness.
The third hypothesis of the theoretical model attempts
to identify the impact of the value of money on materialism.
Macedo et al. (2011) note that money changes interpersonal
relationships among individuals, who seek social
status through their financial behavior. This quest for social
recognition hinders the level of materialism, i.e., people
tend to use more and more money to acquire material
goods (not always necessary) that satisfy their desires and
indicate a better position in society. Therefore, a relation is
established between these two constructs (value of money
and materialism).
Next, two hypotheses related to the construct of risk are
developed, one focused on risk perception and the other on
risk behavior. In these relations, it is notable that the higher
the perceived risk is, the lower the level of debt becomes.
Moreover, the more conservative behavior toward risk is,
the lower the level of debt becomes (Caetano et al., 2011).
The value of money construct may also influence
indebtedness and emotion. Accordingly, Stone and Maury
(2006) investigate the personal aspects that influence
debt and note that factors such as obsession, inadequacy,
and retention of money are important in predicting the
condition of being in debt. People who save a greater
proportion of their income tend to appreciate it more
and therefore tend to have negative emotions, in case
they suffer serious financial problems and shortage of
money. On the other hand, there are people who prefer
to spend more money and experience positive emotions
from purchases, such as pleasure from and satisfaction
with purchases. According to Vohs et al. (2008), attitudes
toward money can cause motivational, emotional, and
behavioral changes.
With respect to the construct of emotions, two
hypotheses can be set forth that relate emotion to
risk. Källmén (2000) examines the hypothesis that the perception
of risk is related to personality differences, through
the locus of control, self-efficacy, and anxiety, where the
last of these (anxiety) involves emotions caused by the environment.
This author notes that people with lower levels
of anxiety and emotion manage risk more efficiently than
those with high levels of anxiety. Zuckerman and Kuhlman
(2000) suggest that high-risk behaviors (such as reckless
driving) may be caused by the need to express emotions.
These authors stress that people search impulsively for
feelings involving positive and negative emotions. As a result,
they tend to develop (or not develop) riskier behaviors
to experience the desired sensation. Thus, the two
hypotheses developed here are impact of emotion on perception
and that on risk behavior.
The final hypothesis of the theoretical model aims
to measure the relation between propensity toward
indebtedness and emotion. It is notable that consumption
affects people emotionally and physically; it may cause
positive emotions, such as joy and satisfaction, or negative
emotions, such as sadness (Quelch and Jocz, 2007). Given
these perspectives, the following hypothesis is analyzed:
the propensity toward indebtedness impacts emotion.
From these hypotheses, a theoretical model is set forth
in Fig. 1.
The setting for this research is the city of Santa Maria,
located in the state of Rio Grande do Sul, Brazil. According
to census data from 2010, available from the Brazilian
Institute of Geography and Statistics (IBGE), the city of
Santa Maria has an estimated 261,031 inhabitants. The
sample is characterized by the confidence level of 95% and
a sampling error of 3.2% at 973 respondents.
The study design was submitted to the National Research
Ethics System; it was approved on April 17,
2012, under the identification number 11831 and Certificate
of Presentation to Ethical Consideration number
02054312.5.0000.5346.
The data collection instrument is a structured questionnaire
with open and closed questions, divided into nine
sections. The first section addresses the profile, whereas
the second section considers aspects relating to expenses;
the third phase of the questionnaire addresses individual
debt. The remaining sections explore behavioral factors
based on the following references: financial literacy,
using the scale of Matta (2007) and Disney and Gathergood
(2011); risk, using the scale of Paulino (2009);
emotions, based on Disney and Gathergood (2011); materialism
and indebtedness, using the scale of Moura (2005)
and Disney and Gathergood (2011); and finally, value of
money, using the scale of Moreira (2000). A five-point Likert
scale for six factors is used: perception of risk–no risk
up to extreme risk; risk behavior and emotions—very unlikely
to very likely; indebtedness, materialism, and money
values—strongly disagree to strongly agree. For the factor
of financial literacy, we use a four-point scale—never to
always.
Model estimation and validation employs structural
equation modeling. In the structural model, many multivariate
equations have been used to predict and explain
a set of endogenous and exogenous constructs. To this
end, the principal equation in modeling is represented by
Eq. (1). (Hair et al., 2006, p. 672).
η = Bη + Γ ξ + ζ (1)
where
η = is an m × 1 vector of endogenous latent variables;
B = is an m × m matrix regression coefficients relating the
latent endogenous variables to each other;
Γ = is an m × k matrix regression coefficients relating
endogenous variables to exogenous variables;
ξ = is a k × 1 vector of exogenous latent variables;
ζ = is an m × 1 vector of disturbance terms.
The convergent validity of each construct is assessed
by observing the magnitude and statistical significance of
the following standardized coefficients and fit index of the
model: χ
2
(chi square) statistic, root mean square residual
(RMR), root mean square error of approximation (RMSEA),
goodness-of-fit index (GFI), comparative fit index
(CFI), normed fit index (NFI), and Tucker–Lewis index (TLI)
or non-normed fit index (NNFI). It is emphasized that to
reduce sensitivity toward sample size, some researchers
divide values by degrees of freedom (χ
2
/GL). For example,
Hair et al. (2006) consider values equal to or less than
5 as acceptable.
For the indexes, acceptable values are as follows: GFI,
CFI, NFI, and NNFI equal to or greater than 0.90, RMSEA less
than 0.08, and RMR less than 0.10 (Kline, 1998; Garver and
Mentzer, 1999; Hair et al., 2006).