Data and code for 'Analysis of UVR Mortgage Credit Projections in Colombia under the Diffusion of Innovations Theory Approach'

This research is guided by the hypothesis that the persistence of misconceptions in the projection of UVR-denominated mortgage amortizations is explained by a set of factors drawn from Rogers’ diffusion of innovations theory. Specifically, it is proposed that these misconceptions are sustained by limited conceptual understanding of UVR behavior, the institutional circulation of simplified or incorrect projection practices, organizational routines that legitimize such practices, and the influence of peer learning within financial institutions. These elements, operationalized through four latent constructs, are expected to jointly account for the continued use of technically inconsistent projection methods.

To evaluate this hypothesis, data were collected through a structured questionnaire aligned with the theoretical framework and the problem statement. The instrument consists of closed-ended items measured on a five-point Likert scale (1 = Not at all, 5 = Completely) designed to capture attitudes, beliefs, and perceptions related to UVR projection practices. The questionnaire was administered in person to a sample of 100 individuals from Colombian financial institutions offering UVR-denominated mortgage loans. Participants included advisors, technical staff, and product managers. A non-probabilistic convenience sampling method was used, consistent with descriptive studies and with recommendations for minimum sample size in CFA models with up to five constructs and more than three indicators per construct.

The data set comprises responses to observable variables grouped into four constructs defined in Table 2. For modeling clarity, item labels were standardized as X1-X8 for exogenous indicators and Y1-Y7 for endogenous indicators, without modifying the wording or conceptual meaning of the variables. The analysis was performed using Confirmatory Factor Analysis (CFA), which allows validation of the proposed measurement structure by examining the relationships between items and latent constructs, correlations among constructs, and measurement errors.

The CFA results show that the model achieves satisfactory fit indices, supporting the internal consistency and theoretical coherence of the constructs. The findings confirm the research hypothesis: misconceptions persist due to identifiable and interrelated factors rather than isolated errors in judgment. The strongest effects appear in constructs related to organizational diffusion of heuristics and the internalization of simplified practices, followed by limitations in conceptual understanding of the UVR system. Correlations among constructs suggest that misconceptions emerge from the interaction between institutional norms, communication channels, and individual cognitive frameworks.