Explore the evidence.
Understand the reach.

In the context of rapid technological advancement and evolving financial ecosystems, this study investigates the interplay between financial literacy, digital finance, and financial behaviour among high school teachers in KwaZulu-Natal (KZN), South Africa. Structural equation modelling (SEM) approach was used to describe intricate relationships between latent constructs, namely, financial literacy, financial Behaviour and, digital finance Knowledge as well as observed variables (B7 to B22, C23 to C44, D45 to D56, demographics). Demographic variables include Gender, Age Category, Qualification Type, Teaching Duration, Financial Literacy Course and Field Employed. The model defined each construct as a linear combination of its observed variables as well as demographic variables under its measurement component, and hypothetical relationships among three latent variables under its structural component. We used Python software and machine learning code to create and estimate the model using observed data. The missing values in each observed variable were imputed with averages of the variables’ data. The tests of normality assumption for SEM were performed using histogram and q-q plots, Shapiro-Wilk test D'Agostino and Pearson's test.  Most data vectors violated the normality assumption. Diagonally Weighted Least Squares (DWLS) was used for parameter estimation as a suitable method that can provide better estimation for more complex models with many parameters. This method has an additional advantage of being less sensitive to outliers. The DWLS parameter estimation results, based on p-values, suggest that financial literacy and financial behaviour have a negative association which means that lacking financial behaviour results in financial literacy. A positive association between has been identified between financial behaviour and digital finance knowledge. Financial literacy and digital finance knowledge have a weak or no relationship. Thus, having or not having digital finance knowledge does not necessarily determine the possibility of experiencing financial literacy. Furthermore, among demographic variables, the financial literacy Course is the only demographic variable that has a significant impact, and only on Financial Behaviour and Financial literacy. The rest of the variables do not significantly influence the latent constructs. This study uniquely applies SEM using the DWLS estimator to explore personal finance dynamics in an emerging economy, offering concise insights into key relationships.

In the context of rapid technological advancement and evolving financial ecosystems, this study investigates the interplay between financial literacy, digital finance, and financial behaviour among high school teachers in KwaZulu-Natal (KZN), South Africa. Structural equation modelling (SEM) approach was used to describe intricate relationships between latent constructs, namely, financial literacy, financial Behaviour and, digital finance Knowledge as well as observed variables (B7 to B22, C23 to C44, D45 to D56, demographics). Demographic variables include Gender, Age Category, Qualification Type, Teaching Duration, Financial Literacy Course and Field Employed. The model defined each construct as a linear combination of its observed variables as well as demographic variables under its measurement component, and hypothetical relationships among three latent variables under its structural component. We used Python software and machine learning code to create and estimate the model using observed data. The missing values in each observed variable were imputed with averages of the variables’ data. The tests of normality assumption for SEM were performed using histogram and q-q plots, Shapiro-Wilk test D'Agostino and Pearson's test.  Most data vectors violated the normality assumption. Diagonally Weighted Least Squares (DWLS) was used for parameter estimation as a suitable method that can provide better estimation for more complex models with many parameters. This method has an additional advantage of being less sensitive to outliers. The DWLS parameter estimation results, based on p-values, suggest that financial literacy and financial behaviour have a negative association which means that lacking financial behaviour results in financial literacy. A positive association between has been identified between financial behaviour and digital finance knowledge. Financial literacy and digital finance knowledge have a weak or no relationship. Thus, having or not having digital finance knowledge does not necessarily determine the possibility of experiencing financial literacy. Furthermore, among demographic variables, the financial literacy Course is the only demographic variable that has a significant impact, and only on Financial Behaviour and Financial literacy. The rest of the variables do not significantly influence the latent constructs. This study uniquely applies SEM using the DWLS estimator to explore personal finance dynamics in an emerging economy, offering concise insights into key relationships.