Factors affecting personal financial management behavior evidence from vietnam

This study provides empirical evidence that the Covid-19 pandemic and many other factors such as financial attitudes, financial well-being, and financial socialization have an impact on

INTRODUCTION

Rationale of the study

The temporary subsiding of Covid-19 in Vietnam offers a valuable opportunity to reflect on how the pandemic drastically disrupted daily life In particular, the crisis has underscored the importance of sound financial management, which has long been a vital aspect of people's lives The Covid-19 pandemic serves as a powerful reminder of the need to adopt responsible financial behaviors to better withstand future uncertainties.

During the COVID-19 pandemic, many families faced financial exhaustion and helplessness, especially those with low income and limited assets, due to unforeseen emergencies A Pureprofile survey of 538 people in April 2020 revealed that 45% experienced income loss, and nearly 60% had to reduce daily spending, highlighting a lack of financial planning for future risks The pandemic has prompted a global awakening, encouraging individuals to prioritize savings and adjust their financial behaviors However, many people's financial stability declined as they struggled to adapt to government-mandated social distancing measures, exposing the urgent need for better financial management and emergency preparedness.

Many people's personal financial management practices are ineffective, leading to financial difficulties However, a resilient group has maintained their financial stability during the challenging epidemic period Even amidst social distancing measures, these individuals could still afford to spend, despite having average incomes before the COVID-19 pandemic Effective financial management is key to weathering economic uncertainties and maintaining financial well-being during crises.

Although numerous studies have explored factors influencing personal financial management behavior, there is a lack of research on the impact of the Covid-19 pandemic once it has subsided, particularly regarding its role in shaping perceptions and behavioral changes Updating and understanding these factors to reflect societal changes is crucial for economists and educators to develop effective policies and guidance Such efforts are essential for ensuring individual financial stability, which in turn supports overall national economic stability.

I have chosen the topic "Factors Affecting Personal Financial Management Behavior: Evidence from Vietnam" for my graduation thesis at Ho Chi Minh City University of Banking This research aims to explore the key factors that influence individual financial management behaviors within the Vietnamese context Understanding these factors can provide valuable insights for enhancing financial literacy and promoting better financial decision-making among consumers The study's findings will contribute to the development of strategies that improve personal financial management practices in Vietnam, supporting economic stability and individual financial well-being.

Research objectives

This thesis aims to identify the key factors influencing personal financial management behavior among Vietnamese individuals and assess the level of their impact The research objectives include exploring various determinants that affect financial habits and understanding their significance Additionally, the study seeks to provide practical solutions to enhance personal financial management practices in Vietnam By analyzing these factors, the thesis offers valuable insights for improving financial literacy and responsible money management among Vietnamese people.

Research questions

Through the objectives, the research question is posed as follows:

 Research question 1: Which factors affect personal financial management behavior?

 Research question 2: To what extent do these factors affect personal finanacial management behavior?

Scope of the study

Research subjects: People living and working in Ho Chi Minh City

Spatial scope: Ho Chi Minh City, Vietnam

Time range: Personal financial management behavioral metrics surveyed for the first 3 months of 2022

Contribution of the study

Empirical evidence demonstrates that the Covid-19 pandemic, along with factors like financial knowledge, financial attitude, financial well-being, and financial socialization, significantly influence personal financial management behavior These elements collectively shape individuals' ability to effectively handle their finances, especially during challenging times such as a global health crisis Understanding the impact of these variables can help in developing targeted financial education and socialization strategies to improve financial decision-making and resilience.

The study reveals that financial knowledge alone does not influence personal financial management behavior, highlighting a significant gap in current financial education This finding underscores the urgent need to改革金融教育 in schools to better equip students with practical financial skills and improve their financial decision-making.

Research method

Utilize descriptive statistics to analyze and summarize data into clear tables, highlighting the basic characteristics derived from empirical research Next, assess data reliability through Cronbach's Alpha, followed by Exploratory Factor Analysis (EFA) to identify the underlying components of each factor Use Pearson correlation to examine the linear relationships between dependent and independent variables Finally, apply the Ordinary Least Squares (OLS) multiple linear regression model to evaluate the influence of independent variables on the dependent variable, providing insights into their impact within the research framework.

Structure of the study

Chapter 3: Research methods and data

Chapter 4: Research results and discussion

This article examines the significant shifts in personal financial management behaviors during the COVID-19 pandemic, highlighting the urgent need to identify factors influencing these changes It emphasizes the importance of understanding how the pandemic affected individuals’ financial decision-making patterns, with a focus on assessing the impact of various influencing factors The research employs a quantitative methodology, analyzing data collected from a survey on personal finance management conducted in the first three months of 2022, to determine the key factors shaping financial behavior post-COVID-19 This study aims to provide insights into the determinants of personal financial management behavior during a period of global economic disruption, informing future financial education and policy strategies.

The article highlights the importance of the topic's theoretical and practical research relevance, emphasizing its significance for advancing knowledge and real-world applications Additionally, it underscores the well-structured design and comprehensive content of each chapter, demonstrating a clear and cohesive approach to the study subject.

LITERATURE REVIEW

Theoretical framework

Personal financial management involves overseeing cash flow, debt, savings, and spending to ensure responsible financial behavior (Hilgert et al., 2003) It encompasses the identification, acquisition, allocation, and utilization of financial resources, reflecting an individual's ability to manage their finances effectively (Horne et al., 2002) Personal financial management behavior is driven by a desire to meet needs in accordance with income levels and includes daily activities like planning, budgeting, monitoring, controlling, and storing financial funds (Kholilah & Iramani, 2013) Developing strong personal financial management skills is crucial for fostering responsible individuals capable of effectively managing their money and assets.

Personal financial management behavior encompasses how individuals respond to financial events by acquiring, allocating, deciding, and using financial resources to achieve specific objectives Effective financial decisions significantly impact both personal well-being and the broader economy, influencing present and future financial stability Poor financial management can lead to undesirable short, medium, and long-term consequences that affect individuals, households, and society as a whole (Fenton et al., 2016) Empirical studies show that sound financial management enhances long-term financial status and satisfaction (The Consumer Financial Protection Bureau, 2015) However, financial management behavior is complex and challenging to implement, highlighting the importance of monitoring spending carefully and practicing frugality as safeguards against risky financial activities.

2.1.2 Factors affecting personal financial management behavior

Based on the research model and experimental results of related previous studies: Joo & Grable (2004), Shim et al (2009), Xiao et al (2009), Ida & Dwinta

(2010), Xiao & Dew (2011), Woodyard & Robb (2012), Mien & Thao (2015), Arifin (2017), Ameliawati & Setiyani (2018), Abel et al (2018), Prihartono et al

Research by 2018, Marko (2020), Gayan (2020), and Le Thanh Tam et al (2021) identifies five key factors influencing the personal financial management behavior of Vietnamese individuals These factors became particularly significant following the severe impact of social distancing measures caused by the Covid-19 pandemic The study highlights how the pandemic has shaped financial decision-making and spending habits among Vietnamese people during this challenging period.

Financial knowledge, as defined by Garman et al (2006), encompasses a comprehensive understanding of personal finance facts and is essential for sound financial management Vitt et al (2005) highlight that financial knowledge involves the ability to understand and effectively apply financial management skills such as planning, debt management, interest rate calculations, and understanding compound interest Financially literate individuals are better equipped to recognize and manage aspects related to money and wealth (Swart, 2012), leading to informed financial decisions and responsible behaviors Responsible financial behavior includes actions like paying credit card debt on time and prioritizing needs over desires, which can be fostered through financial education to ensure long-term financial health.

Personal financial knowledge is crucial for individual financial stability and prosperity, as well as for the overall economic development In emerging economies, financially literate citizens help enhance the effectiveness of the financial sector in driving real economic growth and reducing poverty (Faboyede et al., 2015) Vietnam, as an emerging and developing country, benefits from improving financial literacy to support sustainable economic progress and social welfare.

Vietnamese financial literacy remains very low, especially among students who are the future generation of the country Research by Lusardi & Mitchell (2011) indicates that middle-aged individuals tend to have higher financial literacy compared to younger and older groups Although financial literacy generally increases with age, the pattern is not entirely consistent, with those aged 23–29 and 40 and above demonstrating greater financial knowledge than other age groups, as found by Chen & Volpe (1998).

Numerous studies highlight the significance of gender in financial management, indicating that men often outperform women in this area Research by Kharchenko & Olga (2011), Al-Tamimi & Hussain (2009), Arrondel et al (2013), and Koenen & Lusardi (2011) suggests that men tend to be more skilled at managing finances than women Additionally, studies such as Chen & Volpe (2002), Eitel & Martin (2009), Goldsmith (2006), and Hira & Mugenda (2000) confirm that male students generally demonstrate greater financial savviness compared to their female counterparts These findings emphasize the ongoing gender gap in financial literacy and management skills.

Research indicates that both individual traits and family environment significantly influence students' financial knowledge Factors such as parental education, employment status, and family economic status play a crucial role in shaping financial understanding (Mohamad, 2010) Additionally, Murphy (2005) found that adolescents from educated families tend to have higher financial literacy, and ongoing discussions about money with parents further enhance their financial comprehension.

According to several earlier studies (Chen & Volpe (1998), (2002), Sabri et al

Academic competence significantly influences financial knowledge, as higher academic achievement correlates with improved understanding of finance (Shim et al., 2009; 2010) Jariah et al (2004) found a strong positive relationship between GPA and financial literacy, indicating that students with higher GPAs tend to possess better financial knowledge Furthermore, research shows that students with high GPAs are more likely to learn about finances from peers Additionally, the duration of university study impacts financial perception, with graduates exhibiting greater financial knowledge than current students, and senior students demonstrating higher financial awareness compared to freshmen.

Many factors affect an individual's financial knowledge, and this phrase also can be understood by many different definitions, depending on the individual

Financial literacy is crucial as it influences nearly every aspect of individuals' lives Multiple studies demonstrate that possessing financial knowledge promotes responsible financial behaviors, highlighting a strong link between financial literacy and effective personal financial management (Robb and Woodyard, 2011; Zakaria et al., 2012) Consumers with a basic understanding of finance are more likely to make responsible financial decisions, underscoring the importance of financial education (Hogarth and Hilgert, 2002).

A financial attitude reflects an individual's mindset, encompassing their views, judgments, and thoughts about money management It influences how people perceive and approach their finances, shaping their behaviors and decision-making processes Understanding one’s financial attitude is essential for developing sound financial habits and improving overall financial health.

An excellent financial attitude significantly influences how individuals perceive and handle money, shaped by personal, cultural, and moral values related to financial decisions and products These attitudes guide behaviors such as spending, saving, storing, and wasting money (Furnham, 1984) They also impact personal financial management, including budgeting and investment choices, with Ajzen (1991) noting that financial attitudes result from an individual's behavioral tendencies and can be influenced by both economic and non-economic beliefs According to Borden et al (2008), a financial attitude reflects a person's way of thinking and evaluating finances However, certain financial attitudes can contribute to financial difficulties, especially among young people (Lim, 1997; Madern, 2012).

Financial knowledge is a crucial component influencing financial attitudes, as there is a well-established connection between an individual's financial attitude and their level of financial literacy (Grable & Lytton, 1998; Kasman, Heuberger, & Hammond, 2018) Among young people, a positive perception of money and finance can enhance their financial behaviors and motivation to learn about financial management Conversely, a negative attitude toward money can impair their ability to make sound financial decisions, potentially leading to poor financial outcomes (Shim, Xiao, Barber, & Lyons, 2009; Sohn, Joo) Developing a strong financial literacy foundation is essential for fostering healthy financial attitudes and responsible money management.

Research by Grable, Lee, and Kim (2012) highlights the importance of financial literacy among students Soroshian and Teck (2014) found that students' financial attitudes significantly influence their financial literacy levels Additionally, Ibrahim and Alqaydi (2013) emphasize that education plays a crucial role in improving personal financial attitudes, leading to better financial decision-making skills.

A person's financial attitude significantly influences their financial behavior, with numerous studies confirming this positive relationship Ameliawati & Setiyani (2018) found that a positive financial attitude enhances personal financial management practices, a finding supported by research from Davis and Schumm (1987), Shih and Ke (2014), Amanah et al (2016), Mien & Thao (2015), and Herdjiono & Darmanik (2016) However, Maharani (2016) suggests that there is no direct link between financial attitude and management behavior, as some individuals believe that having a positive attitude does not necessarily lead to effective financial planning Furthermore, Amanah, Rahadian, and Iradianty (2016) noted that financial attitude only slightly impacts financial behavior, indicating that other factors may also play a crucial role.

Based on a personal assessment of one's financial circumstances, financial well-being is the condition of being financially sound, content, and worry-free ( Joo,

Empirical review

Generally, there are relatively few studies related to personal financial management behavior in Vietnam Two related research works the author found are as follows:

Le Thanh Tam et al (2021) identified four key factors influencing personal financial management behaviors during the Covid-19 pandemic: the pandemic itself, financial habits inherited from parents, financial education provided by parents, and overall financial well-being However, since their study was conducted in early 2021, prior to the full outbreak of Covid-19 in Vietnam, particularly in Ho Chi Minh City, its reliability in assessing pandemic-related changes in financial behaviors is limited.

Nguyen Thi Ngoc Mien and Tran Phuong Thao (2015) conducted a comprehensive study on the factors influencing personal financial management behaviors in Vietnam Their research identified four key factors—personal financial attitude, financial knowledge, locus of control, and financial management behaviors—that significantly impact individuals' financial practices This well-researched work has garnered high scientific value and is widely cited internationally However, since it was published in 2015, it does not account for the financial behavior changes brought about by the COVID-19 pandemic.

 Mieske Anggraini Halim and Ignatius Roni Setyawan (2021),

―Determinant Factors of Financial Management Behavior Among People in Jakarta During Covid-19 Pandemic‖ The authors analyzed the influence of financial knowledge, financial attitude and financial literacy

This study examines financial management behavior in Jakarta during the Covid-19 pandemic, a period marked by government restrictions that significantly reduced commercial activities and decreased residents' purchasing power The research specifically focuses on the timeframe in which data was collected, without considering Covid-19 as a direct variable influencing personal financial management behaviors Understanding how individuals adapt their financial strategies during such economic disruptions offers valuable insights into their financial resilience amidst crisis.

Gayan Abeyrathna (2020) highlights that financial attitude, financial knowledge, and locus of control significantly influence the personal financial management behaviors of Sri Lankan government employees However, the study does not address the impact of the Covid-19 pandemic on these behaviors Additionally, the research's observational nature and small sample size may limit the generalizability of its findings.

 Abel Tasman, Deny Ari Efendi, Erni Masdupi (2018), ―Analysis Of Personal Financial Management Behavior In Higher Education Student‖

This research identifies key elements influencing personal financial management behavior, highlighting the significant impact of financial knowledge on students' financial conduct Notably, the study reveals that financial knowledge can have a detrimental effect on students' financial management It is important to note that the primary focus was on students, and the research was conducted prior to the COVID-19 pandemic, which may affect the current relevance of the findings.

Prihartono and Asandimitra (2018) examined the factors that influence financial management behavior, highlighting the roles of income, higher education, financial knowledge, financial literacy, attitudes toward finance, and locus of control Their study focuses specifically on economics students, providing insights into how these variables impact their financial decision-making However, the research is limited in scope as it concentrates solely on students, which may affect the generalizability of the findings to broader populations.

 Ameliawati M & Setiyani R (2018), ―The influence of Financial Attitude, Financial Socialization, and Financial Experience to Financial

This study explores how management behavior is influenced by financial literacy as a mediating variable, examining the roles of financial attitude, socialization, and experience in shaping financial management habits By focusing on financial literacy, the research aims to identify key factors that drive responsible financial behaviors among students Although innovative methods have been introduced to analyze these relationships, significant advancements in understanding the determinants of financial management behavior remain limited The primary objective is to determine the variables that impact financial management practices within the student demographic, providing valuable insights for financial education and behavioral interventions.

This chapter provides a clear discussion of the theoretical foundations of personal financial management behavior It analyzes how factors like the Covid-19 pandemic, financial knowledge, financial attitude, financial well-being, and financial socialization influence individuals' financial behaviors The chapter also reviews previous research, including two domestic and five international studies, highlighting their methodologies and key subjects These insights establish a solid theoretical basis for developing the subsequent research model.

RESEARCH METHODS AND DATA

Research model and research hypotheses

Based on previous research, the proposed model identifies five key factors influencing personal financial management behavior: financial knowledge, financial attitude, financial well-being, financial socialization, and the impact of the Covid-19 pandemic The hypothesis suggests that each of these factors positively affects individuals' financial management practices Understanding these elements can help tailor more effective financial education and intervention strategies, especially in the context of ongoing global challenges like the Covid-19 pandemic Ultimately, fostering financial knowledge, fostering positive attitudes, enhancing financial well-being, leveraging social influences, and considering pandemic-related impacts are crucial for improving personal financial behavior.

(Source: Compiled by the author)

Research data

Primary data was collected online through questionnaires for 560 individuals currently residing in Ho Chi Minh City with random sampling method, of which

A total of 557 participants completed the questionnaire, with responses measured on a 5-point Likert Scale ranging from strongly disagree (1) to strongly agree (5) The study assessed key variables including financial knowledge, financial attitude, and financial behavior, which collectively represent financial well-being Additionally, financial education was examined as a component of financial socialization, alongside the impact of Covid-19 and personal financial management These insights aim to understand the factors influencing financial literacy and behavior amid the pandemic.

20 management (G) In which, Personal financial management behavior (G) is the dependent variable, the remaining variables are independent variables

The questionnaire comprised two sections: the first focused on respondents' demographics, while the second gathered data on the variables of interest Each variable was assessed through specific questions, which served as observed indicators to accurately measure the underlying factors This structured approach ensures comprehensive data collection for effective analysis of the study's key variables.

Research methods

The initial step in data analysis involves using Descriptive Statistics to create summary tables that highlight the basic characteristics of the collected data from empirical research Subsequently, the reliability of the data is assessed with Cronbach's Alpha, followed by Exploratory Factor Analysis (EFA) to identify the key components of each factor and the main influences on personal financial management behavior Pearson correlation analysis is then employed to examine the linear relationships between dependent and independent variables Finally, the impact of independent variables on the dependent variable is evaluated through OLS linear regression modeling All data analysis processes are conducted using SPSS.

The followings are to clarify each step:

Cronbach's Alpha coefficient is a vital tool for assessing the reliability of a scale, indicating how well variables within the same group of factors are related It helps identify and eliminate unsatisfactory observed variables or scales that may produce pseudo-factors, thereby ensuring the scale's validity A higher Cronbach's Alpha value, generally 0.7 or above as recommended by Nunnally (1978) and Hair et al (2009), signifies greater internal consistency reliability, with values of 0.6 being acceptable for early exploratory research Ensuring unidirectionality and reliability, a robust scale typically achieves a Cronbach's Alpha of 0.7 or higher.

Corrected Item - Total Correlation is another crucial metric This value represents the correlation between each observed variable with the rest of the

A scale with 21 variables benefits from stronger positive correlations between observed variables, as indicated by higher Corrected Item-Total Correlation values, which reflect the quality of each item According to Cristobal et al (2007), a reliable scale is characterized by observed variables with Corrected Item-Total Correlation values of 0.3 or higher Items with a Corrected Item-Total Correlation below 0.3 should be considered for removal to improve the overall reliability, as higher correlation coefficients denote better observed variable quality and contribute to more accurate scale measurement.

When evaluating the reliability of a variable using Cronbach's Alpha, it is important to analyze the "Cronbach's Alpha if Item Deleted" values, which indicate how removing a specific item would affect the overall scale's reliability If this value is higher than the group's main Cronbach's Alpha, the quality of that variable may be considered better with its removal; however, this alone should not determine exclusion Since there is no individual Cronbach's Alpha for each variable, decisions should also consider the variable’s contribution to the factor, especially if the "Cronbach's Alpha if Item Deleted" difference is less than 0.1 from the overall alpha and the Corrected Item-Total Correlation exceeds 0.3, suggesting the variable’s importance in discriminant or convergent validity during Exploratory Factor Analysis (EFA).

Exploratory Factor Analysis (EFA) is a statistical technique used to reduce a large set of observed variables into a smaller number of meaningful factors, enhancing data interpretability Often applied in research with numerous correlated variables, EFA identifies key underlying features, allowing researchers to simplify complex data For example, instead of analyzing 20 minor characteristics, EFA can highlight 4 major features that capture the essential information of an object, streamlining analysis and decision-making This method helps in uncovering the underlying structure of data, making it easier to understand relationships among variables and improve the effectiveness of subsequent analyses.

22 smaller features that are associated with one another For researchers, this means additional time and money savings

Cronbach's Alpha reliability testing evaluates the internal consistency of variables within the same group or factor but does not consider relationships across different factors In contrast, Exploratory Factor Analysis (EFA) assesses the relationships among variables across all groups or factors, helping to identify variables that load on multiple factors or may be misclassified initially EFA is an interdependent multivariate analysis method that does not distinguish between dependent and independent variables To perform EFA successfully, certain conditions must be met, including sufficient sample size, shared underlying structure among variables, and appropriateness of data for factor analysis.

The Kaiser-Meyer-Olkin (KMO) coefficient is a key measure used to determine the suitability of data for factor analysis A KMO value of 0.5 or higher (0.5 ≤ KMO ≤ 1) indicates that factor analysis is appropriate for the dataset Conversely, if the KMO score is below 0.5, it suggests that factor analysis may not be suitable for the research data Therefore, ensuring a KMO value of at least 0.5 is essential for reliable and valid factor analysis results.

Bartlett's test of sphericity assesses whether the observed variables within a factor are correlated, which is essential for applying factor analysis A significant result (sig Bartlett's Test < 0.05) indicates that the variables are sufficiently correlated, confirming the suitability of factor analysis Conversely, a non-significant result suggests that factor analysis should not be performed, as the variables do not exhibit the necessary intercorrelations Therefore, obtaining a significant Bartlett's test ensures the observed variables reflect underlying factors effectively.

 Eigenvalue is a commonly used criterion to determine the number of factors in EFA analysis With this criterion, only factors with Eigenvalue

≥ 1 are kept in the analytical model

 Total Variance Explained ≥ 50% indicates that the EFA model is appropriate Considering the variation as 100%, this value shows how

23 much the extracted factors are condensed and how much percentage of the observed variables is lost

Factor loading indicates the correlation between observed variables and underlying factors, with higher values signifying stronger relationships According to Hair et al (2010) in *Multivariate Data Analysis*, a loading of 0.5 or above is considered a good quality indicator for an observed variable, while the minimum acceptable threshold is typically set at 0.3.

 Factor Loading at ± 0.3: Minimum condition for the observed variable to be retained

 Factor Loading at ± 0.5: The observed variable has good statistical significance

 Factor Loading at ± 0.7: The observed variable has very good statistical significance

Before conducting the regression analysis, Pearson correlation testing was performed to assess the linear relationships between the dependent and independent variables This step helps identify multicollinearity issues, especially when independent variables are highly correlated with each other, which can compromise the reliability of statistical inferences Detecting strong correlations ensures the robustness and accuracy of the regression model.

Understanding the types of correlation is essential in research: one type involves the relationship between dependent and independent variables, while the other pertains to correlations among independent variables When building a research model, it's important to carefully examine how independent variables relate to the dependent variable to identify which factors significantly influence the outcome These independent variables are selected based on strong theoretical foundations, prior research, and an assessment of the real-world survey environment As a result, the data analysis is expected to reveal meaningful insights into these relationships, strengthening the validity of the research findings.

The Pearson correlation analysis demonstrates that the independent variables are significantly related to the dependent variable, indicating a meaningful association These findings suggest that the independent variables have a considerable impact on the dependent variable, reinforcing their importance in the regression analysis Prior correlation results highlight the predictive relationship, aiding in understanding how changes in independent variables influence the dependent variable within the regression model.

High correlation between independent variables indicates they may essentially represent the same concept or measure, which can lead to multicollinearity issues in regression analysis Strongly associated independent variables can bias statistical results and reduce model reliability To ensure accurate insights, it's important to identify and address multicollinearity when variables are highly correlated Generally, weak associations between independent variables improve the robustness and interpretability of regression models.

Pearson correlation coefficient (r) fluctuates in the continuous range from -1 to +1 It only has any significance when the observed significance level (sig.) is less than the significance level α = 5%

 If r is between 0.50 and ± 1, it is said to be strongly correlated

 If r is between 0.30 and ± 0.49, it is said to be mean correlation

 If r is below ± 0.29, it is said to be weak correlation

Multiple regression analysis differs from Pearson correlation by not involving symmetric variables and focuses on predicting one dependent variable based on two or more independent variables It extends simple linear regression to assess how multiple factors simultaneously influence the outcome, allowing researchers to evaluate the degree of influence—whether many, few, or none—that each predictor has on changes in the dependent variable To accurately determine the impact of each factor, it is essential to consider specific statistical values that quantify their effect sizes within the model.

R-squared (R²) and Adjusted R-squared are key indicators in regression analysis that measure the explanatory power of independent variables on the dependent variable While R² indicates the proportion of variance explained by the model, Adjusted R² provides a more accurate reflection by accounting for the number of predictors, making it a preferable metric for assessing model fit Both values range from 0 to 1, where higher values signify better explanatory capability of the regression model.

RESEARCH RESULTS AND DISCUSSION

Descriptive statistics

The survey included 557 participants, predominantly young individuals aged 17-25, who made up 86.9% of the sample Females comprised 69.8% (389 women), while males accounted for 30.2% (168 men) Participants aged 26-30 represented 7.7%, those aged 31-40 accounted for 4.1%, and individuals over 40 constituted only 1.3% of the respondents.

Figure 4.1 Age groups of survey participants

(Source: Statistics from the author's survey)

Survey participants have experience across diverse fields including business, accounting, finance and banking, medicine, tourism, education, human resources, hospitality, logistics, computer engineering, law, chemicals, and design The majority of these participants are single, accounting for 64.8%, highlighting the demographic trend within this diverse professional group.

Figure 4.2 Marital status of survey participants

(Source: Statistics from the author's survey)

Most respondents in Ho Chi Minh City live with their parents or relatives, with a smaller proportion residing in rented houses rather than owning their own homes or apartments Approximately 70% of these individuals rely on family support for daily living costs, education, employment, and entertainment expenses This reliance is common among the 17–25 age group, many of whom are students or part-time workers still immersed in education or early career stages, limiting their financial independence.

Table 4.1 Residence of survey participants residence

(Source: Results extracted from Eview software)

Single In relationship Married Divorced

This study highlights the link between parental educational attainment and children's financial literacy Despite the global 4.0 era, in Ho Chi Minh City, men from the Baby Boomers and Gen X generations generally have higher education levels than women Specifically, 33.4% of men hold a university degree compared to only 23% of women Additionally, over half of the respondents' mothers did not complete middle or high school, indicating a significant gender gap in parental education levels that may influence financial literacy education.

Table 4.2 Education level of surveyor's father

Under elementary and elementary school 42 7.5 7.5 7.5

Table 4.3 Education level of surveyor's mother

Under elementary and elementary school 57 10.2 10.2 10.2

Checking the reliability of the scale

A survey named Group B assessed six questions related to financial knowledge, and the internal consistency of these variables was evaluated using Cronbach’s alpha The results indicate that the Cronbach’s alpha value for scale B is 0.746, exceeding the acceptable threshold of 0.6, demonstrating good reliability However, item B6 has a low corrected item-total correlation of 0.054, which is below the recommended minimum of 0.3 Notably, the Cronbach’s alpha coefficient for scale B6 is 0.853, higher than the overall scale, suggesting B6 may not contribute meaningfully to the factor Consequently, B6 was removed to improve the scale’s reliability, and Cronbach’s alpha was re-analyzed to ensure the scale’s internal consistency. -**Sponsor**Looking to optimize your financial knowledge article for SEO? We can help refine it for clarity and impact! Your survey analysis reveals interesting insights; to highlight key findings, consider emphasizing that the Cronbach's Alpha for group B is 0.746, indicating good reliability, but removing variable B6 significantly improves the scale's overall consistency to 0.853 because B6 has low explanatory power For more investment insights and portfolio diversification strategies, explore [Freedom24 ETF - English](https://pollinations.ai/redirect-nexad/uA0VZUZa?user_id=983577) and discover how ETFs can help you achieve your financial goals with transparent fees and expert analysis Refocusing your article on these core points will enhance readability and align with SEO best practices.

Cronbach's Alpha Cronbach's Alpha Based on Standardized Items

Scale Variance if Item Deleted

Cronbach's Alpha if Item Deleted

Excluding variable B6 improved the reliability of factor B, as evidenced by a higher Cronbach's Alpha coefficient Additionally, all indicators in the Corrected Item-Total Correlation column exceed 0.3, indicating strong internal consistency Consequently, factor B and its associated observed variables are now well-explained by the scale, ensuring more accurate and reliable measurement.

Table 4.7 Item – Total Statistics – B (II)

In Group C, similar procedures were applied, focusing on five questions related to financial attitude, which served as observed variables The analysis showed a Cronbach's Alpha of 0.673, indicating acceptable internal consistency However, observed variable C1 had a Corrected Item-Total Correlation of -0.228, suggesting it contributed minimally to the factor and was therefore removed from the scale After removing C1, Cronbach's Alpha was reassessed to ensure the reliability of the remaining items.

The following is the result table after removing C1 and leaving group C with 4 observed variables:

Table 4.11 Item – Total Statistics – C (II)

The Cronbach's Alpha coefficient for Scale C is 0.869, indicating excellent reliability and surpassing the acceptable threshold of 0.6, which reflects improved consistency compared to the initial analysis All observed variables demonstrate Corrected Item-Total Correlation values above 0.3, confirming their strong contribution to the scale Consequently, the results establish that Scale C is dependable and that component C is comprehensively explained by its observable factors.

The variables in group D focus on survey questions concerning financial behavior The Cronbach's Alpha test results show that the overall coefficient for group D is 0.507, which is below the acceptable threshold of 0.6, indicating moderate internal consistency Notably, the observed variable D13 has its own Cronbach's Alpha coefficient, further highlighting areas for potential improvement in the reliability of this survey subset.

Alpha if Item Deleted is greater than 0.6 Take out the observed variable D13 and do a second Cronbach's Alpha analysis

And here is the result after removing D13:

Table 4.15 Item – Total Statistics – D (II)

The Cronbach's Alpha coefficient for Group D at the second measurement is equal to the Cronbach's Alpha if Item Deleted for variable D13, indicating an improved reliability when this item is removed The scale demonstrates acceptable reliability, surpassing the threshold of 0.6 Additionally, all observed variables exhibit a Corrected Item-Total Correlation greater than 0.3, further supporting the scale's internal consistency and reliability.

Test for reliability of Group E by using Cronbach's Alpha The Corrected Item

The Total Correlation index for each of the six observed variables related to financial socialization exceeds the established threshold, indicating that they are strong and reliable variables As a result, these variables are classified as good indicators, eliminating the need to create dummy factors for the outcome variable Consequently, group E does not require the removal of any undesirable variables, ensuring the integrity and relevance of the data for accurate analysis.

F represents a set of variables related to questions addressing the Covid-19 pandemic, comprising six observed variables Reliability testing using Cronbach's Alpha revealed that F1 poses a risk of generating dummy factors when evaluated with the Corrected Item-Total Correlation.

The total correlation is 0.298, which is below the 0.3 threshold, indicating a weak relationship with the overall scale Additionally, the Cronbach’s Alpha if Item Deleted for variable F1 is 0.783, surpassing the Cronbach’s Alpha of the F group at 0.762 Therefore, F1 will be removed from the scale due to its minimal contribution to explaining the F factor, improving the internal consistency and reliability of the measurement.

After removing the observed variable F1, the reliability of the variables in the F factor group was tested, and the following result was obtained:

Table 4.21 Item – Total Statistics – F (II)

The Cronbach's Alpha coefficient for the F scale at the second measurement aligns with the results from the second Cronbach's Alpha analysis in the D factor group, reaching a value of 0.783, which exceeds the acceptable threshold of 0.6 Notably, the Cronbach's Alpha if Item Deleted for variable F1 is identical to the overall Cronbach's Alpha, indicating that removing F1 does not improve internal consistency Additionally, all observed variables demonstrate Corrected Item-Total Correlations greater than 0.3, confirming their reliability and contribution to the scale's coherence.

To conduct multiple regression analysis, it is essential to develop a scale with questions that accurately measure the dependent variable, ensuring data is available for both independent and dependent factors Without a dedicated dependent variable scale, researchers can perform reliability tests such as Cronbach's Alpha and Exploratory Factor Analysis (EFA), but cannot proceed with regression analysis to determine regression coefficients Therefore, after collecting survey responses with four questions related to the dependent variable G, it is important to assess the reliability of these items to ensure valid and reliable measurement for subsequent analysis.

The test results indicate that the Cronbach's Alpha coefficient for G is 0.725, surpassing the acceptable threshold of 0.6, which confirms the scale's reliability Additionally, all observed variables demonstrate Corrected Item-Total Correlation values greater than 0.3, highlighting their strong explanatory power Overall, these findings validate the consistency and validity of the measurement scale.

Using Cronbach's Alpha coefficient to assess scale reliability showed improvement after removing certain poorly performing variables Despite differing numbers of observed variables across groups, all factors achieved high Cronbach's Alpha values, indicating strong internal consistency.

Table 4.24 Checking the reliability of the scale by Cronbach's Alpha coefficient

Factor Number of observed variables

Exploratory factor analysis

Performing EFA analysis for this study, we have the following results:

Table 4.25 KMO and Bartlett test results (I)

Kaiser-Meyer-Olkin Measure of Sampling Adequacy .900

Initial Eigenvalues Extraction Sums of Squared

Extraction Method: Principal Component Analysis

Table 4.27 Results of rotated component matrix (I)

Rotation Method: Varimax with Kaiser Normalization a Rotation converged in 6 iterations

The initial EFA results indicate that factor analysis is suitable, supported by a KMO measure of 0.900 exceeding the minimum threshold of 0.5 and a significant Bartlett's Test (p = 0.000) Five components with eigenvalues greater than 1 were identified, accounting for a cumulative variance of 61.787% To ensure the selection of high-quality observed variables, a load factor threshold of 0.5 will be used instead of relying solely on sample size criteria.

41 variable, D12, that needs to be deleted when comparing this threshold to the results in the rotation matrix since D12 does not have a load factor in all factors

From 24 observed variables in the first EFA analysis, remove the D12 variable and include the remaining 23 observed variables in the second EFA analysis

Table 4.28 KMO and Bartlett test results (II)

Kaiser – Meyer – Olkin Measure of Sampling Adequacy .894 Bartlett's Test of Sphericity Approx Chi-Square 6009.592

KMO = 0.894 > 0.5, sig Barlett’s Test = 0.000 < 0.05 Factor analysis is appropriate

Table 4.29 Total Variance Explained (II)

Extraction Method: Principal Component Analysis.

Based on eigenvalues greater than 1, five key factors were identified, effectively summarizing the information from 23 observed variables in the exploratory factor analysis (EFA) These five factors explain a total variance of 62.66%, exceeding the 50% threshold, indicating a strong underlying structure within the data.

5 extracted factors explain 62.657% of the data variation of 23 observed variables participating in EFA

Table 4.30 Results of rotated component matrix (II)

Rotation Method: Varimax with Kaiser Normalization a Rotation converged in 6 iterations.

The second rotation matrix analysis reveals that all 23 observed variables are clearly grouped into 5 distinct factors Each variable exhibits factor loading coefficients greater than 0.5, indicating strong associations with their respective factors Additionally, the absence of low-loading variables suggests a well-defined and reliable factor structure, enhancing the overall validity of the measurement model.

An exploratory factor analysis (EFA) was conducted twice on the independent variables to ensure validity During the process, the observed variable D12 was excluded from the analysis because it did not meet the necessary criteria in the first round, similar to the previous case with variable 24.

44 observed variables In the second study, five distinct factors were created from 23 observed variables, including:

After analyzing the EFA rotation matrix, it was observed that the variables grouped into different factors compared to their original survey order, highlighting the need to reassign and rename the factor groups accordingly Proper renaming of these factors is essential to ensure accurate interpretation before proceeding to the next research stage.

Table 4 reveals that Group 1 consists of observed variables related to two key factors: Financial Attitude (C) and Financial Well-being (D) Based on the survey content, the authors designated Group 1 as Financial Well-being (D) Groups 2 and 3 correspond to Financial Socialization (E) and Financial Knowledge (B), respectively, with no changes made to these groups in the results Separately from factor F (Covid-19 Pandemic), two additional groups, 4 and 5, were identified; Group 4 has been renamed as Financial Attitude (C) after analyzing shared traits within these observed variables.

Table 4.31 Names of factors after EFA analysis

Component Name of factor Signal

(Source: Compile by the author)

While renaming factors may not strictly adhere to the principle that groups should be named based on the majority of observed variables, it is reasonable to do so by considering the content, meaning, general characteristics, and relationships of the variables within each group This approach ensures meaningful and accurate classification, aligning the factor names with their underlying content for clearer interpretation and analysis.

In this study, the dependent variable, G - Personal Financial Management Behavior, was analyzed using Exploratory Factor Analysis (EFA), similar to the approach used for independent variables The G variable comprises four observed indicators: G1, G2, G3, and G4 The results indicate that these observed variables collectively measure the underlying construct of personal financial management behavior effectively, demonstrating adequate factor loadings and internal consistency in the EFA process.

Table 4.32 KMO and Bartlett test results

Kaiser-Meyer-Olkin Measure of Sampling Adequacy .706

KMO = 0.706 > 0.5, sig Bartlett’s Test = 0.000 < 0.05 Factor analysis is appropriate

Extraction Method: Principal Component Analysis a 1 components extracted

The analysis reveals a significant factor with an eigenvalue of 2.205, exceeding the cutoff of 1, indicating its importance in explaining data variability This factor accounts for 55.126% of the total variation among the four observed variables in the exploratory factor analysis (EFA) Additionally, all variables demonstrate strong loadings above 0.5, confirming their reliability and good quality within the factor structure.

Correlation analysis

We got the following findings from this study's Pearson correlation analysis:

** Correlation is significant at the 0.01 level (2-tailed)

The significance value in the correlation analysis is less than 0.05, indicating a statistically significant linear relationship between independent variables and the dependent variable Additionally, there is no strong correlation among the independent variables, as the absolute value of the correlation coefficients between variable pairs is below 0.5, suggesting a lower risk of multicollinearity.

Regression model

Performing multiple linear regression analysis to evaluate the impact of these independent variables on the dependent variable for this study, the results are shown as follows:

Std Error of the Estimate 48822

Durbin-Watson 1.968 a Predictors: (Constant), F_AVERAGE, B_AVERAGE, E_AVERAGE, D_AVERAGE, C_AVERAGE b Dependent Variable: G_AVERAGE

The coefficient of determination, R², is a key metric for assessing the fit of a linear regression model A high R² indicates that most data points are closely clustered around the regression line, reflecting a strong model fit Conversely, a low R² suggests that data points are widely scattered away from the line, indicating a weaker fit and less predictive power Understanding R² helps in evaluating the accuracy and reliability of linear regression models in data analysis.

The Model Summary table presents R² (R Square) and adjusted R² (Adjusted R Square) to assess the model's goodness of fit An adjusted R² value of 0.376 indicates that the independent variables in the regression analysis explain 37.6% of the variation in the dependent variable, while the remaining 62.4% is attributed to outside variables and random errors.

The table's results also provide Durbin-Watson values for evaluating the phenomena of first-order series autocorrelation Because the DW value = 1.968 is

49 between 1.5 and 2.5, the results do not violate the premise of first-order series autocorrelation (Yahua Qiao, 2011)

Model Sum of Squares df Mean Square F Sig

Total 212.281 556 a Dependent Variable: G_AVERAGE b Predictors: (Constant), F_AVERAGE, B_AVERAGE, E_AVERAGE,

After developing a multivariable regression model, the first consideration must be the model's fit to the data set as measured by the value of Adjusted R Square (or

This study utilizes a linear regression model to analyze sample data, with the R-squared indicating the model’s fit within the research sample Since surveying the entire population is impractical due to its large size, a small, representative sample is selected to infer the population's basic features The ANOVA table’s F-test assesses the model's overall significance, determining whether the regression model is appropriate and generalizable to the broader population.

In this study, the sig value for the F test is 0.000 < 0.05 As a result, the linear regression model developed is appropriate for the population

The t-test is used to assess the significance of each independent variable's regression coefficient in the model There are two types of regression coefficients: unnormalized (B in SPSS) and normalized (Beta in SPSS), each serving a distinct purpose in understanding the model's implications Since unnormalized coefficients are affected by differences in measurement units and standard deviations, they cannot be directly compared across variables Therefore, the study focuses on normalized coefficients (Beta) to evaluate the relative importance of independent variables within a consistent framework, ensuring accurate interpretation of their significance.

A negative regression coefficient (B or Beta) indicates that the independent variable has a negative effect on the dependent variable Conversely, a positive coefficient signifies a positive relationship between the independent and dependent variables The magnitude of the impact is assessed using the absolute value of Beta, with larger values indicating a stronger influence In SPSS, the significance of these effects is determined through the t-test results found in the Coefficients table.

The Coefficients table displays the t-test results to assess the significance of the regression coefficient's hypothesis, while the VIF index assesses multicollinearity and the regression coefficients

Variables D (Financial well-being), E (Financial socialization), C (Financial attitude), and F (COVID-19 pandemic) all have significance values below 0.05, confirming their critical role as explanatory factors in the model In contrast, Variable B (Financial knowledge) has a significance value greater than 0.05, indicating it does not significantly influence the model's outcomes.

The study highlights the significant role of intrinsic power in shaping personal financial management behavior Additionally, the Variance Inflation Factor (VIF) coefficients for all independent variables are below 10, with many less than 2, confirming that the data does not violate multicollinearity assumptions.

All independent variables have positive regression coefficients, indicating a beneficial impact on the dependent variable According to the normalized regression coefficients (Beta), the independent variables influence the dependent variable G in the following order from strongest to weakest: [list the variables in order].

 Financial well-being has the strongest impact on personal financial management behavior

 Financial attitude has the second strongest impact on personal financial management behavior

 The Covid-19 pandemic has the third strongest impact on personal financial management behavior

 Financial socialization has the 4th strongest impact on personal financial management behavior

 The Financial knowledge has almost no impact on personal financial management behavior

Residuals may not follow a normal distribution due to factors such as erroneous model application, non-constant variance, or insufficient residual data To ensure accurate analysis, it is important to employ multiple survey methods, with creating a histogram of residuals being one of the simplest approaches to detect deviations from normality Using this visual tool helps identify potential issues in the model and guides further investigation for reliable statistical analysis.

Figure 4.3 Histogram of residual values

A bell-shaped normal distribution curve is overlaid on the histogram, indicating that the residuals follow a normal distribution pattern The mean is approximately 0, and the standard deviation of 0.995 suggests the residuals are close to a standard normal distribution This alignment confirms that the assumption of normal residual distribution is not violated, supporting the validity of the statistical analysis.

Aside from the Histogram, the P-P Plot is another common chart type for detecting violations of the normalized residual assumption

Figure 4.4 Normal P-P Plot of Regression Standardized Residual

In a Normal P-P Plot, residuals that closely align with the diagonal indicate a more regular and normally distributed residual pattern Conversely, the greater the dispersion of data points away from the diagonal, the less likely the residuals follow a normal distribution, signaling potential deviations from normality in the data This visual assessment helps in evaluating the appropriateness of the normality assumption in statistical analyses.

The chart indicates that the residual data points are clustered close to the diagonal, suggesting that the residuals follow an approximately normal distribution This clustering confirms that the assumption of normally distributed residuals is not violated, supporting the validity of the model's residual analysis.

In regression analysis, it is essential that the dependent and independent variables exhibit a linear relationship To assess this assumption, a scatter plot of normalized residuals versus normalized predictions can be used to identify any violations When the data points in the scatter plot form a tight, straight-line pattern, it indicates that the linearity assumption holds true and is not violated Ensuring this linear relationship is crucial for the accuracy and validity of regression models.

The scatter plot of normalized residuals reveals a linear distribution, indicating that the residuals are evenly spread along a straight line The vertical separation of points, based on their distances, demonstrates consistent variance across the data This confirms that the assumption of homoscedasticity is satisfied, supporting the validity of the linear relationship in the model.

This chapter presents a comprehensive quantitative analysis, including descriptive statistics that summarize the mean, maximum, minimum, standard deviation, and sample size of each variable To identify the key factors influencing the dependent variable, Exploratory Factor Analysis (EFA) and Cronbach's Alpha were employed to assess the data's reliability Correlation analysis was conducted to determine the positive or negative relationships between variables, with correlation coefficients measuring the strength of these associations Regression analysis was used to evaluate the impact of individual factors on the dependent variable The findings indicate that the research model is free from issues related to variable variance and autocorrelation, supporting its robustness.

CONCLUSIONS AND RECOMMENDATIONS

Tiêu đề	Factors affecting personal financial management behavior: Evidence from Vietnam
Tác giả	Bui Nguyen Thien Phuc
Người hướng dẫn	Nguyen Thi My Hanh, PhD
Trường học	Ho Chi Minh City University of Banking
Chuyên ngành	Finance - Banking
Thể loại	graduation project
Năm xuất bản	2022
Thành phố	Ho Chi Minh City

Định dạng
Số trang	90
Dung lượng	2,52 MB