Financial Literacy and Financial Behavior among Indonesian Undergraduates
A secondary analysis of published survey data
1 Introduction
Research problem and objectives
Financial literacy may help students evaluate financial choices, but knowledge alone does not determine whether they can save or invest. Income constraints, family background, and financial experience may influence both literacy and behavior. The research problem is therefore to establish whether the financial literacy measures available in a published Indonesian survey differ across students’ reported saving and market participation outcomes, without confusing those associations with the effects of financial education.
The research question is: Among respondents classified as undergraduates in Sjam’s (2025) survey, how are basic and advanced financial literacy indices associated with saving above consumption and participation in stocks or mutual funds? The unit of analysis is an individual respondent. The study is a secondary, cross-sectional analysis of 817 eligible records; it does not involve new participant recruitment.
The objectives are to describe the two literacy indices and outcome distributions, compare both indices across outcome groups, assess the sensitivity of outcome prevalence to missing responses, and specify a feasible subsequent adjusted analysis. Saving above consumption and market participation are distinct outcomes. Neither is treated as a universal measure of good financial decision-making: suitable choices depend on resources, needs, and risk tolerance.
H1 proposes that greater financial literacy is associated with a higher probability of reporting savings above consumption. H2 proposes that greater financial literacy is associated with a higher probability of market participation. Both are provisional expectations. Index construction and direction must be verified before a higher numeric score is interpreted as greater knowledge. Basic literacy is the primary explanatory measure and advanced literacy provides a secondary comparison. H1 has a more indirect basis in the reviewed evidence because the selected studies do not validate this dataset’s exact saving measure.
The preliminary results show differing basic-index contrasts across the two outcomes. These descriptive findings motivate closer examination of measurement and missingness; they do not establish that literacy causes saving or participation. The literature review develops the theoretical rationale, evaluates competing evidence, and explains the limits of applying it to this sample.
Data
Source and sample selection
The data source is Amelina Sjam’s Survey Data on Financial Literacy and Asset Participation among Young Adults in Indonesia, version 1, published on Mendeley Data on October 1, 2025 (Sjam, 2025; https://doi.org/10.17632/k269sxcw6h.1). The repository identifies the file as survey responses and releases it under CC BY 4.0. The downloaded Excel workbook is retained unchanged in the accompanying project package. The supplied package records a download date of October 31, 2026; repository metadata were checked on November 1, 2026.
The supplied workbook has one sheet, 952 rows of observations, 14 original variables, and 952 distinct respondent identifiers. The analytical subset contains 817 records classified as freshmen, sophomores, juniors, or seniors. Fifty graduate records and 85 records with missing academic year are excluded to maintain an undergraduate scope. No age variable is supplied, so an age restriction is not imposed or inferred. These counts were calculated directly from the workbook. Repository metadata and the file are the evidence for provenance; the original responses have not been independently authenticated.
The source page does not establish the collection dates, institutions, sampling frame, response rate, questionnaire wording, or index construction in sufficient detail for this project. Publication date is not treated as collection date, and a probability sample is not assumed. The results describe the retained records and cannot be generalized to all Indonesian undergraduates. This secondary analysis uses published respondent identifiers only, with no attempt to identify or contact individuals.
Variables and analytical coding
Table 1 Variable definitions and analytical coding
| Variable | Available measure and treatment |
|---|---|
| Financial literacy | Basic and advanced indices retained as numeric values supplied by the depositor. They are not percentages correct, and negative values do not mean negative knowledge. Construction and direction require source clarification. |
| Market participation | Stock market participation: 1=participate becomes 1; 0=otherwise becomes 0. Missing stays missing. Repository documentation defines this field to include stocks and mutual funds; it is not restricted to direct stock ownership. |
| Saving relative to consumption | Source Savings rate has three categories. Derived indicator is 1 for savings>consumption and 0 for savings=consumption or savings<consumption. It does not measure saving frequency or saving as a percentage of income. A zero can still describe a respondent who saves. |
| Background | Gender, academic year, employment, income category, economics education, economics in daily life, and parents’ financial knowledge are available. Preserve source categories. |
| Other assets | Non-market participation pools different assets and is not used as a substitute for regular saving or market participation. |
The income field contains Indonesian rupiah bands; it is not monthly disposable funds after tuition and housing, and no midpoint conversion is imposed. Parents’ financial knowledge is a categorical proxy for family background, rather than a direct measure of parental guidance. Economics education is not equivalent to a finance major or a verified course completion record. These distinctions prevent substituting convenient columns for different constructs.
Data preparation and missing responses
The analysis script reads the first sheet, retains the four specified undergraduate categories, creates the two binary outcomes using explicit mappings, and exports the analytical CSV. Original values are retained alongside derived fields. Empty cells remain missing and are never interpreted as nonparticipation or zero saving. There are no duplicate respondent identifiers in the full workbook. A SHA-256 checksum of the supplied workbook is recorded in the results file to detect later changes; this revision does not independently verify the workbook against a fresh repository download.
All 817 eligible records enter the descriptive dataset. Each statistic uses its own nonmissing denominator. A comparison of an index across outcome groups requires both variables to be observed. Missingness may be selective, especially for participation, so a complete-case comparison may differ from the complete undergraduate subset. No inferential claim is based on an assumption that missing observations are random.
Test extract from the published responses
The following 12 observations are the first 12 eligible records in source order, selected solely to demonstrate the data structure. They are an extract from the published survey, not a new survey or a random subsample. The main descriptive analysis uses all 817 eligible records; a brief calculation on this extract is also reported below. Basic and advanced indices are rounded to three decimals for display. NA means missing; binary outcomes follow the coding above.
Table 2 Twelve record extract from the published survey
| Source ID | Basic index | Advanced index | Saving above consumption | Market participation |
|---|---|---|---|---|
| 1 | -0.448 | 0.849 | NA | 0 |
| 2 | -1.388 | -0.579 | NA | 1 |
| 3 | 0.234 | 1.127 | 0 | 0 |
| 4 | 1.025 | -0.544 | 0 | 0 |
| 5 | -2.064 | -0.579 | NA | 1 |
| 6 | 0.841 | 0.274 | 1 | 0 |
| 7 | -2.064 | -0.190 | 0 | 0 |
| 8 | -1.388 | 1.019 | 0 | 0 |
| 9 | 0.234 | -0.003 | 0 | 0 |
| 10 | -2.064 | 1.460 | 0 | NA |
| 11 | -0.442 | 1.058 | 0 | NA |
| 12 | -1.124 | -0.502 | 0 | 0 |
Preliminary descriptive analysis
Table 3 Descriptive statistics for the undergraduate subset
| Measure | Valid N | Missing N | Result |
|---|---|---|---|
| Basic literacy index | 814 | 3 | Mean 0.026; sample SD 1.003 |
| Advanced literacy index | 814 | 3 | Mean −0.012; sample SD 0.990 |
| Market participation | 699 | 118 | 108 of 699 participate or 15.5% |
| Saving above consumption | 720 | 97 | 112 of 720 or 15.6% |
Among the 720 records with a saving category, 486 report savings below consumption (67.5%), 122 report equal saving and consumption (16.9%), and 112 report saving above consumption (15.6%). This field compares saving and consumption and does not identify whether respondents saved in each of the preceding three months. Market participation is missing for 118 of 817 records (14.4%), and the saving category is missing for 97 of 817 (11.9%).
{{COMPARISONS}}
The basic-index contrasts do not consistently follow the two directional hypotheses. They are reported without selecting only favorable comparisons. No p values or causal effects are claimed, and no hypothesis is declared supported. Before interpreting the numeric direction as a knowledge gradient, obtain documentation of index scoring and confirm that the participation labels match the original questionnaire. Resources, family background, missing responses, and selection could also influence the observed contrasts.
{{SENSITIVITY}}
Further analysis and limitations
After source coding is clarified, fit separate binary logistic models for saving above consumption and market participation. Use basic literacy as the primary predictor and advanced literacy in a separate sensitivity model, checking correlation before considering both together. The initial adjusted specification should prioritize income band and academic year, with parents’ financial knowledge considered as a family-background proxy. Employment, gender, and economics education are candidates for sensitivity analyses only if the retained outcome counts and cell sizes permit. Covariates should be selected for substantive reasons rather than automated significance screening. Report the complete-case N and outcome counts before choosing model complexity. Do not treat missing categories as substantive answers.
Review sparse cells, separation, convergence, collinearity, and the suitability of a linear index term. Report odds ratios and 95% confidence intervals, with Holm adjustment if the two primary literacy coefficients are tested as one family at .05. For saving, preserve the original three categories in a secondary descriptive comparison because dichotomization combines respondents whose savings equal consumption with those whose savings are lower. If the information needed to interpret the indices remains unavailable, retain descriptive analysis only. Any adjusted estimates remain associational. This project does not test whether a course caused behavioral change, and secondary data use must be acceptable under the course assignment.
2 Literature review
The review is organized around the mechanisms relevant to the research question: knowledge acquisition, family influences, participation decisions, and the distinction between education effects and observed associations. It is a focused thematic review of six academic papers, not a systematic review or a claim to cover all Indonesian research.
Financial knowledge and the transition to adulthood
Lusardi and Mitchell (2014) synthesize theoretical and empirical research on financial literacy and economic decision-making. Their human-capital perspective provides a reason to expect knowledge to matter: understanding financial concepts may improve the evaluation of financial opportunities and trade-offs. It also introduces an important complication. Individuals who expect to save or invest more may have greater incentives to acquire financial knowledge. Therefore, an observed relationship between literacy and behavior may reflect both the benefits of knowledge and the decision to learn.
Lusardi et al. (2010) bring this issue closer to the proposed population by examining young adults using the National Longitudinal Survey of Youth. Fewer than one-third demonstrated basic understanding of interest rates, inflation, and risk diversification. Financial literacy was also associated with sociodemographic characteristics and family financial sophistication. Their findings suggest that students may enter university with unequal financial preparation. However, the sample concerns young adults more broadly, so its results should not be treated as a direct estimate of literacy at a particular university.
Taken together, these papers support measuring financial knowledge directly rather than assuming that age or educational attainment is an adequate substitute. For this project, the supplied knowledge indices provide the available measures. Family-background categories offer limited context because differences in knowledge may partly reflect earlier opportunities to learn about money.
Financial literacy and investment participation
Van Rooij et al. (2011) examine financial literacy and stock market participation using the Dutch Household Survey. Their study distinguishes basic knowledge from more advanced understanding of financial instruments and diversification. People with lower financial literacy were less likely to participate in the stock market. This extends the review from general financial behavior to an outcome directly relevant to investment decisions and suggests that basic numeracy alone may not capture the knowledge associated with investing.
Applying these findings to university students requires care. Adult households in the Netherlands differ from students in income, accumulated assets, and financial responsibilities. A knowledgeable student may reasonably avoid investing because available funds are needed for tuition or emergencies. Consequently, the secondary analysis will examine participation alongside available income categories and distinguish participation from investment quality. Owning an investment does not, by itself, establish that a student has chosen an appropriate risk level or a diversified portfolio.
Whether financial education changes behavior
Evidence that knowledgeable people behave differently does not automatically show that financial education changes behavior. Fernandes et al. (2014) review 201 studies reported in 168 papers and find very small behavioral effects of financial education interventions in their evidence base. They also report that effects weaken over time and that estimated relationships between literacy and behavior become smaller when additional psychological characteristics are considered. Their findings caution against assuming that a course will reliably produce lasting changes in saving or investing.
Kaiser et al. (2022) reach a more positive conclusion in a meta-analysis of 76 randomized experiments involving more than 160,000 participants. They find positive average causal effects of financial education on both knowledge and financial behavior. Random assignment strengthens the interpretation that the programs produced the measured changes. Nevertheless, an average effect across programs and populations does not guarantee that every course will work equally well for university students.
The two reviews should be read together because they examine different evidence bases and use different approaches to synthesis. The later randomized evidence supports the potential value of financial education, while the earlier review highlights concerns about persistence and the gap between correlation and intervention effects. For the proposed project, this means distinguishing the available economics-education category, measured knowledge, and reported behavior. A one-time student survey can examine how these variables relate, but it cannot independently resolve whether education causes lasting behavioral change (Fernandes et al., 2014; Kaiser et al., 2022).
Synthesis and research gap
The reviewed literature connects financial knowledge with financial decisions, identifies family influences, and debates the effectiveness of education. However, the selected papers do not directly establish how the supplied literacy indices, financial resources, and family background jointly relate to saving and investment behavior in the Indonesian undergraduate subset analyzed here. This is a contextual gap within the reviewed evidence, rather than a claim that no previous research has examined students. The project contributes a transparent descriptive application to a clearly defined subset of published student responses. It does not claim a new causal identification strategy or national representativeness.
The analytical framework treats the supplied financial literacy indices as explanatory measures and saving relative to consumption and market participation as separate outcomes. Income and family background are possible shared influences. H1 and H2 are stated in the Introduction and remain provisional until index coding is verified. The preliminary basic-index market-participation contrast is numerically opposite to the expected direction, illustrating why results must be checked rather than assumed. Self-reports, missing responses, sample selection, and reverse causation limit interpretation.
The saving hypothesis deserves particular caution. General evidence on financial behavior and education provides a rationale for examining saving, but does not establish that a higher savings-to-consumption category is the appropriate benchmark for every student. The empirical contribution is consequently descriptive: documenting both outcomes, exposing inconsistent numerical patterns, and identifying which measurement questions must be resolved before stronger claims are made.
3 Reference
Fernandes, D., Lynch, J. G., Jr., & Netemeyer, R. G. (2014). Financial literacy, financial education, and downstream financial behaviors. Management Science, 60(8), 1861–1883. https://doi.org/10.1287/mnsc.2013.1849
Kaiser, T., Lusardi, A., Menkhoff, L., & Urban, C. (2022). Financial education affects financial knowledge and downstream behaviors. Journal of Financial Economics, 145(2), 255–272. https://doi.org/10.1016/j.jfineco.2021.09.022
Lusardi, A., & Mitchell, O. S. (2014). The economic importance of financial literacy: Theory and evidence. Journal of Economic Literature, 52(1), 5–44. https://doi.org/10.1257/jel.52.1.5
Lusardi, A., Mitchell, O. S., & Curto, V. (2010). Financial literacy among the young. Journal of Consumer Affairs, 44(2), 358–380. https://doi.org/10.1111/j.1745-6606.2010.01173.x
Shim, S., Barber, B. L., Card, N. A., Xiao, J. J., & Serido, J. (2010). Financial socialization of first-year college students: The roles of parents, work, and education. Journal of Youth and Adolescence, 39(12), 1457–1470. https://doi.org/10.1007/s10964-009-9432-x
Sjam, A. (2025). Survey data on financial literacy and asset participation among young adults in Indonesia (Version 1) [Data set]. Mendeley Data. https://doi.org/10.17632/k269sxcw6h.1
van Rooij, M., Lusardi, A., & Alessie, R. (2011). Financial literacy and stock market participation. Journal of Financial Economics, 101(2), 449–472. https://doi.org/10.1016/j.jfineco.2011.03.006