Understanding Personal Expenditure Patterns: An Exploratory and Inferential Analytics Study of a Five-Year Household Expense Tracker (2021–2025)
Author
Anurika Orabuche
Published
May 19, 2026
1. Executive Summary
This study analyses five years of daily household expenditure (January 2021 to December 2025) recorded personally through a structured daily expense tracker, representing a grand total of ₦155.57 million across eleven spending categories. The data was collected through daily manual entry into a personal spreadsheet — making it a genuinely primary dataset with full ownership and traceability.
Five analytical techniques reveal a clear and actionable financial story. Exploratory data analysis identifies the category “Others” (which includes project, capital, and miscellaneous spending) as the dominant budget item at 44.5% of all expenditure, while food (14.7%), family support (9.8%), and investment (6.9%) round out the top four. Data visualisation tells the broader story: annual totals oscillate between ₦21.5M and ₦43.6M, driven largely by the size of capital and project spending captured in “Others”, while all recurring essential categories — food (+341%), school fees (+347%), fuel (+576%) — show strong upward trends driven by Nigeria’s inflationary environment. Hypothesis testing confirms that food spending differs significantly across years (Kruskal-Wallis, p < 0.001), and that essential spending significantly exceeds discretionary spending (Wilcoxon, p = 0.031). Correlation analysis reveals that food and family support move in perfect lockstep (ρ = +1.0), and that investment and food are strongly negatively correlated (ρ = −0.90) — showing that as the cost of living rose, investment was progressively crowded out. Linear regression confirms that Others and food spending are the strongest predictors of annual total expenditure (R² = 0.98).
The primary recommendation is a structured three-bucket budget for 2026: Essentials (₦16M, inflation-indexed), Investment (₦4M, ring-fenced as a first deduction), and Others/Capital (capped at ₦12M, requiring explicit approval for any overage).
2. Professional Disclosure
Job Title: Chief Financial Officer Organisation: Orabuche Household — Nigerian Household Data Source: Personal daily household expense tracker, self-collected 2021–2025
Technique Justifications
Exploratory Data Analysis (EDA): As a finance professional I regularly audit portfolio and operational data before drawing conclusions. Applying the same discipline to personal expenditure data reveals the true composition of household spending — which categories are growing, which contain anomalies, and where data quality requires deliberate handling. This analysis identified that electricity data was absent across multiple years and was therefore excluded, and that personal “Others” spending ballooned significantly in 2025.
Data Visualisation: Budget discussions with family and financial advisers — like presentations to a credit committee — require visual storytelling. A chart showing that food spending quadrupled from ₦1.67M (2021) to ₦7.42M (2024) conveys the cost-of-living pressure instantly in a way that a table cannot. The five-plot structure here mirrors the portfolio dashboards I produce professionally.
Hypothesis Testing: Before recommending that the household budget be formally restructured, I must demonstrate that observed year-on-year spending increases represent statistically real structural shifts — not random variation. The Kruskal-Wallis and Wilcoxon tests applied here provide that formal rigour.
Correlation Analysis: Understanding which spending categories move together reveals the underlying financial dynamics of the household. The near-perfect negative correlation between investment and food spending (ρ = −0.90) is the single most important insight in the dataset — it quantifies the degree to which rising living costs are crowding out wealth-building.
Linear Regression: Regression of total annual expenditure on category predictors enables forward planning. A model that explains 98% of annual variance allows scenario-based budgeting for 2026 — the same forecasting discipline applied to loan portfolio projections at work.
3. Data Collection & Sampling
3.1 Source and Collection Method
The dataset was collected through daily manual entry into a structured Microsoft Excel spreadsheet, maintained by the author from January 2021 to December 2025. Each entry records a date, description, category, and amount for every expenditure event. Annual totals for each category are drawn from the validated footer rows of each annual sheet tab. The spreadsheet was accessed via Google Sheets and the annual summary figures extracted for this analysis. No external datasets supplement this study.
3.2 Sampling Frame
This is a complete census of all recorded household expenditures over the five-year period — not a sample. Every transaction entered into the tracker is included. The only potential gap is unrecorded small cash transactions, which may result in a slight undercount of very minor daily expenses.
3.3 Variables
Variable
Type
Description
year
Integer
Calendar year (2021–2025)
annual_total
Numeric (₦)
Total annual expenditure
food
Numeric (₦)
Groceries, provisions, and daily meals
gas
Numeric (₦)
Cooking gas refills
school_fees
Numeric (₦)
Children’s tuition and school-related costs
fuel
Numeric (₦)
Vehicle fuel (petrol/diesel)
medication
Numeric (₦)
Healthcare and pharmaceutical expenses
personal_care
Numeric (₦)
Grooming, clothing, and personal items
family_support
Numeric (₦)
Financial support to extended family members
data_airtime
Numeric (₦)
Mobile data and airtime
investment
Numeric (₦)
Investment outflows
loan_repayment
Numeric (₦)
Debt servicing payments
others
Numeric (₦)
All other expenditure including capital, project, and miscellaneous costs
Note on category consolidation: Electricity bill data was absent for multiple years and has been excluded from all analyses. Project and construction-related spending has been consolidated into the others category, which captures all capital, project, and miscellaneous outgoings as a single reporting line.
3.4 Time Period and Ethics
Period: January 2021 to December 2025 (5 complete calendar years, 60 months). This dataset is personal and relates solely to the author’s own household expenditure. No third-party personal or financial information is included.
Exploratory Data Analysis (EDA), formalised by Tukey (1977), uses statistical summaries and graphical displays to understand a dataset’s structure before applying formal models. Key activities include distributional assessment, missing-value classification, trend identification, and data quality auditing. Adi (2026) illustrates with Anscombe’s Quartet that summary statistics alone can conceal radically different structures — reinforcing why visual and numeric EDA are both essential before any inferential work.
5.2 Business Justification
Every sound personal financial plan begins with an honest audit of where money actually goes. EDA applied to five years of daily records reveals the true composition of household spending, identifies structural anomalies such as the investment collapse in 2024, and surfaces the data handling decisions — exclusion of electricity data, consolidation of project spending into Others — that must be documented before the analysis proceeds.
notes <-tibble(`#`=1:2,Decision =c("Electricity excluded from analysis","Project/construction costs consolidated into Others" ),Reason =c("Electricity data absent for multiple years — inclusion would introduce systematic zeros that distort distributional statistics and correlations","Project spending is a capital item that varies based on construction phase, not monthly consumption. Consolidating with miscellaneous Others captures total non-recurring spend in a single transparent line" )) kable(notes,caption ="Table 3: Data handling decisions and rationale") |>kable_styling(bootstrap_options =c("striped","hover"), full_width =FALSE)
Table 3: Data handling decisions and rationale
#
Decision
Reason
1
Electricity excluded from analysis
Electricity data absent for multiple years — inclusion would introduce systematic zeros that distort distributional statistics and correlations
2
Project/construction costs consolidated into Others
Project spending is a capital item that varies based on construction phase, not monthly consumption. Consolidating with miscellaneous Others captures total non-recurring spend in a single transparent line
Dominant category: “Others” (incorporating project, construction, and all capital spending) accounts for 44.5% of all expenditure across five years — ₦69.3M out of ₦155.6M total. This is the single largest budget driver and the primary source of year-to-year volatility. Managing the size and timing of capital spending is the most powerful lever available to control total household expenditure.
Inflationary pressure on essentials: All core recurring categories — food, school fees, and fuel — more than tripled over the five-year period. Fuel saw the sharpest increase (+576%), reflecting the removal of the fuel subsidy in 2023. These are structural cost increases, not lifestyle changes, and they must be factored into any forward budget as a permanent step-up.
Investment collapse: Investment fell from ₦6.28M in 2021 to zero in 2024, as rising living costs absorbed all available cash. The partial recovery to ₦2M in 2025 is a positive sign but still well below the 2021 level. This is the most significant personal financial risk identified in the data.
2024 as the leanest Others year: The ₦5.0M “Others” spend in 2024 — the lowest in the five-year period — was offset by the highest food and fuel costs, confirming that 2024 was a year in which capital spending was deliberately deferred to manage rising living costs.
6. Technique 2 — Data Visualisation
6.1 Theory
The grammar of graphics (Wilkinson, 2005), implemented in R’s ggplot2, maps data attributes to visual channels — position, colour, size — in a principled framework. Effective business visualisation selects chart types matched to the relationship being shown, eliminates non-data ink, and ensures the key message is legible without specialist training (Adi, 2026, Ch. 5). The five plots below form a deliberate narrative arc from portfolio overview to the critical investment trade-off.
6.2 Business Justification
Budget discussions within the household — and periodic reviews with a financial adviser — require clear visual communication. A chart showing that food spending grew from ₦139K per month (2021) to ₦618K per month (2024) communicates urgency in a single glance. The five plots here directly mirror the analytical dashboards I prepare for management audiences professionally.
6.3 Five-Plot Visual Narrative
The five plots tell one story: total household spending has grown substantially and unevenly, essentials have risen sharply due to inflation, and investment has been systematically crowded out — requiring deliberate structural budget protection.
import matplotlib.pyplot as pltfig, axes = plt.subplots(1, 2, figsize=(13, 4))axes[0].plot([2021,2022,2023,2024,2025], [1.67,2.29,4.11,7.42,7.36], 'o-', color="#C0392B", linewidth=2, label="Food")axes[0].plot([2021,2022,2023,2024,2025], [6.28,2.33,0.09,0,2.0], 's--', color="#0E8A72", linewidth=2, label="Investment")axes[0].set_title("Food vs Investment (₦M)", fontsize=11)axes[0].set_ylabel("₦ million"); axes[0].legend(fontsize=9)axes[0].set_xticks([2021,2022,2023,2024,2025])cats_2025 = {'Others/Capital':20.18,'Food':7.36,'Family Support':3.21,'Personal Care':3.61,'School Fees':3.51,'Fuel':2.96}axes[1].barh(list(cats_2025.keys()), list(cats_2025.values()), color=["#C0392B"if v>5else"#1A56A0"for v in cats_2025.values()], alpha=0.85)axes[1].set_title("2025 Spend by Category (₦M)", fontsize=11)axes[1].set_xlabel("₦ million")plt.tight_layout()plt.savefig("viz_py.png", dpi=150, bbox_inches="tight")plt.show()
6.4 Business Interpretation
Figure 2 establishes that annual spending oscillated between ₦21.5M and ₦43.6M — driven primarily by the size of capital and project spending within the Others category. Figure 3 reveals that every recurring essential category is on an upward trajectory, with fuel (+576%) and school fees (+347%) representing the steepest inflation. Figure 4 shows 2025 as the highest-spending year, with Others at ₦20.2M and food at ₦7.4M together accounting for 63% of total annual spend. Figure 5 delivers the strategic core of the analysis: as food costs quadrupled, investment was eliminated entirely by 2024. This is not a coincidence — it is the financial signature of a cost-of-living crisis at the household level. Figure 6 contextualises the five-year totals: over half of all money spent (57.7%) sits in the Others/Capital bucket — highlighting that controlling capital spending timelines is the primary budget management lever.
7. Technique 3 — Hypothesis Testing
7.1 Theory
Hypothesis testing provides a formal framework for distinguishing real trends from random year-to-year fluctuation. The analyst specifies H₀ (no effect) and H₁ (an effect exists), selects a test appropriate to data type and distributional assumptions, and evaluates both p-value and effect size. A null result is as analytically valuable as a significant result — it prevents over-reaction to patterns that are within normal variation (Adi, 2026, Ch. 6).
7.2 Business Justification
Before formally restructuring the household budget — committing to higher monthly food and fuel allocations — it is important to establish that the observed increases reflect statistically real structural shifts rather than one-off fluctuations. The tests below provide that rigour.
7.3 Hypothesis 1 — Does Monthly Food Spending Differ Significantly Across Years?
H₀: Monthly food expenditure is identically distributed across all five years (2021–2025). H₁: At least one year has a significantly different monthly food expenditure distribution. Test: Kruskal-Wallis — appropriate for non-normal data across multiple independent groups. α = 0.05
df |>transmute(Year = year,`Annual Food (₦M)`=round(food/1e6, 3),`Monthly Avg (₦)`=comma(round(food/12)),`% of Annual Budget`=paste0(round(food/annual_total*100, 1), "%") ) |>kable(caption ="Table 6: Food spending by year") |>kable_styling(bootstrap_options =c("striped","hover"), full_width =FALSE)
Table 6: Food spending by year
Year
Annual Food (₦M)
Monthly Avg (₦)
% of Annual Budget
2021
1.668
139,030
5.6%
2022
2.290
190,823
10.7%
2023
4.109
342,455
11.4%
2024
7.422
618,478
29.8%
2025
7.362
613,508
16.9%
Code
df_monthly |>mutate(year_f =factor(year)) |>ggplot(aes(x = year_f, y = food/1e3, fill = year_f)) +geom_boxplot(alpha =0.7, outlier.colour ="#C0392B") +scale_fill_brewer(palette ="Set1", guide ="none") +scale_y_continuous(labels =label_number(suffix ="K")) +labs(title ="Figure 7: Monthly food expenditure distribution by year (₦ thousands)",subtitle =glue("Kruskal-Wallis χ²({kw1$parameter}) = {round(kw1$statistic,1)}, p < 0.001 — year-on-year differences confirmed"),x ="Year", y ="Monthly food (₦ thousands)" ) +theme_minimal(base_size =11)
Code
from scipy import stats as scfood_by_year = [[df_py[df_py['year']==yr]['food'].values[0]/12]*12for yr in [2021,2022,2023,2024,2025]]h, p = sc.kruskal(*food_by_year)print(f"Kruskal-Wallis (Food ~ Year): H = {h:.3f}, p = {p:.2e}")
Kruskal-Wallis (Food ~ Year): H = 59.000, p = 4.71e-12
Code
for yr in [2021,2022,2023,2024,2025]: val = df_py[df_py['year']==yr]['food'].values[0]print(f" {yr}: ₦{val/1e6:.3f}M annual (₦{val/12:,.0f}/month)")
Business interpretation: Food spending differences across years are statistically confirmed — not noise. Monthly food costs rose from an average of ₦139,029 (2021) to ₦618,478 (2024), a 345% increase. For the household budget: “A 2026 food budget must be set at a minimum of ₦7.5–8M annually. Basing any budget on 2021 or 2022 figures would understate food costs by more than ₦5M and create a systematic monthly deficit.”
H₀: Essential spending (food + school fees + fuel + medication + gas) is no greater than discretionary spending (personal care + data/airtime) across the five years. H₁: Essential spending significantly exceeds discretionary spending. Test: Wilcoxon signed-rank test (paired, one-sided). α = 0.05
wilcox_res <-wilcox.test(df$essentials, df$discretionary,paired =TRUE, alternative ="greater")cat(glue("Wilcoxon Signed-Rank (Essentials > Discretionary): V = {wilcox_res$statistic}, p = {round(wilcox_res$p.value, 4)}\n"))
Wilcoxon Signed-Rank (Essentials > Discretionary):
V = 15,
p = 0.0312
Code
df |>transmute(Year = year,`Essentials (₦M)`=round(essentials/1e6, 2),`Discretionary (₦M)`=round(discretionary/1e6, 2),`Essentials / Discret.`=round(essentials/discretionary, 1) ) |>kable(caption ="Table 7: Essentials vs Discretionary spending by year") |>kable_styling(bootstrap_options =c("striped","hover"), full_width =FALSE)
Table 7: Essentials vs Discretionary spending by year
Year
Essentials (₦M)
Discretionary (₦M)
Essentials / Discret.
2021
3.08
1.43
2.2
2022
4.31
1.10
3.9
2023
6.47
2.22
2.9
2024
12.25
2.66
4.6
2025
14.04
3.84
3.7
Code
from scipy.stats import wilcoxoness = [df_py[df_py['year']==yr][['food','school_fees','fuel','medication','gas']].sum(axis=1).values[0]for yr in [2021,2022,2023,2024,2025]]disc = [df_py[df_py['year']==yr][['personal_care','data_airtime']].sum(axis=1).values[0]for yr in [2021,2022,2023,2024,2025]]stat, p = wilcoxon(ess, disc, alternative='greater')print(f"Wilcoxon (Essentials > Discretionary): V = {stat}, p = {p:.4f}")
Wilcoxon (Essentials > Discretionary): V = 15.0, p = 0.0312
Code
for i,yr inenumerate([2021,2022,2023,2024,2025]): ratio = ess[i]/disc[i]print(f" {yr}: Essentials ₦{ess[i]/1e6:.2f}M vs Discretionary ₦{disc[i]/1e6:.2f}M (ratio {ratio:.1f}×)")
2021: Essentials ₦3.08M vs Discretionary ₦1.43M (ratio 2.2×)
2022: Essentials ₦4.31M vs Discretionary ₦1.10M (ratio 3.9×)
2023: Essentials ₦6.47M vs Discretionary ₦2.22M (ratio 2.9×)
2024: Essentials ₦12.25M vs Discretionary ₦2.66M (ratio 4.6×)
2025: Essentials ₦14.04M vs Discretionary ₦3.84M (ratio 3.7×)
Result: V = 15, p = 0.031. Reject H₀.
Business interpretation: Essential spending is statistically and practically larger than discretionary spending across all five years, and the ratio has grown from 2.2× (2021) to 3.7× (2025). For the household budget: “Discretionary spending — personal care and data/airtime — is the only adjustable line in the budget when cashflow is under pressure. All other recurring categories are either rising due to inflation (food, fuel, school fees) or non-negotiable (family support). Any budget reduction must target discretionary items first.”
8. Technique 4 — Correlation Analysis
8.1 Theory
Correlation analysis measures the strength and direction of association between pairs of numeric variables. Spearman’s ρ is used throughout as the non-parametric rank-based alternative, appropriate given the small sample (n=5 years) and non-normal distributions. The core principle: correlation is not causation — a strong correlation may reflect a common external driver rather than a direct relationship between the two categories (Adi, 2026, Ch. 8).
8.2 Business Justification
Understanding which spending categories move together — and which trade off against each other — is the foundation of intelligent budgeting. If investment and food are strongly negatively correlated, I cannot allow food costs to rise freely without consciously protecting the investment line. Correlation analysis makes this dynamic quantitative and actionable.
1. Food ↔︎ Family Support (ρ = +1.0, perfect positive) These two categories have moved in perfect lockstep over five years. Both are driven by rising living costs and household obligations — when food prices rise, the support needed by extended family members rises proportionally. Budget implication: these must be planned together as a single “household obligations” line, not managed independently.
2. Investment ↔︎ Food (ρ = −0.90, strong negative) As food spending quadrupled, investment was progressively eliminated. This is the most critical finding in the dataset. Budget implication: investment cannot be left as a residual item after other spending. It must be automated as a fixed monthly deduction before living costs are allocated — otherwise, inflationary pressure on food will continue to crowd it out.
3. School Fees ↔︎ Fuel (ρ = +1.0, perfect positive) Both categories rose in perfect tandem — both are non-negotiable costs driven by inflation and life-stage obligations. Together they now account for ₦6.47M annually (2025). Budget implication: these are core non-discretionary essentials and should be the first items confirmed in any annual budget exercise.
4. Loan Repayment ↔︎ Others (ρ = −0.50, moderate negative) As loans were repaid over the period, Others spending grew — suggesting that cash freed from debt servicing was absorbed into capital and miscellaneous spending rather than redirected to investment. Budget implication: when the remaining loan is fully retired, the freed monthly cash must be consciously redirected to the investment bucket rather than allowed to diffuse into untracked expenditure.
Causation note: Most of these correlations are driven by Nigeria’s inflationary environment as a common external factor. A controlled experiment would be needed to isolate direct causal relationships.
9. Technique 5 — Linear Regression
9.1 Theory
Ordinary Least Squares (OLS) linear regression models the relationship between a continuous outcome variable and one or more predictors by minimising the sum of squared residuals. In a log-log specification, coefficients represent elasticities — the percentage change in the outcome associated with a one-percent change in the predictor. Model fit is assessed using R² (proportion of outcome variance explained by the model) (Adi, 2026, Ch. 9).
Important caveat on sample size: With only n=5 annual observations, OLS regression is used here as a descriptive and scenario-planning tool rather than an inferential one. Coefficient p-values cannot be interpreted in the conventional hypothesis-testing sense. The value of the model lies in quantifying relative category contributions to total spending and enabling forward scenario planning — not in statistical significance.
9.2 Business Justification
A regression of annual total expenditure on category-level spending answers the most actionable budgeting question: which categories, when they increase, pull the annual total up most strongly? If Others spending has an elasticity of 0.40, a 10% reduction in capital and project spending would reduce the annual total by approximately 4% — providing a direct estimate of the savings achievable through project timing decisions.
Base (Others ₦8M): ≈ ₦30.8M
Mid (Others ₦15M): ≈ ₦39.4M
High (Others ₦20M): ≈ ₦44.2M
9.4 Business Interpretation of Coefficients
The log-log model achieves R² ≈ 0.98, confirming that these five categories together explain essentially all variation in annual total expenditure. The coefficients are elasticities:
Others/Capital (β ≈ 0.40): A 1% increase in Others spending is associated with a 0.40% increase in total annual expenditure. Since Others captures all project and capital outgoings, this confirms that managing the size and timing of capital spending is the most powerful budget lever available. A decision to defer ₦5M of project spending (approximately 25% of a ₦20M Others budget) would reduce total annual expenditure by approximately 10%.
Food (β ≈ 0.25): Food spending has a meaningful elasticity. Given the structural inflation already observed (food rose 341% from 2021 to 2025), further increases are expected. A 2026 food budget of ₦8M — a 9% increase over 2025 — is a conservative but realistic estimate. For every ₦1M overspend on food, the household total increases by approximately ₦250K.
Fuel (β ≈ 0.30): Fuel has the second-highest elasticity and is the most externally driven category — prices are set by the market, not by household behaviour. The 576% increase since 2021 has already materially impacted the budget. Fuel costs must be budgeted conservatively at ₦3.5–4M for 2026.
Deployment recommendation: Build a three-scenario annual budget — Base (Others ₦8M, estimated total ~₦30M), Mid (Others ₦15M, ~₦37M), High (Others ₦20M, ~₦44M) — and use the model to estimate total expenditure under each scenario. This replaces ad hoc budgeting with a data-driven planning framework.
10. Integrated Findings
The five analytical techniques construct a clear, interconnected financial narrative.
EDA established the data landscape and confirmed the two key structural decisions — excluding electricity and consolidating project costs into Others — before analysis proceeded. It revealed that Others (44.5%), food (14.7%), and family support (9.8%) together account for nearly 70% of all five-year spending, and that every recurring category has trended sharply upward. Visualisation transformed these patterns into strategic clarity: the cost-of-living increase across all essentials is steep and sustained, and 2025 is the highest-cost year on record at ₦43.6M.
Hypothesis testing confirmed that food spending increases are statistically real (p < 0.001) — not noise — and that essentials now significantly exceed discretionary spending across all years (p = 0.031), with the gap continuing to widen. Correlation analysis identified the two most actionable financial relationships: food and family support move together perfectly (ρ = +1.0), requiring joint planning; and investment and food are strongly negatively correlated (ρ = −0.90), confirming that rising living costs are actively crowding out wealth-building. Linear regression quantified the relative sensitivity of total spending to each category, confirming Others/capital and fuel as the primary variance drivers and enabling scenario-based forward planning.
Single integrated recommendation: Implement a Three-Bucket Protected Budget for 2026: (1) Essentials bucket — set at ₦16M, covering food (₦8M), fuel (₦3.5M), school fees (₦4M), and medication/gas; inflation-indexed annually; (2) Investment bucket — ring-fenced at ₦4M, deducted automatically before any other discretionary spending to reverse the 2024 collapse; (3) Others/Capital bucket — capped at ₦12M, with any project or capital spending above this ceiling requiring an explicit household budget review. Together these three rules address every finding across all five analytical techniques and provide a structured, evidence-based framework for 2026 financial management.
11. Limitations & Further Work
1. Only five annual observations. With n=5, regression coefficients and correlation values are directionally informative but statistically fragile. Any additional year of data would materially improve model reliability. A daily-level transaction dataset would enable proper time-series analysis.
2. Monthly distributions assumed uniform. Annual totals were divided equally across 12 months for hypothesis testing. Real spending is seasonal — school fees in January and September, fuel spikes in certain quarters. The true monthly distribution would produce stronger and more reliable test statistics.
3. “Others” category requires decomposition. At ₦20.2M in 2025 (46.3% of total), Others is now the largest single line in the budget by a considerable margin. Its consolidated nature — mixing construction costs, gifts, transport, and miscellaneous — makes it difficult to manage precisely. Breaking it into at least “Project/Construction” and “Miscellaneous” would unlock significantly more analytical value.
4. No income data included. Without monthly income figures, it is impossible to calculate savings rates or assess whether expenditure trajectories are sustainable. Adding an income column to the tracker would transform it from an expense record into a complete personal financial management tool.
5. Inflation adjustment not applied. All values are in nominal naira. Adjusting by Nigeria’s CPI would separate genuine consumption growth from pure price-level effects — particularly important for fuel and food, where most of the increase reflects market price changes rather than increased consumption.
References
Adi, B. (2026). AI-powered business analytics: A practical textbook for data-driven decision making — from data fundamentals to machine learning in Python and R. Lagos Business School / markanalytics.online. https://markanalytics.online
Allaire, J. J., Teague, C., Scheidegger, C., Xie, Y., & Dervieux, C. (2022). Quarto (Version 1.x) [Computer software]. https://doi.org/10.5281/zenodo.5960048
Orabuche, A. (2026). Personal household expense tracker — January 2021 to December 2025 [Dataset]. Self-collected daily expenditure records, maintained by the author. Data available on request.
R Core Team. (2024). R: A language and environment for statistical computing (Version 4.x). R Foundation for Statistical Computing. https://www.R-project.org/
Tukey, J. W. (1977). Exploratory data analysis. Addison-Wesley.
Van Rossum, G., & Drake, F. L. (2009). Python 3 reference manual. CreateSpace.
Wickham, H. (2016). ggplot2: Elegant graphics for data analysis. Springer. https://doi.org/10.1007/978-3-319-24277-4
Wickham, H., Averick, M., Bryan, J., Chang, W., McGowan, L., François, R., Grolemund, G., Hayes, A., Henry, L., Hester, J., Kuhn, M., Pedersen, T. L., Miller, E., Bache, S. M., Müller, K., Ooms, J., Robinson, D., Seidel, D. P., Spinu, V., … Yutani, H. (2019). Welcome to the tidyverse. Journal of Open Source Software, 4(43), 1686. https://doi.org/10.21105/joss.01686
Wilkinson, L. (2005). The grammar of graphics (2nd ed.). Springer.
Appendix: AI Usage Statement
Claude (Anthropic) was used during the preparation of this assignment to assist with structuring the Quarto document template and generating R and Python code scaffolding for the five analytical sections. All analytical decisions — the choice to consolidate project spending into Others, the decision to exclude electricity data, the identification of the investment-versus-food crowding-out dynamic as the primary analytical finding, the selection and justification of each statistical test, the interpretation of all regression coefficients and correlation values, the three-bucket budget recommendation, and all limitations identified — were made independently by the author based on direct review of five years of their own personally maintained financial records. The author takes full responsibility for all conclusions and is prepared to explain and defend every result in the viva voce examination.
Data Analytics 1 — Capstone Case Study | Lagos Business School | April 2026Submitted to: Prof Bongo Adi (badi@lbs.edu.ng)
