Yila
All use cases
Data Analysis

Milk-tea taste data: describe, compare, correlate, regress

Upload a 180-row teaching mock survey. Yila checks the fields and missingness first, then in one reproducible script runs descriptives, a three-group sweetness ANOVA, a taste–purchase-intent correlation, and a multiple regression controlling for price, sweetness preference, and age — methods justified by diagnostics, association not treated as cause.

This is mock milk-tea taste data for a teaching demo. Read the actual file — do not invent numbers — and finish four analyses in one reproducible script: descriptives, a three-group sweetness comparison, a taste–purchase-intent correlation, and a multiple regression controlling for acceptable price, sweetness preference, and age. Organize each part as question / why this method / results / how to write it, keep association separate from causation, and deliver clear figures, result tables, and a reproducible script.

Data check complete: 180 rows, 9 Chinese survey columns, 60 people in each ice-level group; 4 missing cells in sweetness preference, acceptable price, and purchase intent; no duplicate rows. A few IQR-flagged points are kept for diagnosis, not auto-deleted.

All four parts ran in one script. Group-wise Shapiro–Wilk and Levene tests were non-significant, so sweetness was compared with ordinary one-way ANOVA: F(2, 177) = 53.09, p < .001, η² = .375; Tukey HSD shows regular ice < less ice < no ice. Taste and purchase intent: Pearson r = .400 (n = 178). After controlling for price, sweetness preference, and age, taste remains positively associated with purchase intent (B = 0.50), model R² = .250 — cross-sectional data support association, not cause.

All four parts ran in one script. Group-wise Shapiro–Wilk and Levene tests were non-significant, so sweetness was compared with ordinary one-way ANOVA: F(2, 177) = 53.09, p < .001, η² = .375; Tukey HSD shows regular ice < less ice < no ice. Taste and purchase intent: Pearson r = .400 (n = 178). After controlling for price, sweetness preference, and age, taste remains positively associated with purchase intent (B = 0.50), model R² = .250 — cross-sectional data support association, not cause.
chart.pngTeaching figure · PNG
chart.svgEditable vector · SVG
stats_results.csvStatistics table · CSV
analysis_data.csvAnalysis data · CSV
analysis.pyReproduction script · Python
Task completed

Yila task replay complete.

Start with this method