This project presents a Conjoint Analysis conducted to understand consumer preferences for mobile phone features. The analysis uses linear regression to compute the part-worth utilities (coefficients) of different attribute levels and interpret which combinations are most desired by consumers.
| File/Folder | Description |
|---|---|
data.xlsx |
Dataset with 36 mobile phone profiles ranked by users |
Conjoint_analysis.docx |
Full report including methodology and results |
images/ |
Contains regression outputs and charts used in the report |
To determine how various mobile phone features—price, battery life, camera resolution, and screen size—influence user preferences, using Conjoint Analysis and SPSS regression output.
We aim to identify which attributes most impact consumer ranking to inform product design and marketing strategy.
Each row in the dataset represents one mobile phone profile shown to participants. The following attributes were included:
| Attribute | Levels |
|---|---|
| Price | ₹599, ₹699, ₹799 |
| Battery | 2 days, 3 days |
| Camera | 48MP, 108MP |
| Screen | 6.1", 6.7", 7.2" |
| Ranking | 1 (Most preferred) to 16 (Least preferred) |
📌 Total Profiles Ranked: 36
📌 Respondents: 10
- Used dummy variable regression in SPSS to model rankings as a function of product features.
- Computed part-worth utilities for each attribute level.
- Evaluated model fit using R², F-statistic, and Significance levels.
- Interpreted coefficients to derive consumer preference insights.
- R = 0.949 → Strong linear relationship
- R² = 0.900 → 90% of the variance in rankings explained by features
- Adjusted R² = 0.895
- Standard Error = 1.502
✅ Indicates excellent model fit.
| Metric | Value |
|---|---|
| F-statistic | 150.827 |
| Significance (p) | < 0.001 |
✅ Statistically significant model, showing feature variables significantly affect rankings.
Interpretation Highlights:
| Feature | Coefficient | Interpretation |
|---|---|---|
| Intercept | 11.889 | Baseline ranking |
| Price ₹699 | +6.462 | Most preferred price point |
| Battery 2 days | -2.723 | Surprisingly more preferred than 3 days |
| Camera 48MP | -1.451 | Preferred over 108MP |
| Screen 6.1" | -1.195 | Most preferred screen size |
| Screen 7.2" | -4.627 | Least preferred screen |
📌 Consumers do not always favor higher specs—pricing, compact screen, and mid-range cameras are appreciated more.
This output from SPSS shows t-values, standard errors, and significance for each feature level. All significant at p < 0.05.
This table shows the average utility scores and rankings for each profile. Higher utility → better preference.
- 📉 High price (₹799) receives no preference uplift — customers prefer ₹699 or even ₹599.
- 🔋 2-day battery is unexpectedly preferred over 3 days—possible user perception or trade-off effect.
- 📷 48MP cameras outperform 108MP, which could indicate diminishing returns.
- 📱 Compact screens like 6.1" are clearly favored over 7.2".
Conjoint analysis reveals that price and screen size have the strongest influence on preference rankings. Contrary to expectations, mid-tier features are often more desirable. These insights can help companies design optimal products and set strategic pricing.
- SPSS – Regression analysis, dummy coding, ANOVA
- Excel – Profile construction, utility calculation
- Markdown – Documentation
- GitHub Pages – Project deployment
View hosted project at:
➡️ https://suryaprakashj123.github.io/Cluster_Analysis
- Green, P.E. & Rao, V.R. (1971). Conjoint Measurement for Quantifying Judgmental Data. Journal of Marketing Research.
- Hair et al. (2010). Multivariate Data Analysis. Pearson.