Statistical Modelling in User Experience Design of Detergent Packages †
1
Department, Industrial Engineering Design. Technical University of Sofia, 1756 Sofia, Bulgaria
2
Machine Elements and Non-Metallic Constructions, Technical University of Sofia, 1756 Sofia, Bulgaria
3
Department of Computing and Information Technology, The University of the West Indies, St. Augustine 330912, Trinidad and Tobago
*
Author to whom correspondence should be addressed.
†
Presented at the 15th International Scientific Conference TechSys 2026—Engineering, Technologies and Systems, Plovdiv, Bulgaria, 14–16 May 2026.
Eng. Proc. 2026, 150(1), 72; https://doi.org/10.3390/engproc2026150072
Published: 24 July 2026
(This article belongs to the Proceedings of The 15th International Scientific Conference TechSys 2026—Engineering, Technology and Systems)
Abstract
A method for user experience design of detergent packages is proposed. Using a checklist, user experience with detergent packages is assessed. Through a statistical model, the most important package design elements, which provide positive emotional user experience, are determined. User experience design recommendations for good and bad design of detergent packages are defined. Within a case study, the method was experimentally used for design of detergent packages. Further research in this new and very promising area is discussed.
1. Introduction
Consumer perceptions and emotional responses to a product play a crucial role in purchasing decisions. “The creation of appealing packaging must take into account the emotional reactions they trigger in consumers” [1,2]. To incorporate these emotional considerations into packaging design, firms must adopt techniques that translate subjective and unconscious emotional responses into tangible design features [3,4].
Designers typically begin with user requirements, prioritizing both functionality and visual appeal in package interaction. Establishing a user-friendly and manageable packaging process fosters a positive user experience, which is fundamental to influencing consumers’ buying decisions [5,6,7].
Packaging design transcends visual appeal; it integrates the consumers’ thoughts, emotions, and actions to forge a lasting emotional bond with the detergent. Factors like tactile feedback, visual appeal, and user-friendliness play a critical role in shaping favourable brand perception.
A method for gathering UX data through interaction with detergent packages is proposed and illustrated through a case study [8]. Design guidelines are established and applied to create an optimal detergent package solution.
2. Method for User Experience Design of Detergent Packages
The method includes UX evaluation followed by statistical modelling and analysis of the collected data [9,10,11,12,13,14].
In step one, a checklist is defined and used to assess user impressions of pre-selected detergent packages using a 5-point semantic differential scale [15,16].
In step two, packages are assessed using Kansei words [17].
In step three, statistical modelling is performed. Due to the large number of variables and multicollinearity, a Partial Least Squares Regression (PLSR) model is applied [9,11,13,18]. Ordinary Least Squares (OLS) is unsuitable because correlated independent variables violate model assumptions. PLSR instead finds multidimensional directions in the X-space that best predict Y.
In step four, design recommendations are derived from the statistical analysis.
In step five, an optimal UX-driven detergent package is designed. The method steps are shown in Figure 1.
3. Case Study
The method proposed is illustrated by a case study of eleven detergent packages from package brands available on the market.
Within the process of creating the checklist (step 1), four pilot tests with small groups of five participants (students at the Department of Industrial Design Engineering of Technical University of Sofia) were performed.
In the second step, 70 users (BSc students in industrial design engineering) rated their experience with 11 detergent packages for 23 design elements comprising 72 units using 21 visual, tactile and usage Kansei impression words.
In the third step, statistical modelling and analysis of UX data was performed. Because of the large number of variables (23) and their units (72), a multivariate statistical PLRS model was applied.
All design elements (23) and their 72 units are represented as independent variables, and all Kansei word pairs describing user experience (visual, tactile and usage) are presented as 21 dependent UX variables. Some of the regression coefficients obtained are show in Table 1.
The regression coefficients concerning the values of visual, tactile and usage impressions for all 72 categories of the 23 design elements were taken from the total UX data set. For all 21 Kansei word pairs, statistical analysis was performed to find the most important design elements above the mean for the specific SD.
The range of the regression coefficient values was calculated from the regression coefficients of design units for each of the design elements. The distribution of all 72 values of UX impressions for the design elements and their units above average values (0.0981) is presented in Table 2.
The most important design elements and their recommended design units with range values above the average value (0.0981) for the positive impressions of the consumers are the following: package shape (0.3782)—square, value image (0.1819)—light, position of trade name (0.1623)—below, logo size (0.1502)—small, style of letters in trade name (0.1346)—lower case, package width (0.1339)—narrow, image size (0.1337)—small, package height (0.1254)—high, front trade name optical value (0.1047)—bold, and dosimeter (0.1007)—available.
The non-recommended design units (bad design) for the relevant design elements’ values are as follows: package shape—trapezoid, value image—neutral, position of trade name—bottom right, logo size—large, style of letters in trade name—upper lowercase, package width—high, image size—large, package height—low, front trade name optical value—light, and dosimeter—unavailable.
In the fourth step, UX design recommendations for detergent packages are defined (cf. Table 2). They can serve as a basis for introducing well design practices in detergent package design.
In step five, an optimal detergent package was designed based on UX design recommendations. It combines the best UX evaluations of user impressions from UX assessments. The optimal package is expected to trigger a positive user experience due to the optimal values of design elements presented in Figure 2.
4. Conclusions
A method focusing on the impact of user experience on design is proposed to highlight recommendations that could help designers to determine the most important design elements in detergent package design. It uses user experience design to stimulate, attract, and convince customers, and motivate their buying behaviour and decision-making. It was found that it is possible to measure emotional user experience through a statistical model. User experience recommendations for detergent packages are defined. This research presents a first step in this very promising research field of emotional user experience detergent package design.
Author Contributions
Conceptualization, A.R. and A.N.; methodology, A.N.; software, A.R.; validation, A.R., L.D. and A.N.; formal analysis, A.N.; investigation, A.R.; resources, A.R.; data curation, A.R.; writing—original draft preparation, A.R.; writing—review and editing, L.D.; visualization, A.R.; supervision, A.N.; project administration, L.D.; funding acquisition, L.D. All authors have read and agreed to the published version of the manuscript.
Funding
This research was funded by Technical university of Sofia.
Institutional Review Board Statement
Ethical review and approval were waived for this study due to the non-invasive nature of the survey, which involved only anonymous evaluations of detergent package design elements by adult participants without collecting personal or sensitive data.
Informed Consent Statement
Informed consent was obtained from all subjects involved in the study.
Data Availability Statement
The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.
Acknowledgments
The authors would like to thank the Research and Development Sector at the Technical University of Sofia for the financial support.
Conflicts of Interest
The authors declare no conflict of interest.
Abbreviations
The following abbreviations are used in this manuscript:
| UX | User Experience |
| PLSR | Partial Least Squares Regression |
| OLS | Ordinary Least Squares |
| SD | Semantic Differential |
References
- Park, J.; Han, S.H.; Kim, H.K.; Oh, S.; Moon, H. Modeling user experience: A case study on a mobile device. Int. J. Ind. Ergon. 2013, 43, 187–196. [Google Scholar] [CrossRef]
- Desmet, P.; Overbeeke, K.; Tax, S. Designing Products with Added Emotional Value: Development and Application of a Method for Research Through Design. Des. J. 2001, 4, 32–47. [Google Scholar] [CrossRef]
- Jokinen, J.P. Emotional user experience: Traits, events, and states. Int. J. Hum.-Comput. Stud. 2015, 76, 67–77. [Google Scholar] [CrossRef]
- Saariluoma, P.; Jokinen, J.P. Emotional dimensions of user experience: A user psychological analysis. Int. J. Hum.-Comput. Interact. 2014, 30, 303–320. [Google Scholar] [CrossRef]
- Sauro, J.; Lewis, J.R. Quantifying the User Experience: Practical Statistics for User Research; Morgan Kaufmann: San Francisco, CA, USA, 2016. [Google Scholar]
- Beales, H. The Value of Behavioral Targeting; Network Advertising Initiative: Washington, DC, USA, 2010. [Google Scholar]
- Chen, Y.; Pavlov, D.; Canny, J.F. Large-scale behavioral targeting. In Proceedings of the 15th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining; Association for Computing Machinery: New York, NY, USA, 2009; pp. 209–218. [Google Scholar]
- Law, E.L.-C.; Van Schaik, P. Modelling User Experience—An Agenda for Research and Practice; Oxford University Press: Oxford, UK, 2010. [Google Scholar]
- Li, T.; Fu, Q.; Meng, F.; Wang, Z.; Wang, X. Study on Partial Least-Squares Regression Model of Simulating Freezing Depth Based on Particle Swarm Optimization. J. Comput. 2011, 6, 532–539. [Google Scholar] [CrossRef]
- Svezhenov, Y.; Nikov, A.; Dimitrov, L. modelling in dynamic user experience design of educational toys packages. AIP Conf. Proc. 2019, 2072, 090015. [Google Scholar] [CrossRef]
- Abdi, H. Partial Least Squares Regression (PLS-Regression); Sage: Thousand Oaks, CA, USA, 2003; pp. 792–795. [Google Scholar]
- Song, W.; Xie, X.; Huang, W.; Yu, Q. The Design of Automotive Interior for Chinese Young Consumers Based on Kansei Engineering and Eye-Tracking Technology. Appl. Sci. 2023, 13, 10674. [Google Scholar] [CrossRef]
- Sawatsky, M.L.; Clyde, M.; Meek, F. Partial least squares regression in the social sciences. Quant. Methods Psychol. 2015, 11, 52–62. [Google Scholar] [CrossRef]
- Thioulouse, J.; Chessel, D.; Dole’dec, S.; Olivier, J.-M. ADE-4: A multivariate analysis and graphical display software. Stat. Comput. 1997, 7, 75–83. [Google Scholar] [CrossRef]
- Traş, F. An Exploration of User Experience on Products: Multiple Outlets as a Case of Design; Fen Bilimleri Enstitüsü: İstanbul, Turkey, 2015. [Google Scholar]
- Khalaj, J.; Pedgley, O. A semantic discontinuity detection (SDD) method for comparing designers’ product expressions with users’ product impressions. Des. Stud. 2019, 62, 36–67. [Google Scholar] [CrossRef]
- Wöbbekind, L.; Mandl, T.; Womser-Hacker, C. Construction and first testing of the UX kids questionnaire (UXKQ) a tool for measuring pupil’s user experience in interactive learning apps using semantic differentials. In Proceedings of Mensch und Computer 2021; Association for Computing Machinery: New York, NY, USA, 2021; pp. 444–455. [Google Scholar]
- Wold, H. Estimation of principal components and related models by iterative least squares. Multivar. Anal. 1966, 1, 391–420. [Google Scholar]
Figure 1.
Method steps for UX design of detergent packages.
Figure 2.
UX design recommendations for well-designed detergent package.
Table 1.
Some regression coefficients of detergent package design elements.
| Detergent Package Design Elements | Design Units | Regression Coefficients | Detergent Package Design Elements | Design Units | Regression Coefficients | ||
|---|---|---|---|---|---|---|---|
| 1. | Package Shape | Square 1 | 0.1082 | 6. | Package Width | High 1 Average 2 Narrow 3 | −0.0845 0.0275 0.0425 |
| Trapezoid 2 | −0.1142 | ||||||
| Polygon 3 | 0.0191 | ||||||
| Cylinder 4 | 0.0044 | ||||||
| Cone 5 | −0.0328 | ||||||
| 2. | Value Image | Dark 1 | −0.0456 | 7. | Image Size | Large 1 Medium 2 Small 3 | −0.0199 0.0009 0.0327 |
| Neutral 2 | −0.0842 | ||||||
| Contrast 3 | 0.0261 | ||||||
| Light 4 | 0.0782 | ||||||
| 3. | Position of Trade Name | Above 1 | −0.0065 | 8. | Package Height | High 1 Average 2 Low 3 | 0.0548 0.0326 −0.0758 |
| Below 2 | 0.0817 | ||||||
| Left 3 | 0.0442 | ||||||
| Middle 4 | −0.0343 | ||||||
| Right 5 | −0.0005 | ||||||
| Top Left 6 | −0.0122 | ||||||
| Top Right 7 | 0.0519 | ||||||
| Bottom Left 8 | 0.0064 | ||||||
| Bottom Right 9 | −0.0982 | ||||||
| 4. | Logo size | Large 1 | −0.0315 | 9. | Font Trade Name Optical Value | Bold 1 Semi-Bold 2 Light 3 | 0.0670 −0.0216 −0.0548 |
| Middle 2 | −0.0533 | ||||||
| Small 3 | 0.0857 | ||||||
| 5. | Style of Letters in Trade Name | Capitals 1 | −0.0025 | 10. | Dosimeter | Unavailable 1 Available 2 | −0.0293 0.0293 |
| Upper Lowercase 2 | −0.0456 | ||||||
| Lowercase 3 | 0.0737 | ||||||
Table 2.
Some design recommendations including the most important detergent package design elements.
Table 2.
Some design recommendations including the most important detergent package design elements.
| Detergent Package Design Elements | Range | Good Design | Bad Design |
|---|---|---|---|
| 1. Package Shape | 0.3782 | Square 1 | Trapezoid 2 |
| 2. Value Image | 0.1819 | Light 4 | Neutral 2 |
| 3. Position of Trade Name | 0.1623 | Below 2 | Bottom Right 9 |
| 4. Logo Size | 0.1502 | Small 3 | Large 1 |
| 5. Style of Letters in Trade Name | 0.1346 | Lowercase 3 | Upper Lowercase 2 |
| 6. Package Width | 0.1339 | Narrow 3 | High 1 |
| 7. Image Size | 0.1337 | Small 3 | Large 1 |
| 8. Package Height | 0.1254 | High 1 | Low 3 |
| 9. Font Trade Name Optical Value | 0.1047 | Bold 1 | Light 3 |
| 10. Dosimeter | 0.1007 | Available 2 | Unavailable 1 |
| Average of All 23 Design Elements | 0.0981 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.
Share and Cite
MDPI and ACS Style
Radoslavov, A.; Dimitrov, L.; Nikov, A. Statistical Modelling in User Experience Design of Detergent Packages. Eng. Proc. 2026, 150, 72. https://doi.org/10.3390/engproc2026150072
AMA Style
Radoslavov A, Dimitrov L, Nikov A. Statistical Modelling in User Experience Design of Detergent Packages. Engineering Proceedings. 2026; 150(1):72. https://doi.org/10.3390/engproc2026150072
Chicago/Turabian StyleRadoslavov, Alexander, Lyubomir Dimitrov, and Alexander Nikov. 2026. "Statistical Modelling in User Experience Design of Detergent Packages" Engineering Proceedings 150, no. 1: 72. https://doi.org/10.3390/engproc2026150072
APA StyleRadoslavov, A., Dimitrov, L., & Nikov, A. (2026). Statistical Modelling in User Experience Design of Detergent Packages. Engineering Proceedings, 150(1), 72. https://doi.org/10.3390/engproc2026150072
