Application of Multi-Diamond Processing Model to Hairstyling ^†

Abstract

In 2003, the British Design Council advocated for strategic design and its management. They researched ways to create a universal framework for describing the design process and came up with the Double Diamond Model, which became a part of the universal design language. The Double Diamond Model applies to innovation and problem-solving involving divergence and convergence at four stages: Discover, Define, Develop, and Deliver. Continuous divergence, convergence, and iteration are necessary at each stage of the design process. Therefore, we reinterpreted design process methods to optimize and apply the entire design process in various fields. A case study applying the Multi-Diamond Model to users’ experiences during a hairstyling appointment was conducted to understand user needs and behavioral differences. Users’ behaviors and their understanding of hairstyling were analyzed by employing the fuzzy comprehensive evaluation method to choose hairstyles that were suitable for various face shapes at different perceived difficulty levels.

1. Introduction

The Double Diamond Model was proposed by the Design Council of the United Kingdom in 2005 to understand a problem and identify potential challenges. It is a design process model consisting of four phases: Discover, Define, Develop, and Deliver [1]. Various solutions were generated during the development phases through prototyping, testing, and evaluation [2]. These phases belong to a cycle of divergence and convergence. The first diamond model defines the problem, while the second diamond model explores possibilities through iterative computation, testing, and development. Widely applied in designing products and services, the Double Diamond Model provides a structured approach to the design process [3] (Figure 1).

A three-diamond design model was also proposed as the two-diamond model is usually employed for problem discovery and problem-driven solutions in the design process. Data are used for computation, analysis, or planning to represent and quantify the problem. By incorporating data on user needs and data constraints, the third diamond model is introduced to enhance the framework of the two-diamond model in the design process. Design experiences involve diverse processes with divergences and convergences in continuous iterations. A simple two or three-diamond model cannot fully capture the complexities of a design project. Teachers often outline phased objectives in product design courses, breaking each phase down for better comprehension for students. The Double Diamond Model is applied to each design stage with dispersion, convergence, and iteration.

In the beauty industry, a user service process can include choosing a hairstyle and hair color that expresses individual personality and style. Satisfaction hinges on an individual’s expectations regarding image, style, and personality. Effective communication ensures that serviced hairstyles and colors align with client expectations.

We delved into how to select design thinking methodologies to support human-centered artificial intelligence and innovation. Using a multi-diamond design framework, we reinterpreted the Double Diamond Model as seven phases: Observation, Requirements, Market Research, Design Goals, Conceptual Solution, Conceptual Evaluation, and Solution Validation. Each phase involves divergence, convergence, and iteration. Black box system analysis, the six-whys questioning method, and a goal tree were applied in each diamond model to illustrate divergence and convergence actions. Using the user’s hair experience as a case study, customer behavior was analyzed by gender, age, and profession. Customer understanding of face shape and hair color, difficulty level, and personal evaluation were used to verify the feasibility of the model and understand the customer’s experience during the process, their comprehension of hairstyle and color, and the difficulty.

2. Theoretical Background

2.1. Maslow’s Theory of Needs

Maslow’s Theory of Needs was developed by psychologist Abraham Maslow to outline human motivations based on inherited, instinctive needs. These needs encompass physiological (air, water, food), safety (personal and property safety), social (relationships, love), respect (self-worth, achievement), and self-actualization (personal potential, dreams realization). Physiological and safety needs are basic, social and respect needs are psychological, and self-actualization needs are self-satisfaction. The theory provides a framework for understanding universal human needs and the development of higher-level emotions.

2.2. Black Box System

In the design process, various constraints make reliance on intuition and subjectivity difficult. System theory is used to recognize and analyze problems such as data collection, organization, and goal establishment in design. A system is a conceptual entity with no independent existence, focusing on understanding its characteristics of input, output, and conversion. The purpose of the conversion process is to establish connections with other elements (Figure 2). In the system, substances, energy, and information undergo orderly flow and conversion in the interactive process (Figure 3), representing the essence of black box analysis [4].

2.3. Objective Tree

An objective tree is used to set goals which are categorized into primary and secondary goals in a hierarchical structure [5]. The goals are interrelated to ensure consistency and validity. When creating an objective tree, the following points are usually included (Figure 4).

(1)

Determine the overarching goal to achieve, whether it is an organizational, team, or individual objective aligned with the long-term strategic direction.

(2)

Define key objectives by splitting the final goal into several primary objectives, which are the stages or nature of the end goal to be achieved.

(3)

Define secondary objectives to segment each primary goal into specific, quantifiable, measurable, and achievable secondary goals.

(4)

Establish a hierarchy of objectives according to the logical relationship between the primary and secondary objectives and establish a hierarchical structure of objectives, usually in the form of a tree diagram.

(5)

Determine the relevance and constraints within the objective tree and identify associations and constraints among the objectives, including sequential relationships, dependencies, or conditional factors.

Figure 4. Structure of objective tree.

Figure 4. Structure of objective tree.

2.4. Evaluation Using Fuzzy Theory

Zadeh introduced ‘fuzzy sets’ to explore quantitative processing in human cognitive processes and created a new branch of mathematics. Practical phenomena are categorized as definite, random, and fuzzy in the fuzzy set. Mathematical tools such as geometry and algebra address definite objects, probability theory deals with random phenomena, and fuzzy mathematics is used to explain fuzzy phenomena [6]. In assessing entities with multiple attributes and diverse influences, the Fuzzy Comprehensive Evaluation involves six main steps [7].

(1)

Establishment of factor sets

$$U=\left\{u_1,u_2,\dots ,u_m\right\}$$

(1)

(2)

Creating weight sets

$$\underset{˜}{A}=\left(a_1,a_2,\dots ,a_m\right)$$

(2)

(3)

Creating a rating set

$$V=\left\{V_1,V_2,\dots ,V_n\right\}$$

(3)

(4)

Single Factor Fuzzy Evaluation

$${\underset{˜}{R}}_i=({\displaystyle \frac{r_{i1}}{V_1}}+{\displaystyle \frac{r_{i2}}{V_2}}+\dots +{\displaystyle \frac{r_{in}}{V_n}})$$

(4)

$$\underset{˜}{R}=\left(\begin{array}{ccc}{r}_{11}& r_{12}& \cdots r_{1n}\\ r_{21}& r_{22}& \cdots r_{2n}\\ \cdots & & \\ r_{m1}& r_{m2}& \cdots r_{mn}\end{array}\right)$$

(5)

(5)

Fuzzy Comprehensive Evaluation

$$\underset{˜}{B}=\underset{˜}{A}\circ \underset{˜}{R}$$

(6)

$$\underset{˜}{B}=\left(a_1,a_2,\dots ,a_m\right)\circ \left(\begin{array}{ccc}{r}_{11}& r_{12}& \cdots r_{1n}\\ r_{21}& r_{22}& \cdots r_{2n}\\ \cdots & & \\ r_{m1}& r_{m2}& \cdots r_{mn}\end{array}\right)=\left(b_1,b_2,\dots ,b_n\right)$$

(7)

included among these

$$b_j=\underset{i=1}{\overset{m}{{\displaystyle \vee }}}\left(a_i\wedge r_{ij}\right)\left(j=1,2,\dots ,n\right)$$

(8)

3. Methods

For the foundational design flow, we developed the stages of dispersion, iterative computation, convergence, and delivery in design. At each stage, outputs were used as inputs. In this approach, we used a multi-diamond model and employed the methodology in Section 2 for the research, analysis, and evaluation of the model’s feasibility.

3.1. Multi-Diamond Model

The design framework of the Multi-Diamond Model comprises seven stages: observation, requirement proposal, market research, design objectives, conceptual proposals, concept evaluation, and proposal validation. In the Double Diamond Model’s design process, the discovery phase corresponds to observation, requirements, and market analysis. The definition phase aligns with design objectives derived from market analysis, while the development phase corresponds to conceptual solutions. The delivery phase aligns with evaluating conceptual solutions, involving testing and feasibility assessment. The process starts with an input point, flowing through the diamond, generating an output (Figure 5). This output becomes the input for the next stage, continuing the process within the diamond. Iteration results in a complete result denoted by a bolded hollow circle, symbolizing integrated output and input. In the figure, arrows present the potential for backtracking and iteration when facing challenges.

In the observation phase, everyday experiences or Maslow’s hierarchy of needs are presented for inspiration. The emphasis is on identifying and exploring issues and understanding stakeholder perspectives using surveys, interviews, and the 5W1H technique. Simultaneously, market analysis and target trees are used to shape the final design. Utilizing image boards, design scripts, and scenario planning, conceptual design solutions are envisioned within specific contexts. By evaluating these solutions using merits, drawbacks, and feasibility, followed by prototyping for validation, the groundwork is established for subsequent development and assessment.

3.2. Detailed Process

Using Maslow’s hierarchy of needs, life’s different facets, personal experiences, behaviors, emotions, inner feelings, and preferences are determined and documented. This process generates meaningful and interesting problem requirements for further exploration. After identifying the requirement, we employed the 5W1H method to analyze each item and presented the results in a target tree to identify design targets and requirements. Utilizing the organized design targets and potential problem requirements, we interviewed and administered a questionnaire survey on the design targets. Data were collected to understand the importance of the requirement items. This process aided in determining the design objectives. The fuzzy theory or grey theory was applied to analyze the data to explore behavioral patterns, design objectives, and demand values. After setting the design objectives, a literature review was conducted to identify relevant technologies and methods. We set design boundaries by defining the research scope for convergence. With these conditions, the conceptual design phase was initiated to generate multiple conceptual solutions. We evaluated the strengths and weaknesses of these concepts. For the evaluation, the PUGH evaluation method was used to analyze data based on weights or degrees, simplifying the process of ranking and selecting the final solution. Following the selection of the final solution, feasibility was validated. A retrospective examination of the preceding stage was performed, followed by iterative corrections for the final product.

4. Case Study

Instagram and TikTok are vital for personalized communication. Maslow’s Needs Theory and the first diamond model were used for such social media to investigate appearance anxiety regarding hairstyles and hair color.

Process

The output from the first diamond model was used to summarize problems and demands as the input for the next diamond model through divergence and convergence. The second diamond model was used to explore consumer demands, employing the 5W1H analysis. For each goal, divergence and convergence were conducted in a target tree to illustrate their relationship at each level. Systematically analyzing potential consumer needs, design objectives, and potential demand was obtained as the final output. In the second diamond model, the output comprised design objectives and problem demands. Transitioning to the third diamond model of the market survey, the input was derived from the output of the second stage. Based on the results of divergence and convergence from the second stage, potential problems were analyzed further through a questionnaire survey and interviews such as a market survey. Interviews with hairstylists revealed that customers browsed hairstylists’ portfolios on social media, using images for communication. The questionnaire was designed to understand how customers chose hairstyles and colors, identified their problems, and validated the consistency of the requirements in the second stage. The questionnaire consisted of basic demographic information, a survey on face shape and hairstyle/hair color preferences, and customers’ feelings regarding their appearance. Integrating the findings, demands, and customer behavior patterns was confirmed as the final output.

According to the survey results (Figure 6), 80% of the respondents conducted research before visiting a hair salon. The primary methods included using social media and referencing hairstylist portfolios. Approximately 20% did not perform pre-research due to a lack of ideas or a preference for direct communication with hairstylists. These findings represent the customer experience process before visiting a hair salon.

The design objective of the fourth diamond model was to confirm the ultimate purpose of the analysis by analyzing users’ understanding of their hairstyles, face shapes, and behaviors to explore their needs and entry points for hairstylists. The fifth to seventh diamond models involved a coherent operation that converted qualitative text into quantitative data. In the approach, the fuzzy logic of the fuzzy evaluation method was used to understand and analyze subjective user feedback involving qualitative concepts.

In the questionnaire, customers were asked about their understanding of their face shape, hairstyle, and hair color, as well as the difficulty in selecting a hairstyle and hair color. Scores of 0, 0.25, 0.5, 0.75, and 1 were used, ranging from “not at all understanding” to “very understanding,” and from “not at all difficult” to “very difficult.”

Table 1. Understanding face shape, hairstyle, and hair color.

	Not at All Understanding	Not Understanding	Average	Well Understanding	Very Well Understanding
Face Shape	0.15	0.15	0.26	0.43	0.11
Hair Shape	0.16	0.21	0.29	0.29	0.05
Hair Color	0.26	0.3	0.12	0.25	0.07

and

Table 2. Difficulty in choosing hairstyle and hair color.

	Very Little Difficult	Not Difficult	Average	Difficult	Very Difficult
Hair Style	0.09	0.23	0.18	0.36	0.14
Hair Color	0.1	0.27	0.11	0.33	0.19

present the percentage results of these scores.

5. Results and Discussion

The fuzzy comprehensive evaluation was used as follows.

The customer’s knowledge of face shape, hairstyle, and hair color was analyzed.

(1)

Factor set U1 = { Face Shape, Hair Style, Hair Color }

(2)

Weight set ${\underset{˜}{A}}_1=\left(0.3,0.35,0.35\right)$

(3)

Ratings V₁= { Very unaware, unaware, average, aware, very aware }

(4)

Single Factor Fuzzy Evaluation: Face Shape (x1), Hair Shape (x2), Hair Color (x3), we get

$$\begin{gathered} \underset{˜}{R}\left(x1\right)=\left(0.05 0.15 0.26 0.43 0.11\right) \\ \underset{˜}{R}\left(x2\right)=\left(0.16 0.21 0.29 0.29 0.05\right) \\ \underset{˜}{R}\left(x3\right)=\left(0.26 0.3 0.12 0.25 0.07\right) \end{gathered}$$

Therefore, the one-factor evaluation matrix is given by

$${\underset{˜}{R}}_1=\left(\begin{array}{ccc}0.05& 0.15& 0.26\\ 0.16& 0.21& 0.29\\ 0.26& 0.3& 0.12\end{array}\begin{array}{cc} 0.43& 0.11\\ 0.29& 0.05\\ 0.25& 0.07\end{array}\right)$$

(5)

Fuzzy Comprehensive Evaluation $\underset{˜}{B}=\underset{˜}{A}\circ \underset{˜}{R}$

$$\underset{˜}{B}=\left({\displaystyle \frac{0.26}{\mathrm{Very} \mathrm{unaware}}}+{\displaystyle \frac{0.21}{\mathrm{unaware}}}+{\displaystyle \frac{0.29}{\mathrm{average}}}+{\displaystyle \frac{0.3}{\mathrm{aware}}}+{\displaystyle \frac{0.11}{ \mathrm{very} \mathrm{aware}}}\right)$$

Since the sum of the values of the factors in the evaluation result exceeded 1, normalization was required.

$$\because 0.26+0.21+0.29+0.3+0.11=1.17$$

The rating result was adjusted to

$$\left(0.22 0.18 0.25 0.26 0.09\right)$$

The results of this evaluation were as follows.

• In total, 22% of people did not understand their face shape and suitable hair style and color at all.
• In total, 18% did not understand their face shape and hair color.
• In total, 25% understood their face shape and hair color.
• In total, 26% understood their face shape and hair color well.
• In total, 9% understood their face shape and hair color very well.

The customer’s difficulty in choosing their own hair style and color was as follows.

(1)

Factor set U2 = { Hair Style, Hair Color }

(2)

Weight set ${\underset{˜}{A}}_2=\left(0.6, 0.4\right)$

(3)

Rating V₂ = { Very little difficult, not difficult, average, difficult, very difficult }

(4)

Single Factor Fuzzy Evaluation: Face Shape (x1), Hair Shape (x2), Hair Color (x3), we get

(5)

$\underset{˜}{R}$ ($x_4$) = (0.09 0.23 0.18 0.36 0.14)

$\underset{˜}{R}$ ($x_5$) = (0.1 0.27 0.11 0.33 0.19)

Therefore, the one-factor evaluation matrix is given by

$${\underset{˜}{R}}_2=\left(\begin{array}{ccc}0.09& 0.23& 0.18\\ 0.1& 0.27& 0.11\end{array}\begin{array}{cc} 0.36& 0.14\\ 0.33& 0.19\end{array}\right)$$

(6)

Fuzzy Comprehensive Evaluation $\underset{˜}{B}=\underset{˜}{A}\circ \underset{˜}{R}$

$$\underset{˜}{B}=\left({\displaystyle \frac{0.1}{\mathrm{Very} \mathrm{undifficult}}}+{\displaystyle \frac{0.27}{\mathrm{not} \mathrm{difficult}}}+{\displaystyle \frac{0.18}{ \mathrm{average}}}+{\displaystyle \frac{0.36}{\mathrm{difficult}}}+{\displaystyle \frac{0.19}{\mathrm{very} \mathrm{difficult}}}\right)$$

Since the sum of the values of the factors in the evaluation result exceeded 1, normalization was required.

$$\because 0.1+0.27+0.18+0.36+0.19=1.1$$

The rating result was adjusted to

$$\left(0.09 0.25 0.16 0.33 0.17\right)$$

The results of this evaluation were as follows.

• In total, 9% felt ‘very little difficulty’ choosing their hairstyles and colors.
• In total, 25% felt ‘no difficulty’ in choosing their hairstyles and colors.
• In total, 16% felt ‘average difficulty’ in choosing their hairstyles and colors.
• In total, 33% felt ‘difficulty’ in choosing their hairstyles and colors.
• In total, 17% felt ‘very difficult’ in choosing their hairstyles and colors.

6. Conclusions

In the multi-diamond design, services, event planning, and corporate operational decision-making need to be determined. This flexible and dynamic model provides a framework for adaptability based on different activities in various domains. Divergence, convergence, and iteration are emphasized at each stage, departing from traditional approaches. This approach adds granularity to the design process, enabling designers to address challenges more precisely and adapt to rapidly changing demands. The multi-diamond design model offers a framework for design thinking and integrates divergence, convergence, and iteration. This approach is a problem-solving method for innovation. This model can guide designers to successfully solve diverse challenges in different domains. Simultaneously, it encourages the exploration of new possibilities while solving problems.