SUPPORTING THE CHOICE OF THE BEST-FIT AGILE MODEL USING FITRADEOFF

Schramm, Vanessa B.; Damasceno, Adriana C.; Schramm, Fernando

doi:10.1590/0101-7438.2023.043spe1.00264750

Vanessa B. Schramm ^**Corresponding author

Development of Systems for Supporting Sustainable Decisions (DeSiDeS), Federal University of Campina Grande (UFCG), Rua Aprígio Veloso, 882, Bloco CB (REENGE), Bairro Universitário, 58429-900 Campina Grande, PB, Brazil - E-mail: vanessa@labdesides.ufcg.edu.br

https://orcid.org/0000-0001-9276-5251

Adriana C. Damasceno

Development of Systems for Supporting Sustainable Decisions (DeSiDeS), Federal University of Campina Grande (UFCG), Rua Aprígio Veloso, 882, Bloco CB (REENGE), Bairro Universitário, 58429-900 Campina Grande, PB, Brazil - E-mail: adriana.damasceno@gmail.com

https://orcid.org/0000-0002-5252-2826

Fernando Schramm

Development of Systems for Supporting Sustainable Decisions (DeSiDeS), Federal University of Campina Grande (UFCG), Rua Aprígio Veloso, 882, Bloco CB (REENGE), Bairro Universitário, 58429-900 Campina Grande, PB, Brazil - E-mail: fernando@labdesides.ufcg.edu.br

https://orcid.org/0000-0003-3303-615X

*Corresponding author

c _i	Criterion	Description	Evaluation scale	Direction of preference
c ₁	Explicitness and Unambiguity	Is the development process defined explicitly and unambiguously?	Yes or No	“Yes” is preferable to “No” (Max)
c ₂	Rationale	Has the process been rationalized through providing extensive and precise explanations?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃	Completeness	A complete process definition includes definitions for: development cycle, lifecycle, roles, activities, modeling language, artifacts, practices/ techniques, rules, and umbrella activities.	Ratio of the number of existing definitions to the total.	A higher value is preferable to a smaller value (Max)
c ₄	Generic development lifecycle coverage	Which phases of the generic development cycle are covered by the development process? Generic phases include: inception, analysis, design, implementation, test, deployment, maintenance, support, and postmortem.	Ratio of the number of covered phases to the total number of generic phases.	A higher value is preferable to a smaller value (Max)
c ₅	Smooth transition	Is the transition between the phases smooth?	Yes or No	“Yes” is preferable to “No” (Max)
c ₆	Seamless transition	Is the transition between the phases seamlessness?	Yes or No	“Yes” is preferable to “No” (Max)
c ₇	Adequate products	Does the development process produce the products typically associated with the generic development activities (feasibility analysis, requirement specification, design, modeling, documentation, test, training and deployment)?	Ratio of product types supported to the ideal number of product types.	A higher value is preferable to a smaller value (Max)
c ₈	Modeling coverage	Do the products include models (analysis and design)?	Yes or No	“Yes” is preferable to “No” (Max)
c ₉	Consistency	This criterion evaluates the level at products complement each other.	High, Medium (overlapping exist that can result in inconsistencies) or Low	“High” is preferable to “Low” (Max)
c ₁₀	Tangibility/Visibility	This criterion evaluates the level at the products are tangible, understandable, and testable to end-users.	High, Medium or Low	“High” is preferable to “Low” (Max)
c ₁₁	Standards	Are there any specific standards for the products?	Yes or No	“Yes” is preferable to “No” (Max)
c ₁₂	Process based on functional/non-functional requirements	Is the development process based on requirements?	Yes or No	“Yes” is preferable to “No” (Max)
c ₁₃	Non-functional requirements verification	Are the non-functional requirements addressed?	Yes or No	“Yes” is preferable to “No” (Max)
c ₁₄	Traceability	Can the products be traced to the requirements?	Yes or No	“Yes” is preferable to “No” (Max)
c ₁₅	Requirements change	Does the development process let changes in requirements?	Yes or No	“Yes” is preferable to “No” (Max)
c ₁₆	Available and published documents	Is the development process published and available to users?	Published and available (2), Published but not available (1); No published and not available (0)	A higher value is preferable to a smaller value (Max)
c ₁₇	Process enactment documentation	Is the running process documented?	Yes or No	“Yes” is preferable to “No” (Max)
c ₁₈	Size/Complexity	Size/complexity is defined as a function of building blocks of the development process.	Total number of practices, roles, products, and phases/ stages.	A lower value is preferable to a higher value (Min)
c ₁₉	Practicality	This criterion evaluates the level at the process is practical, based on issues affecting practicality, such as support for umbrella activities, independence from specific tools, pragmatic techniques, concrete rules, and non-overlapping activities.	High, Medium or Low	“High” is preferable to “Low” (Max)
c ₂₀	Practicability	This criterion evaluates the level at process development is practicable, through providing effective activities or techniques such as suitability filters and instantiation/ adaptation methods.	High, Medium or Low	“High” is preferable to “Low” (Max)

c _i	Criterion	Description	Evaluation scale	Direction of preference
c ₂₁	Speed	How quickly does the methodology produce results?	1/[iteration length (in days) + deployment interval (in days)]	A higher value is preferable to a smaller value (Max)
c ₂₂	Sustainability	Are speed and quality maintained until the end? Are they controlled or monitored?	Yes or No	“Yes” is preferable to “No” (Max)
c ₂₃	Flexibility	Are expected/ unexpected changes captured and handled in the project?	Ratio of the number of supporting activities and practices to the total	A higher value is preferable to a smaller value (Max)
c ₂₄	Learning	Does the process ”learn” from past projects and previous iterations?	Ratio of the number of supporting activities and practices to the total	A higher value is preferable to a smaller value (Max)
c ₂₅	Responsiveness	Does the method provide feedback?	Ratio of the number of supporting activities and practices to the total	A higher value is preferable to a smaller value (Max)
c ₂₆	Leanness	Does the method value shorter time spans, using economical and simple quality-assured means for production?	Ratio of the number of supporting activities and practices to the total	A higher value is preferable to a smaller value (Max)

c _i	Criterion	Description	Evaluation scale	Direction of preference
c ₂₇	Computational complexity	This criterion evaluates the level at computational complexity.	High (scientific and complex), Medium (business-oriented) or Low (simple and personal usage).	“Low” is preferable to “High” (Min)
c ₂₈	Project management	Support for development process management; including planning, scheduling, controlling and monitoring, and process review.	Ratio of the number of covered activities to the total.	A higher value is preferable to a smaller value (Max)
c ₂₉	Team management	Does the methodology provide any process for team and people management?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₀	Quality assurance	Does the methodology support for quality assurance techniques such as technical review, continuous verification and validation, and strategies/ techniques enhancing requirements traceability?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₁	Risk management	Does the methodology support for risk management techniques such as feasibility analysis, risk-based planning, active user involvement, continuous verification and validation, iterative process/ product/ plan reviews, and continuous integration?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₂	Adaptation and customization	Does the methodology provide methods for customizing it based on the parameters of the project at hand?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₃	Flexibility	Does the methodology allow the process and modeling language to be changed during its execution?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₄	Scalability	Is the methodology suitable for projects with different sizes, criticalities and complexities?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₅	Extensibility	Does the methodology provide any extension points?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₆	Tutorials and training	Are tutorials and training documents available?	Yes or No	“Yes” is preferable to “No” (Max)
c ₃₇	Empirical evidence	Does empirical evidence exist?	Yes or No	“Yes” is preferable to “No” (Max)

c _i	Criterion	Description	Evaluation scale	Direction of preference
c ₃₈	Constraints	Any general constraints in the methodology that influences practicality?	Yes or No	“No” is preferable to “Yes” (Min)

Criterion (c _i )	Agile Models
	DSDM	SCRUM	XP2	Crystal
c ₁	No	No	Yes	Yes
c ₂	No	No	No	Yes
c ₃	6/9	5/9	5/9	6/9
c ₄	6/9	4/9	8/9	7/9
c ₅	Yes	Yes	Yes	Yes
c ₆	No	No	No	No
c ₇	4/8	4/8	7/8	8/8
c ₈	No	No	Yes	No
c ₉	Medium	Medium	High	High
c ₁₀	Medium	High	High	Medium
c ₁₁	Yes	No	Yes	Yes
c ₁₂	Yes	Yes	Yes	No
c ₁₃	Yes	Yes	Yes	Yes
c ₁₄	Yes	Yes	Yes	No
c ₁₅	Yes	Yes	Yes	Yes
c ₁₆	2	2	2	2
c ₁₇	Yes	Yes	Yes	Yes
c ₁₈	30	23	40	46
c ₁₉	Medium	Medium	High	Medium
c ₂₀	Low	Low	High	High
c ₂₁	1/14	1/14	1/14	1/56
c ₂₂	Yes	Yes	Yes	Yes
c ₂₃	15/18	8/8	5/9	2/8
c ₂₄	16/18	7/8	3/9	3/8
c ₂₅	15/18	8/8	7/9	4/8
c ₂₆	6/18	8/8	5/9	4/8
c ₂₇	Medium	High	Medium	High
c ₂₈	3/4	4/4	3/4	4/4
c ₂₉	Yes	Yes	Yes	Yes
c ₃₀	Yes	Yes	Yes	Yes
c ₃₁	Yes	Yes	Yes	Yes
c ₃₂	No	Yes	No	Yes
c ₃₃	No	Yes	No	No
c ₃₄	No	Yes	Yes	Yes
c ₃₅	No	No	No	No
c ₃₆	Yes	Yes	Yes	Yes
c ₃₇	Yes	Yes	No	Yes
c ₃₈	Yes	No	No	No

Rank j	1	2	3	4	5	6	7	8	9	10	11	12
Criterion (c _i )	c ₂₄	c ₄	c ₁₈	c ₂₆	c ₂₃	c ₇	c ₂₅	c ₃	c ₂₈	c ₂₁	c ₂₇	c ₉
					c ₃₃
Rank j	13	14	15	16	17	18	19	20	21	22	23	-
Criterion (c _i )	c ₁₀	c ₁₉	c ₁	c ₃₂	c ₈	c ₁₄	c ₂	c ₁₂	c ₃₈	c ₃₄	c ₃₇	-
		c ₂₀						c ₁₁

Cycles	Consequence A (x _i )	Consequence B: Best of	Decision maker’s choice	Potentially optimal alternatives
1	x₂₄ = 0.61	c ₃₇	B	DSDM, SCRUM, XP2
2	x₂₄ = 0.75	c ₄	B	SCRUM, XP2
3	x₂₄ = 0.82	c ₄	B	SCRUM, XP2
4	x₄ = 0.778	c ₁₈	B	SCRUM, XP2
5	x₄ = 0.834	c ₁₈	B	SCRUM, XP2
6	x₁₈ = 28.75	c ₂₆	B	SCRUM, XP2
7	x₁₈ = 25.875	c ₂₆	B	SCRUM, XP2
8	x₂₆ = 0.833	c ₂₃	I	SCRUM

Cycles	Consequence A (x_i)	Consequence B: Best of	Decision maker’s choice	Potentially optimal alternatives
1	x₂₄ = 0.61	c₃₇ = 1	A	DSDM, SCRUM, XP2
2	x₂₄ = 0.61	c₄ = 0.89	A	DSDM, SCRUM, XP2
3	x₄ = 0.665	c₁₈ = 23	B	DSDM, SCRUM
4	x₁₈ = 34.5	c₂₆ = 1	B	DSDM, SCRUM
5	x₂₆ = 0.665	c₂₃ = 1	B	DSDM, SCRUM
6	x₇ = 0.75	c₂₅ = 1	I	DSDM, SCRUM
7	x₂₅ = 0.75	c₃ = 0.75	I	DSDM, SCRUM
8	x₃ = 0.69	c₂₈ = 1	B	DSDM, SCRUM
9	x₂₈ = 0875	c₂₁ = 0.714	A	DSDM, SCRUM
10	x₂₁ = 0.045	c₂₇ = 1	B	DSDM, SCRUM
11	x₂₇ = 2	c₉ = 3	B	DSDM, SCRUM
12	x₉ = 2	c₁₀ = 3	A	DSDM, SCRUM
13	x₁₀ = 2	c₁₉ = 3	I	DSDM, SCRUM
14	x₁₉ = 2	c₁ = 1	I	DSDM, SCRUM
15	x₂₄ = 0.47	c₄ = 0.89	I	DSDM, SCRUM
16	x₄ = 0.778	c₁₈ = 23	B	SCRUM

\underset{k_{1}, k_{2}, \dots, k_{i}}{m a x} \sum_{i = 1}^{n} k_{i} v_{i} (x_{i j}), j = 1, 2, \dots, m s . t . \sum_{i = 1}^{n} k_{i} v_{i} (x_{i j}) \geq \sum_{i = 1}^{n} k_{i} v_{i} (x_{i z}), z = 1, 2, \dots m, z \neq j k_{i + 1} \leq k_{i} v_{i} (x_{i}^{'}) - ϵ \ for i = 1 to n - 1 k_{i + 1} \leq k_{i} v_{i} (x_{i}^{''}) - ϵ for i = 1 to n - 1 \sum_{i = 1}^{n} k_{i} = 1 k_{i} \geq 0, i = 1, 2, \dots, n

[1] *Corresponding author

Brasil

Brasil

SUPPORTING THE CHOICE OF THE BEST-FIT AGILE MODEL USING FITRADEOFF

ABSTRACT