Towards an Analyzability Model for Hybrid Software

Ana D

ıaz-Mu

noz

1,2 a

, Jos

e A. Cruz-Lemus

2 b

, Mois

es Rodr

ıguez

2 c

and Teresa Baldasarre

3 d

AQCLab Software Quality, Ciudad Real, Spain

University of Castilla-La Mancha, Ciudad Real, Spain

University of Bari Aldo Moro, Bari, Italy

Keywords:

Software Quality Model, Analyzability, Hybrid Systems, ISO/IEC 25010, Quantum Circuits, Quantum

Metrics, Empirical Study.

Abstract:

This paper presents an initial validation of a software quality model focused on analyzability, aligned with the

ISO/IEC 25010 standard. The model targets hybrid systems that integrate classical and quantum components,

combining established classical metrics with quantum-speciﬁc measures designed to capture the complexity of

quantum circuits. In this ﬁrst empirical study, we evaluate only the quantum dimension of the model through a

quasi-experimental setup involving computer engineering students. The results show that the model’s analyz-

ability levels correlate with participants’ comprehension performance, supporting its utility in distinguishing

circuit complexity. These ﬁndings offer promising evidence for further model reﬁnement and lay the ground-

work for future evaluations involving real-world hybrid code bases.

1 INTRODUCTION

Quantum computing has emerged as a promising so-

lution to problems intractable for classical comput-

ers, offering signiﬁcantly enhanced processing ca-

pabilities (Bernhardt, 2019)(Piattini et al., 2021).

However, current approaches do not entirely replace

classical computing; instead, they aim to integrate

both paradigms into hybrid systems that combine the

strengths of classical and quantum processing.

As these systems grow in complexity, ensur-

ing their quality becomes critical. Software qual-

ity is a key factor in their development and adoption

(Rodr

ıguez et al., 2015), as issues related to maintain-

ability and usability may hinder industrial implemen-

tation (Rodr

ıguez et al., 2016). The ISO/IEC 25010

standard (ISO/IEC, 2011) provides a structured ap-

proach to software quality, with analyzability (a key

sub-characteristic of maintainability) being particu-

larly relevant due to the interdisciplinary nature and

dual architecture of hybrid systems.

Although established models exist for classi-

cal software (Piattini et al., 2020) (Verdugo et al.,

https://orcid.org/0000-0001-6515-8835

https://orcid.org/0000-0002-0470-609X

https://orcid.org/0000-0003-2155-7409

https://orcid.org/0000-0001-8589-2850

2024), there are still signiﬁcant gaps in assessing the

quality of systems that integrate both classical and

quantum components. Recent work has begun ad-

dressing this by proposing hybrid models that com-

bine classical metrics—such as cyclomatic complex-

ity—with quantum-speciﬁc ones, like quantum cy-

clomatic complexity and circuit depth (D

ıaz-Mu

noz

et al., 2024a)(D

ıaz-Mu

noz et al., 2024b). These ap-

proaches aim to expand our understanding of hybrid

software quality and promote its adoption in industrial

contexts.

Nevertheless, the practical application of such

models remains limited, and empirical validations are

scarce. In particular, no consolidated methodology

has yet emerged to operationalize integrating classi-

cal and quantum aspects within a uniﬁed and usable

quality model.

This work represents a ﬁrst step toward validat-

ing and reﬁning the quantum part of a previously

developed hybrid quality model (D

ıaz-Mu

noz et al.,

2024a)(D

ıaz-Mu

noz et al., 2024b), which integrates

classical and quantum metrics to assess analyzabil-

ity and maintainability in hybrid systems. However,

this initial empirical validation focuses exclusively on

the quantum dimension of the model due to the lack

of publicly available, integrated hybrid code bases.

Through a quasi-experimental study with computer

engineering students, we examine whether the ana-

Díaz-Muñoz, A., Cruz-Lemus, J. A., Rodríguez, M. and Baldasarre, T.

Towards an Analyzability Model for Hybrid Software.

DOI: 10.5220/0013533800004525

In Proceedings of the 1st International Conference on Quantum Software (IQSOFT 2025), pages 97-104

ISBN: 978-989-758-761-0

lyzability levels indicated by the quantum metrics of

the model align with participants’ subjective under-

standing of quantum circuits. The insights gained

from this study will contribute to reﬁning the model

and guiding future research toward a comprehensive

framework for evaluating the quality of hybrid soft-

ware systems.

2 RELATED WORK

The ISO/IEC 25010 quality model (ISO/IEC, 2011),

part of the ISO/IEC 25000 series (SQuaRE)

(ISO/IEC, 2014), establishes a structured framework

for assessing the quality of software and systems.

This standard deﬁnes eight fundamental quality char-

acteristics encompassing various software evaluation

dimensions: functionality, reliability, usability, efﬁ-

ciency, maintainability, portability, compatibility, and

security. These characteristics are further divided into

sub-characteristics, enabling a more detailed and spe-

ciﬁc assessment.

Among these, analyzability is a crucial sub-

characteristic of maintainability. It evaluates how eas-

ily software can be understood, diagnosed, and mod-

iﬁed, ensuring its long-term evolution without signif-

icant obstacles. In essence, analyzability determines

how effectively the behavior and structure of software

can be identiﬁed and comprehended, facilitating error

detection, quality enhancements, and the integration

of new functionalities.

In classical software, analyzability-related proper-

ties are essential for ensuring systems remain under-

standable, maintainable, and easy to modify. These

properties focus on how software design and code or-

ganization impact developers’ ability to interpret and

alter systems efﬁciently. Table 1 presents the primary

metrics used to evaluate these aspects (Rodr

ıguez and

Piattini, 2014).

These analyzability properties have proven effec-

tive in evaluating software quality in industrial envi-

ronments. They are widely applied to identify prob-

lematic code areas and enhance maintainability and

comprehension. The study in (Verdugo et al., 2024)

provides an in-depth analysis of software evaluation

and certiﬁcation practices at the AQCLab laboratory

over the past 25 years. It highlights the practical

use of classical quality metrics, such as analyzabil-

ity, demonstrating their effectiveness in real-world in-

dustrial applications and emphasizing their continued

relevance.

However, hybrid systems introduce new chal-

http://www.aqclab.es

lenges due to integrating quantum components.

The additional complexity of quantum comput-

ing —such as non-determinism, entanglement, and

circuit depth— signiﬁcantly impacts analyzability.

These characteristics require new metrics to evaluate

quantum circuits’ complexity, understandability, and

maintainability.

Several authors have proposed metrics for quan-

tum software systems. For instance, Kumar (Ku-

mar, 2023) formalized structural coverage criteria

for quantum software testing. Other works have

discussed metrics such as quantum circuit depth,

gate complexity, and quantum cyclomatic complexity

ıaz-Mu

noz et al., 2024b). However, these contribu-

tions focus exclusively on quantum systems and lack

integration into broader software quality models.

Only a few studies have attempted to build uni-

ﬁed models integrating classical and quantum per-

spectives. Our previous work (D

ıaz-Mu

noz et al.,

2024a) proposed a hybrid model that combines classi-

cal analyzability metrics with quantum-speciﬁc indi-

cators. However, there is still a lack of systematic val-

idation of the quantum part of such models, primarily

through empirical studies in real environments.

This paper aims to ﬁll that gap by empirically ex-

ploring the feasibility of applying the quantum prop-

erties of a hybrid quality model to evaluate the an-

alyzability of software systems with quantum com-

ponents. By doing so, we contribute to developing

a structured and integrated framework for assessing

quantum and hybrid software quality, which remains

largely unexplored in current literature.

3 EXPERIMENTAL ANALYSIS

A controlled experiment was conducted to explore the

feasibility of applying the proposed model in assess-

ing the analyzability of quantum circuits. Although

the model was originally conceived as hybrid —com-

bining classical and quantum metrics— this ﬁrst vali-

dation focuses exclusively on the quantum dimension.

This focus is not only due to the lack of publicly avail-

able real-world hybrid code bases but also because the

quantum part of the model had not yet been empir-

ically validated on its own. Speciﬁcally, the study

examined the relationship between the analyzability

level of a set of quantum circuits and participants’ per-

formance in solving related tasks.

The experiment cannot be strictly controlled, as

the sample was selected conveniently. Nonetheless,

participants were randomly assigned to different treat-

ments to minimize internal bias and enhance the va-

lidity of the results. Ethical considerations were also

IQSOFT 2025 - 1st International Conference on Quantum Software

Table 1: Classical metrics.

Metric Description

Coding rules Predeﬁned coding guidelines that enhance code readability and comprehension.

These include naming conventions, structural organization, and design princi-

ples. Adhering to these rules improves code clarity and reduces errors.

Code documentation Essential for explicitly describing a program’s functionality, objectives, and

behavior. Inline comments, supplementary documents (e.g., README ﬁles),

and design speciﬁcations enhance code comprehensibility, facilitating modi-

ﬁcations by other developers.

Cyclomatic complexity Introduced by McCabe, this metric quantiﬁes the complexity of a program’s

control ﬂow by counting the number of independent execution paths. Higher

values indicate increased complexity, making code harder to analyze and mod-

ify. Reducing cyclomatic complexity improves software comprehension.

Package structuring The organization of modules and packages directly inﬂuences analyzability. A

well-structured system allows developers to identify relevant components with-

out grasping the code base. Hierarchical organization and modularity enhance

clarity and maintainability.

Class structuring In object-oriented programming, clear and well-deﬁned class responsibilities

facilitate analyzability. The Single Responsibility Principle (SRP) and low-

class coupling improve software organization and easier maintenance.

Method size Excessively long methods complicate comprehension. Keeping methods con-

cise and focused on a single functionality enhances readability. Best practices

suggest limiting method length to 10–15 lines to simplify analysis and modiﬁ-

cations.

Duplicate code Code redundancy increases system complexity and complicates maintenance,

as changes must be applied in multiple locations. Eliminating duplicate code

through refactoring techniques signiﬁcantly enhances analyzability and main-

tainability.

addressed: all participants took part voluntarily and

signed informed consent forms by institutional guide-

lines.

In what follows, we provide a detailed description

of the study’s components.

3.1 Analyzability Model

This section introduces the proposed model as a pre-

liminary approach for exploring the analyzability of

hybrid systems. Rather than presenting a deﬁnitive

assessment framework, this study applies the quan-

tum part of the model to investigate its feasibility,

identify potential reﬁnements, and contribute to its it-

erative development.

The classical properties included in the model

align with those outlined in the previous section

regarding software analyzability. Additionally, the

model incorporates speciﬁc quantum metrics to as-

sess quantum circuits, aiming to capture key complex-

ity and understandability factors introduced by quan-

tum computing. These adapted quantum metrics are

shown in Table 2.

It is important to note that some quantum prop-

erties in the model are derived from classical soft-

ware metrics, such as quantum cyclomatic complex-

ity, which adapts the classical deﬁnition to the quan-

tum context. However, other metrics are newly de-

ﬁned speciﬁcally for quantum systems, such as cir-

cuit width, circuit depth, and auxiliary qubit usage, as

these concepts have no direct equivalent in classical

software due to the fundamentally different nature of

quantum computation.

3.2 Relevant Variables

The experimental design incorporates the variables

and inﬂuencing factors outlined below.

The independent variable in the experiment is

the analyzability level of the quantum circuits, cate-

gorized into three levels:

• Low Analyzability: A circuit with a highly com-

plex structure characterized by deep circuit depth,

a high density of operations, and excessive use of

auxiliary qubits, making it difﬁcult to understand.

This level is rated 1/5 according to the model.

• Medium Analyzability: A circuit of moderate

complexity, exhibiting a balanced structure with a

reasonable number of operations and qubits, rated

Towards an Analyzability Model for Hybrid Software

Table 2: Quantum metrics.

Metric Description

Circuit width This metric quantiﬁes the number of qubits utilized in a quantum circuit at any

moment. It reﬂects the quantum resources required to execute an algorithm.

A larger circuit width generally increases complexity, making the circuit more

challenging to comprehend and debug.

Circuit depth This refers to the number of sequential layers of quantum gates applied to qubits

before obtaining a result. Greater depth typically implies higher complexity,

making the circuit more difﬁcult to analyze and optimize. Additionally, deeper

circuits have longer execution times and are more susceptible to errors due to

qubit interference.

Gate complexity This metric evaluates the overall complexity of the quantum gates used in a

circuit. It considers both the type and number of gates applied. While some

gates, such as Hadamard or phase gates, are relatively simple, others, like swap

or controlled gates, introduce greater complexity. Many complex gates can

make a circuit more difﬁcult to interpret and analyze.

Conditional instructions These are operations where a quantum gate’s execution depends on the state

of a control qubit. While they enable decision-making based on quantum in-

formation, they also increase circuit complexity. Many conditional instructions

can make the circuit more intricate and harder to understand, as execution paths

may vary dynamically.

Quantum cyclomatic com-

plexity

Adapted from the classical software metric, this evaluates the complexity of a

quantum circuit based on the number of independent execution paths it contains

(Kumar, 2023). Higher values indicate a more complex structure, making the

circuit harder to analyze and debug.

Measurement operations These operations extract results from qubits and are fundamental in quantum

computing. This metric counts how frequently measurements occur within a

circuit. Many measurements, particularly those dependent on complex condi-

tions, may indicate increased difﬁculty understanding the circuit’s behavior.

Initialization and reset op-

erations

These refer to operations that either prepare qubits in a speciﬁc state before ex-

ecution or reset them afterward. While essential for maintaining computational

coherence, excessive or complex use of these operations can contribute to over-

all circuit complexity, complicating analysis.

Auxiliary qubits Additional qubits used to facilitate complex quantum operations without inter-

fering with the main computation. A higher number of auxiliary qubits can

indicate increased circuit complexity, introducing more dependencies that must

be managed and understood.

3/5 according to the model.

• High Analyzability: A well-structured, opti-

mized circuit designed to minimize complexity

and enhance clarity, making it easier to compre-

hend. This level is rated 5/5 according to the

model.

The dependent variable is the average score ob-

tained for each circuit. This metric quantiﬁes partic-

ipants’ ability to read, interpret, and analyze the cir-

cuit, reﬂecting their capacity to understand its behav-

ior and the ﬁnal states of the qubits.

Additionally, several external factors that may

impact the results were considered, including:

• Prior Experience in Quantum Programming:

Participants with greater exposure to quantum

programming are expected to demonstrate a

higher ability to analyze complex circuits.

• Background in Classical Programming: Fa-

miliarity with traditional programming paradigms

may inﬂuence participants’ speed and accuracy in

understanding quantum circuits.

• Educational Background: A higher level of edu-

cation is assumed to correlate with improved com-

prehension skills.

Another potential confounding factor is the dif-

ference in abstraction between classical and quantum

programming. Unlike classical code, quantum cir-

cuits rely on probabilistic logic and visual-spatial rea-

soning, which may affect participants’ understanding

regardless of their general programming experience.

IQSOFT 2025 - 1st International Conference on Quantum Software

100

3.3 Research Assumptions

The primary aim of this experiment is to examine the

relationship between the analyzability level of quan-

tum circuits, as assessed using the proposed model,

and the performance outcomes of the participants.

The following hypotheses were formulated:

Null Hypothesis (H

): There is no statistically

signiﬁcant correlation between the analyzability level

of the quantum circuits and the average score obtained

by the participants. In other words, the analyzability

level of the circuit does not inﬂuence the participants’

performance in understanding and analyzing the cir-

cuit.

Alternative Hypothesis (H

): This hypothesis

contradicts the null hypothesis, asserting that there is

a statistically signiﬁcant relationship between the an-

alyzability level and the participants’ performance.

These hypotheses aim to evaluate whether the

model’s analyzability rating corresponds with mea-

surable differences in participants’ ability to under-

stand and evaluate the circuits. Rejecting the null hy-

pothesis would indicate the model effectively captures

meaningful complexity differences among quantum

circuits.

3.4 Subjects

The experiment involved 109 computer engineering

students from Aldo Moro University in Bari, Italy.

Tables 3 and 4 visually depict the distribution of par-

ticipants across various factors, such as age, gender,

and educational background. Most participants are

between 18 and 21 years old, with a higher proportion

of male participants. Regarding educational back-

ground, many participants pursue their undergraduate

degree after completing secondary education, while a

smaller percentage report holding advanced degrees,

such as a master’s.

Table 3: Number of participants classiﬁed by age group and

gender.

Age group Male Female Others

18-21 years old 75 8 1

22-25 years old 19 1 0

26-30 years old 4 0 0

31-40 years old 1 0 0

Data analysis reveals a relatively homogeneous

sample of gender, age, and educational background.

In terms of gender, males represent the majority of

participants, comprising 79.57% of the sample, while

females make up 18.28%, and a small percentage

identify with a different gender. In terms of age,

Table 4: Number of participants classiﬁed by education

level and gender.

Education level Male Female Others

High school 90 8 1

Bachelor’s degree 8 1 0

Master’s degree 1 0 0

75.27% of participants are between 18 and 21 years

old, with signiﬁcantly smaller proportions in the 22-

25 and 26-30 age groups, representing 18.28% and

6.45%, respectively. As for educational background,

the majority of participants (90.32%) have completed

secondary education (High School), with 10.75%

holding a university degree and 2.15% possessing a

master’s degree.

3.5 Earlier Preparation

Before the experiment, participants underwent intro-

ductory training covering the basics of quantum com-

puting and quantum circuit design over several ses-

sions. This training aimed to equip participants with

the necessary knowledge to understand the concepts

involved in the experimental tasks. Topics included

fundamental principles of quantum computing, such

as qubit usage and operation, quantum gates, and the

measurement of quantum states. Additionally, partic-

ipants were introduced to practical examples of con-

structing and simulating quantum circuits using tools

like Qiskit and Quirk, which would be employed dur-

ing the experiment.

To further support the participants’ preparation,

guided exercises reﬂecting the structure and tasks they

would encounter during the experiment were also

conducted, albeit with reduced difﬁculty. This ap-

proach was intended to ensure participants became

sufﬁciently familiar with the core concepts and tools,

thereby minimizing potential biases arising from a

lack of prior knowledge. Despite this preparation, it

must be acknowledged that participants’ actual exper-

tise levels were not formally assessed before the ex-

periment. Therefore, variability in prior experience

may inﬂuence the results and is considered a poten-

tial threat to internal validity.

3.6 Data Collection Approach

A detailed procedure was established for data col-

lection, ensuring the random assignment of exercises

and the traceability of participant responses. For this

purpose, each participant received a personalized PDF

document via email. This document, written as an in-

dividualized letter, contained three links to randomly

assigned exercises to mitigate biases related to fatigue

Towards an Analyzability Model for Hybrid Software

101

and learning effects.

Each exercise included two key components:

• Qiskit Code: Each exercise provided the code

in the Qiskit quantum programming language,

which participants were to use as a reference for

analyzing the circuit’s behavior.

• Quirk Simulator: As part of the task, partici-

pants were required to design the corresponding

circuits for the Qiskit code they received in the

Quirk simulator

, an interactive tool that enables

real-time visualization of qubit evolution.

The results of the exercises were gathered using

Google Forms, designed to assess participants’ abil-

ity to understand and analyze quantum circuits. The

goal of the exercises was for participants to analyze

the provided circuit and determine the ﬁnal states of

the qubits after its execution. These results would be

entered into the forms as responses to speciﬁc ques-

tions.

The exercises were designed to assess partici-

pants’ ability to determine the ﬁnal quantum state of

each qubit after circuit execution. Each task consisted

of eight multiple-choice and short-answer questions,

targeting core aspects of comprehension such as gate

behavior, measurement interpretation, and identiﬁca-

tion of quantum interference. Each correct answer

contributed one point to the total circuit score, with

a maximum of eight points per circuit. After the data

collection phase, a manual review was conducted to

discard incomplete or invalid responses. The cleaned

dataset was exported in CSV format and processed in

Excel and Python for statistical analysis.

4 ANALYSIS AND REFLECTION

ON RESULTS

This section provides an in-depth discussion of the

experiment’s results, aiming to place the ﬁndings in

context and analyze their signiﬁcance concerning the

analyzability model for quantum circuits. The results

are interpreted to assess the validity of the proposed

hypotheses and to reﬂect on the implications and lim-

itations of this initial validation. Additionally, the po-

tential inﬂuence of these ﬁndings on future research

and practical applications in the quantum computing

ﬁeld is discussed.

https://algassert.com/quirk

4.1 Analysis of Data Distribution

A descriptive analysis was ﬁrst performed on the

scores obtained by participants, considering various

demographic variables. This analysis helped identify

patterns and preliminary differences relevant to inter-

preting the results of the statistical tests.

Using violin plots (Figure 1), the distribution of

scores was visualized across different analyzability

levels and demographic subgroups. The X-axis rep-

resents the analyzability level of the circuit —“high”

(green), “medium” (orange), and “low” (dark blue)—

and the Y-axis ranges from 0 to 8, representing the

number of correct responses out of eight questions per

circuit.

Next, an interpretation of the various results ob-

tained is provided:

• Score Dispersion. The width of the violin in-

dicates score variability. Educational level, for

example, shows a larger dispersion, possibly due

to differences in academic preparation or famil-

iarity with abstract reasoning. In contrast, gen-

der and order of exercises show narrower distri-

butions, suggesting more consistent performance

among those groups.

• Score Distribution. Most participants, espe-

cially in younger and less-experienced subgroups,

scored higher when analyzing highly analyzable

circuits. This suggests that improved circuit struc-

ture beneﬁts a wide range of users, reinforcing the

utility of the analyzability rating in distinguishing

complexity levels.

• Comparison of Average Scores. The mean

scores (white circles) conﬁrm a clear positive

trend between analyzability level and participant

performance.

• Comparison Between Analyzability Levels. For

nearly all demographic groups, scores increased

with circuit analyzability. This reinforces the hy-

pothesis that more analyzable circuits facilitate

understanding. However, the relative ﬂatness ob-

served in some subgroups (e.g., educational level)

suggests that factors such as cognitive style, ab-

straction ability, or prior training may also play an

important role.

4.2 Evaluation of Assumptions

To statistically evaluate the relationship between cir-

cuit analyzability and participant performance, the

Kruskal-Wallis test (Kruskal and Wallis, 1952) was

applied. This non-parametric method is appropri-

ate for comparing three or more independent groups

IQSOFT 2025 - 1st International Conference on Quantum Software

102

(a) Age

(b) Gender

(d) Classical Computing Experience

(e) Quantum Computing Experience

(f) Order of Exercises

Figure 1: Descriptive Analysis of the Data.

when normal distribution cannot be assumed, as in

this study.

The analysis, conducted with Python’s

scipy.stats library, yielded a test statistic of

58.3928 and a p-value <0.001. This result is well

below the 0.05 threshold, so the null hypothesis

) can be rejected. Therefore, it can be concluded

that the differences between groups are statistically

signiﬁcant.

This supports the interpretation that the analyz-

ability levels assigned to the circuits using the model

correlate with measurable differences in comprehen-

sion performance. However, it is essential to empha-

size that this validation applies only to the quantum

portion of the model and within the study sample and

environment constraints. Further testing must assess

whether similar results would be obtained with hybrid

(classical–quantum) code in real-world scenarios.

4.3 Limitations and Impact on Validity

Despite promising results, this study has several lim-

itations. The sample was limited to undergraduate

students from a single institution, reducing diversity

in academic background and technical experience.

While participants received training, their quantum

and classical programming proﬁciency was not for-

mally assessed. Moreover, although the model tar-

gets hybrid systems, this evaluation focused only on

quantum circuits due to the lack of publicly available

hybrid code bases. Additional confounding factors

—such as differences in reasoning style, cognitive

load, or familiarity with graphical tools like Quirk—

may also have inﬂuenced performance.

Future work should involve a more diverse, inter-

national sample with varying levels of quantum expe-

rience, ideally classiﬁed through pre-tests. The model

must also be validated on hybrid systems integrat-

Towards an Analyzability Model for Hybrid Software

103

ing classical and quantum components. Compara-

tive studies across tools and environments will fur-

ther support the generalization and scalability of the

model. These steps are key to reﬁning its applicabil-

ity in real-world development and quality assurance

settings.

5 CONCLUSIONS AND FUTURE

WORK

This paper presented a preliminary validation of a

quality model for assessing the analyzability of classi-

cal–quantum software. The model integrates classical

and quantum metrics within a uniﬁed framework. Al-

though hybrid in design, this ﬁrst evaluation focused

solely on quantum circuits due to the lack of accessi-

ble, mature hybrid code bases.

An empirical study with 109 participants showed

statistically signiﬁcant differences in comprehension

performance across analyzability levels assigned by

the model. These results support the model’s utility

in distinguishing circuit complexity and reinforce the

need for structured metrics in hybrid software quality

evaluation.

However, this study is an initial step. Its scope

was limited to a homogeneous participant sample and

quantum-only code. Future work will expand the

study to include diverse proﬁles, assess prior exper-

tise, and test the model on real hybrid systems. Fur-

ther iterations will reﬁne the metrics and assess their

scalability across tools and contexts, moving toward

a generalizable hybrid software quality assessment

framework.

ACKNOWLEDGEMENTS

This research was supported by the projects

QSERV: Quantum Service Engineering: Devel-

opment Quality, Testing & Security of Quantum

Microservices (PID2021-124054OB-C32), funded

by the Spanish Ministry of Science and Innova-

tion and ERDF; Q2SM: Quality Quantum Soft-

ware Model (13/22/IN/032) project ﬁnanced by

the Junta de Comunidades de Castilla-La Man-

cha and FEDER funds; and AETHER-UCLM: A

holistic approach to smart data for context-guided

data analysis (PID2020-112540RB-C42) funded by

MCIN/AEI/10.13039/501100011033/. It also re-

ceived ﬁnancial support for the execution of applied

research projects within the UNION - UCLM Own

Research Plan framework, co-ﬁnanced at 85% by

the European Regional Development Fund (FEDER)

(2022-GRIN-34110).

REFERENCES

Bernhardt, C. (2019). Quantum Computing for Everyone.

MIT Press.

ıaz-Mu

noz, A., Rodr

ıguez, M., and Piattini, M. (2024a).

Implementing an environment for hybrid software

evaluation. Science of Computer Programming,

236:103109.

ıaz-Mu

noz, A., Rodr

ıguez, M., and Piattini, M.

(2024b). Towards a set of metrics for hybrid (quan-

tum/classical) systems maintainability. JUCS - Jour-

nal of Universal Computer Science, 30(1):25–48.

ISO/IEC (2011). Iso/iec 25010:2011 systems and software

engineering — systems and software quality require-

ments and evaluation (square) — system and software

quality models.

ISO/IEC (2014). Iso/iec 25000:2014: Systems and software

engineering — systems and software quality require-

ments and evaluation (square) — guide to square.

Kruskal, W. H. and Wallis, W. A. (1952). Use of ranks in

one-criterion variance analysis. Journal of the Ameri-

can Statistical Association, 47:583–621.

Kumar, A. (2023). Formalization of structural test cases

coverage criteria for quantum software testing. Inter-

national Journal of Theoretical Physics, 62:1–16.

Piattini, M., Hevia, J. L. O., P

erez-Castillo, R., Serrano,

M., and Nodarse, G. P. (2021). Toward a quantum

software engineering. IT Professional, 23:62 – 66.

Piattini, M., Nodarse, G. P., P

erez-Castillo, R., Oliver, J.

L. H., Serrano, M., Gonz

alez, G. H., Guzm

an, I., Pa-

radela, C. A., Polo, M., Murina, E., Navajas, L. J.,

Marque

no, J., Gallego, R., Tura, J., Phillipson, F.,

Murillo, J., Ni

no, A., and Rodr

ıguez, M. (2020). The

talavera manifesto for quantum software engineering

and programming.

Rodr

ıguez, M., Oviedo, J. R., and Piattini, M. (2016). Eval-

uation of software product functional suitability: A

case study. Software Quality Professional Magazine,

18.

Rodr

ıguez, M. and Piattini, M. (2014). Software product

quality evaluation using iso/iec 25000. ERCIM News,

99:41–42.

Rodr

ıguez, M., Piattini, M., and Fern

andez, C. M. (2015).

A hard look at software quality: Pilot program uses

iso/iec 25000 family to evaluate, improve and certify

software products. IEEE Software, 48(09):30–36.

Verdugo, J., Oviedo, J. R., Rodr

ıguez, M., and Piattini, M.

(2024). Connecting research and practice for software

product quality evaluation and certiﬁcation: A soft-

ware laboratory’s 25-year journey. IEEE Software,

41:33–40.

IQSOFT 2025 - 1st International Conference on Quantum Software

104