In recent years with the advances in Machine Learning and Artificial Intelligence, the demand for ever smarter automation solutions could seem insatiable. One such demand was identified by Fortnox AB, but undoubtedly shared by many other industries dealing with contractual services, who were looking for an intelligent solution capable of predicting the expiration date of a contractual period. As there was no clear evidence suggesting that Machine Learning models were capable of learning the patterns necessary to predict a contract's expiration, it was deemed desirable to determine subject feasibility while also investigating whether it would perform better than a commonplace rule-based solution, something that Fortnox had already investigated in the past. To do this, two different solutions capable of predicting a contractual expiration were implemented. The first one was a rule-based solution that was used as a measuring device, and the second was a Machine Learning-based solution that featured Tree Decision classifier as well as Neural Network models. The results suggest that Machine Learning models are indeed capable of learning and recognizing patterns relevant to the problem, and with an average accuracy generally being on the high end. Unfortunately, due to a lack of available data to use for testing and training, the results were too inconclusive to make a reliable assessment of overall accuracy beyond the learning capability. The conclusion of the study is that Machine Learning-based solutions show promising results, but with the caveat that the results should likely be seen as indicative of overall viability rather than representative of actual performance.

The focus of this thesis is on the implementation, the evaluation and the useful application of the third dimension in software visualization. Software engineering is characterized by a complex interplay of different stakeholders that produce and use several artifacts. Software visualization is used as one mean to address this increasing complexity. It provides role- and task-specific views of artifacts that contain information about structure, behavior, and evolution of a software system in its entirety. The main potential of the third dimension is the possibility to provide multiple views in one software visualization for all three aspects. However, empirical findings concerning the role of the third dimension in software visualization are rare. Furthermore, there are only few 3D software visualizations that provide multiple views of a software system including all three aspects. Finally, the current tool support lacks of generating easy integrateable, scalable, and platform independent 2D, 2.5D, and 3D software visualizations automatically. Hence, the objective is to develop a software visualization that represents all important structural entities and relations of a software system, that can display behavioral and evolutionary aspects of a software system as well, and that can be generated automatically. In order to achieve this objective the following research methods are applied. A literature study is conducted, a software visualization generator is conceptualized and prototypically implemented, a structured approach to plan and design controlled experiments in software visualization is developed, and a controlled experiment is designed and performed to investigate the role of the third dimension in software visualization. The main contributions are an overview of the state-of-the-art in 3D software visualization, a structured approach including a theoretical model to control influence factors during controlled experiments in software visualization, an Eclipse-based generator for producing automatically role- and task-specific 2D, 2.5D, and 3D software visualizations, the controlled experiment investigating the role of the third dimension in software visualization, and the recursive disk metaphor combining the findings with focus on the structure of software including useful applications of the third dimension regarding behavior and evolution.

Abstract Modern software development must adapt to demanding schedules while keeping the software at a high level of quality. Agile software development has been adopted in recent years to meet such a need. Test-driven development (TDD) is one practice that has arisen within the agile software development movement that leverages unit tests to develop software in incremental cycles. TDD supporters claim that the practice increases the productivity of the practitioners who employ it, as well as the internal and external quality of the software they develop. In order to validate or refute such claims, the software engineering research community has studied TDD for the last decade; the results of the empirical studies on the effects of TDD have been mostly inconclusive. This dissertation has studied two factors that may impact the manifestation of the claimed effects of TDD on software’s external quality and developers’ productivity: the developers’ conformance to the process (i.e., their ability to follow TDD) and their skills. The research was performed in four phases. In the first phase, the literature was reviewed to identify a set of factors that have been considered to affect TDD. In the second phase, two experiments were executed within academia. A total of 77 students at the University of Oulu, took part in the studies. These experiments investigated the quality of the software, as well as the subject’s productivity with respect to their programming and testing skills. A follow-up study, using data collected during the second experiment, explored the relation between the quality, productivity and the subjects’ process conformance. In the third phase, four industrial experiments, involving 30 professional, were performed. Process conformance and skills were investigated in relation to the TDD’s effects on external quality and productivity. The fourth phase synthesized the evidence gathered in the two previous phases. The results show that TDD is not associated with improvements in external quality, or developers’ productivity. Further, improvements in both external quality and productivity are associated with skills, rather than with the process, at least in the case of professional developers. Hence, process conformance has a negligible impact. The productivity of novice developers, on the other hand, can benefit from the test-first approach promoted by TDD. / Tiivistelmä Modernin ohjelmistokehityksen täytyy mukautua haastaviin aikatauluihin säilyttäen ohjelmistojen korkea laatu. Ketterä ohjelmistokehitys on viime vuosina omaksuttu tähän tarpeeseen ja suuntauksessa on saanut alkunsa testivetoisen kehityksen käytäntö, joka hyödyntää yksikkötestausta ohjelmiston inkrementaalisessa, syklisessä kehityksessä. Testivetoisen kehityksen puolestapuhujat väittävät tämän käytännön lisäävän ohjelmistokehittäjien tuottavuutta sekä ohjelmiston sisäistä ja ulkoista laatua. Ohjelmistotuotannon tutkimusyhteisö on tutkinut testivetoista kehitystä viimeisen vuosikymmenen aikana vahvistaakseen tai kumotakseen nämä väitteet. Empiiriset tutkimukset testivetoisen kehityksen vaikutuksista ohjelmistotuotantoon ovat suurelta osin tuloksettomia. Tämä väitöstyö tutkii kahta tekijää, jotka voivat vaikuttaa testivetoisen kehityksen väitettyjen vaikutusten ilmentymiseen ohjelmiston ulkoisena laatuna ja ohjelmistokehittäjien tuottavuutena: ohjelmistokehittäjien taitoja ja prosessin mukaista toimintaa. Tutkimus toteutettiin neljässä vaiheessa. Ensimmäisessä vaiheessa tehtiin kirjallisuuskatsaus, jolla selvitettiin tekijöitä, joiden on katsottu vaikuttavan testivetoiseen kehitykseen. Toisessa vaiheessa tehtiin Oulun yliopistolla kaksi koetta, joihin osallistui kaikkiaan 77 opiskelijaa. Kokeissa tutkittiin ohjelmiston laadun ja osallistujien tuottavuuden suhdetta heidän ohjelmointi- ja testaustaitoihinsa. Toisen kokeen aikana kerättyä aineistoa hyödynnettiin jatkotutkimuksessa, jossa tarkasteltiin laadun, tuottavuuden ja prosessin mukaisen toiminnan suhdetta. Kolmannessa vaiheessa tehtiin neljä koetta, joihin osallistui 30 ohjelmistoalan ammattilaista. Prosessin mukaista toimintaa ja taitoja tutkittiin suhteessa testivetoisen kehityksen vaikutuksiin ohjelmiston ulkoiseen laatuun ja tuottavuuteen. Neljännessä vaiheessa syntetisoitiin kahden edellisen vaiheen löydökset. Tulokset osoittavat, ettei testivetoinen kehitys liity ulkoisen laadun parantumiseen eikä ohjelmistokehittäjien tuottavuuteen. Parannukset laadussa ja tuottavuudessa liittyvät ennemmin taitoihin kuin prosessiin, ainakin ohjelmistokehityksen ammattilaisten kohdalla. Näin ollen prosessin mukaisella toiminnalla on vähäpätöinen vaikutus. Toisaalta testivetoisen kehityksen suosiman test-first-menettelytavan hyödyntäminen voi edistää aloittelevien ohjelmistokehittäjien tuottavuutta.

Abstract Background: Confirmation bias is the tendency to search for evidence that confirms a person’s preconceptions. Confirmation bias among software testers is their tendency to validate the correct functioning of the program rather than testing it to reveal errors. Psychology literature suggests that time pressure may promote confirmation bias because time pressure impedes analytical processing of the task at hand. Time pressure is perceived negatively for its effects in software engineering (SE), therefore, its effect on confirmation bias may exacerbate software quality. Objective: We aim to examine confirmation bias among software testers. Additionally, we examine the effect of time pressure on confirmation bias and how time pressure affects the testers’ perception of the performance. We also question what other antecedents to confirmation bias exist in software testing and how they lead to it. Method: We first examined the state of the art research on cognitive biases in SE using systematic mapping. Then, we empirically examined the feasibility of using students in further experiments. An experiment with 42 students (novice professionals) investigated the manifestation of confirmation bias and whether time pressure promotes it. Another experiment with 87 novice professionals examined the perception of the performance of software testers under time pressure. A grounded theory study based on the interview-data of 12 practitioners explored other antecedents to confirmation bias in software testing and how they lead to it. Results: Time pressure emerged as a major antecedent to confirmation bias in the grounded theory. Testers prefer to validate the correct functioning of the program under time pressure. However, time pressure could not significantly promote confirmation bias among testers. Software testers significantly manifest confirmation bias irrespective of time pressure. The perception of performance is also sustained irrespective of time pressure. Conclusion: Testers should develop self-awareness of confirmation bias and improve their perception of performance to improve their actual testing. In the industry, automated testing may alleviate confirmation bias due to time pressure by rapidly executing the test suites. / Tiivistelmä Tausta: Vahvistusharha tarkoittaa taipumusta hakea ennakko-odotuksia vahvistavaa todistusaineistoa. Ohjelmistotestaajien vahvistusharha tarkoittaa taipumusta varmistaa ohjelmiston oikea toiminta mieluummin kuin hakea siitä virheitä. Psykologinen tutkimus esittää, että aikataulupaine voi lisätä vahvistusharhaa heikentämällä työn analyyttista tarkastelua. Aikataulupainetta pidetään ohjelmistotekniikan soveltamiseen kielteisesti vaikuttavana asiana, minkä johdosta sen vaikutus vahvistusharhaan voi heikentää ohjelmiston laatua. Tavoite: Tarkastelimme ohjelmistotestaajien vahvistusharhaa tutkimalla aikataulupaineen vaikutusta vahvistusharhaan ja testaajien käsitykseen testauksen tehokkuudesta. Lisäksi kysymme, mitkä muut tekijät johtavat ohjelmistotestauksen vahvistusharhaan, ja millä tavoin. Menetelmä: Ensiksi tarkastelimme ohjelmistotekniikan kognitiivisten harhojen viimeisintä tutkimusta systemaattista kirjallisuuskartoituksella. Sitten tutkimme kokeellisesti, miten yliopisto-opiskelijat soveltuvat käytettäväksi tutkimusjoukkona vahvistusharhan kokeellisessa tutkimuksessa. Kokeellinen tutkimus, johon osallistui 42 opiskelijaa (aloittelevaa ammattilaista), tarkasteli vahvistusharhan lisääntymistä aikataulupaineen vaikutuksesta. Toinen kokeellinen tutkimus, johon osallistui 87 aloittelevaa ammattilaista, tarkasteli ohjelmistotestaajien käsitystä testauksen tehokkuudesta aikataulupaineen alla. Kahdeltatoista ammattilaiselta haastattelemalla kerätystä tutkimusaineistosta tarkasteltiin ankkuroidun teorian menetelmällä muiden mahdollisten tekijöiden vaikutusta vahvistusharhaan. Tulokset: Ankkuroidussa teoriassa aikataulupaine osoittautui merkittäväksi vahvistusharhan tekijäksi. Ammattimaiset ohjelmistotestaajat haluavat mieluummin validoida ohjelmiston oikean toiminnan aikataulupaineessa. Toisessa kokeellisessa tutkimuksessa aikataulupaine ei kuitenkaan lisännyt merkittävästi testaajien vahvistusharhaa, vaan testaajien vahvistusharha ilmeni merkittävästi aikataulupaineista riippumatta. Myös käsitys työn tehokkuudesta säilyi riippumatta aikataulupaineesta. Johtopäätös: Ohjelmistotestaajien on syytä kehittää tietoisuuttaan vahvistusharhasta ja parantaa käsitystään työn tehokkuudesta parantaakseen testaustyötä. Teollisuudessa automaattinen testaus voi lieventää aikataulupaineen aiheuttamaa vahvistusharhaa nopeuttamalla testisarjoja.

The focus of this thesis is on the implementation, the evaluation and the useful application of the third dimension in software visualization. Software engineering is characterized by a complex interplay of different stakeholders that produce and use several artifacts. Software visualization is used as one mean to address this increasing complexity. It provides role- and task-specific views of artifacts that contain information about structure, behavior, and evolution of a software system in its entirety. The main potential of the third dimension is the possibility to provide multiple views in one software visualization for all three aspects. However, empirical findings concerning the role of the third dimension in software visualization are rare. Furthermore, there are only few 3D software visualizations that provide multiple views of a software system including all three aspects. Finally, the current tool support lacks of generating easy integrateable, scalable, and platform independent 2D, 2.5D, and 3D software visualizations automatically. Hence, the objective is to develop a software visualization that represents all important structural entities and relations of a software system, that can display behavioral and evolutionary aspects of a software system as well, and that can be generated automatically. In order to achieve this objective the following research methods are applied. A literature study is conducted, a software visualization generator is conceptualized and prototypically implemented, a structured approach to plan and design controlled experiments in software visualization is developed, and a controlled experiment is designed and performed to investigate the role of the third dimension in software visualization. The main contributions are an overview of the state-of-the-art in 3D software visualization, a structured approach including a theoretical model to control influence factors during controlled experiments in software visualization, an Eclipse-based generator for producing automatically role- and task-specific 2D, 2.5D, and 3D software visualizations, the controlled experiment investigating the role of the third dimension in software visualization, and the recursive disk metaphor combining the findings with focus on the structure of software including useful applications of the third dimension regarding behavior and evolution.

Tesis por compendio / [ES] El Desarrollo Dirigido por Modelos, MDD por sus siglas en inglés (Model Driven Development), es un enfoque de ingeniería del software que centra la creación y evolución de productos software en el modelado. Desde hace casi dos décadas, la comunidad científica ha descrito muchas de las ventajas de MDD frente a otros enfoques, sin embargo, su adopción en el entorno industrial es muy poco frecuente. Con el objetivo de entender por qué MDD no ha reemplazado otros enfoques de ingeniería software, he realizado una investigación empírica a través de tres experimentos controlados. Con el primer experimento pretendo aclarar si los beneficios de MDD frente al Desarrollo centrado en Código, CcD por sus siglas en inglés (Code Centric Development), son fieles a la realidad del desarrollo software actual. En el segundo experimento comparo la valoración que realizan los ingenieros de los modelos que utilizan, con su utilidad para ser utilizados en contextos MDD. En el tercer experimento analizo el desempeño de los profesionales software en tareas de mantenimiento en contextos MDD. Nuestros resultados confirman los beneficios de MDD frente a otros enfoques, pero también, que la intención de uso de MDD no alcanza valores máximos. Los sujetos subestiman el potencial de los modelos que desarrollan y utilizan en contextos MDD. El problema de adopción parece estar ligado a factores humanos, no a factores técnicos. / [CA] El Desenvolupament Dirigit per Models, MDD (Model Driven Development), és un enfocament d'enginyeria del programari que centra la creació i evolució de productes programari en el modelatge. Des de fa quasi dues dècades, la comunitat científica ha descrit moltes dels avantatges de MDD enfront d'altres enfocaments, no obstant això, la seua adopció en l'entorn industrial és molt poc freqüent. Amb l'objectiu d'entendre per què MDD no ha reemplaçat altres enfocaments d'enginyeria programari, he realitzat una investigació empírica a través de tres experiments controlats. Amb el primer experiment pretenc aclarir si els beneficis de MDD enfront d'altres enfocaments, com el Desenvolupament centrat en Codi, CcD (Code Centric Development), són fidels a la realitat del desenvolupament programari actual. En el segon experiment compare la valoració que realitzen els enginyers dels models que utilitzen, amb la seua utilitat per a ser utilitzats en contextos MDD. En el tercer experiment analitze l'acompliment del professional programari en tasques de manteniment en contextos MDD. Els nostres resultats confirmen els beneficis de MDD enfront d'altres enfocaments, però també, que la intenció d'ús de MDD no aconsegueix valors màxims. Els subjectes subestimen el potencial dels models que desenvolupen i utilitzen en contextos MDD. El problema d'adopció sembla estar lligat a factors humans, no a factors tècnics. / [EN] Model Driven Development (MDD) is a software engineering approach in which the code of a software product is generated and evolutionated from conceptual models that abstractly represents the system. For nearly two decades, the scientific community has described many of the advantages of MDD over other approaches. Despite the benefits of MDD, its use in real practical developments is merely anecdotal. To understand why MDD has not replaced other software engineering approaches, I have conducted an empirical investigation through three controlled experiments. The first experiment aims to clarify whether the benefits of MDD compared to code-centric development (CcD) match the reality of development in real environments. In the second experiment, I compare engineers' assessment of the models they develop with the usefulness of these models to be used in MDD contexts. In the third experiment, I analyze the performance of software professionals in maintenance tasks in MDD contexts. Our results confirm the benefits of MDD over other approaches; however, the intention to use MDD does not reach maximum values. Subjects underestimate the potential of the models they develop and use in MDD contexts. The adoption problem seems to be linked to human factors, not to technical factors. / Domingo Montes, MÁ. (2022). Experimentation to Evaluate the Benefits of Model Driven Development [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/185748 / TESIS / Compendio