Background: Data-driven decision-making (DDDM) has emerged as a primary approach to decision-making in many organizations. It uses data and analytics to guide decision-making processes and can lead to better business outcomes. Prior research has focused on DDDM in large corporations operating in large economies, and therefore this thesis will examine DDDM in small and medium enterprises in Sweden.  Purpose: The purpose of this research study is to examine the effect DDDM has on the financial performance of Swedish SMEs to investigate if the utilization of DDDM benefits companies financially and to understand the effect of managerial experience, technical skills, information quality, and firm size on the data-driven decision-making process. Method: This study is based on the positivism paradigm, following deductive reasoning and a quantitative approach of gathering data through digital surveys. The sample consisted of 55 Swedish SMEs gathered through simple random sampling. Further, the data was analyzed using Pearson correlation, Spearman rank correlation, and regression analysis to test hypotheses. Findings: The literature review identified a research gap on DDDM, factors that effect DDDM, and Financial Performance. Four hypotheses were developed to answer the research questions. The OLS regression found that DDDM had no significant effect on Financial Performance, the first hypothesis was not supported. The Information Quality variable had a significant positive effect on DDDM resulting in support for the second hypothesis. However, Managerial Experience and Technical Skills did not have a significant effect in the main regression model, hypotheses three and four were not supported. Conclusion: The thesis showed that DDDM did not have a significant effect on financial performance in Swedish SMEs. Additionally, managerial expertise and technical skills did not have an effect on DDDM. However, Information quality did have an effect on the DDDM process and was correlated with technical skills, which is in line with the theories used in the study: Organizational Information Processing Theory (OIPT) and Absorptive Capacity. This further supports that information quality is vital for the DDDM process and can explain why DDDM might not always lead to improvements in financial performance for Swedish SMEs.

<p>  </p> <p>First-Year Engineering (FYE) programs are a critical part of engineering education, yet they are quite complex settings. Given the importance and complexity of FYE programs, research to better understand student learning and inform design and assessment in FYE programs is imperative. Therefore, this dissertation showcases various uses of data analytics and educational theory to support decision-making when designing and assessing FYE programs. Three case studies shape this dissertation work. Each study encompasses a variety of educational data sources, analytical methods, and decision-making tools to produce valuable findings for FYE classrooms. In addition, this dissertation also discusses the potential for incorporating data analytics into FYE programs. A more detailed description of the research methods, a summary of findings, and a list of resulting publications for each case study follows.</p> <p>The first case study investigated the relationship between two related Computational Thinking (CT) practices, data practices and computational problem-solving practices, in acquiring other CT competencies in a large FYE course setting. This study explored the following research questions: (1) What are the different student profiles that characterize their foundational CT practices at the beginning of the semester? and (2) Within these profiles, what are the progressions that students follow in the acquisition of advanced CT practices? To answer these questions, N-TARP Clustering, a novel machine learning algorithm, and sound statistical tools were used to analyze assessment data from the course at the learning objective level. Such a hybrid approach was needed due to the high-dimensionality and homogeneity characteristics of the assessment. It was found that early mastery of troubleshooting and debugging is linked to the successful acquisition of more complex CT competencies. This research was published in an article in the journal <em>IEEE Access</em>.</p> <p>The second case study examined self-regulation components associated with students' successful acquisition of CT skills using students' reflections and assessment data. This research was grounded in three subprocesses of the Self-Regulated Learning (SRL) theory: strategic planning, access to feedback, and self-evaluation. This study responded to the following research question: What is the relationship between SRL subprocesses: access to feedback, self-evaluation, strategic planning, and the acquisition of CT skills in an FYE course? Results from a structural equation model, which reflects the complexity and multidimensionality of the analysis, provided evidence of the relevance of the three subprocesses in the acquisition of CT skills and highlighted the importance of self-assessment as key to success in the acquisition of programming skills. Furthermore, self-assessment was found to effectively represent the task strategy and access to feedback from the students. This analysis led to the understanding that even though the three SRL subprocesses are relevant for the student's success, self-evaluation serves as a catalyst between strategic planning and access to feedback. A resulting article from this case study will be submitted to the <em>International Journal of Engineering Education</em> in the future.</p> <p>Lastly, the third study aimed to predict the students' learning outcomes using data from the Learning Management System (LMS) in an FYE course. The following research questions were explored in this case study: (1) What type of LMS objects contain information to explain students' grades in a FYE course? (2) Is the inclusion of a human operator during the data transformation process significant to the analysis of learning outcomes? Two different sections of a large FYE course were used, one serving as a training data set and the other one as a testing data set. Two logistic regression models were trained. The first model corresponded to a common approach for building a predictive model, using the data from the LMS directly. The second model considered the specifics of the course by transforming the data from aggregate user interaction to more granular categories related to the content of the class. A comparison was made between the predictive measures, e.g., precision, accuracy, recall, and F1 score for both models. The findings from the transformed data set indicate that students' engagement with the career exploration curriculum was the strongest predictor of students' final grades in the course. This is a fascinating finding because the amount of weight the career assignments contributed to the overall course grade was relatively low. This study will be presented at the 2022 American Society of Engineering Education (ASEE) national conference in Minneapolis, Minnesota.</p>

Background: Over the years, Data Analytics (DA) has gained much attention enabling the extraction of valuable insights from the massive amount of data that is being produced every day. To exploit DA practices, various requirements for its successful usage are needed. Organizational Culture (OC) is provenly one critical intangible resource that is required for DA practices. However, there is a lack of existing research about what factors and values of OC facilitate DA practices. Purpose: The purpose of this study is to explore what role OC plays for DA practices and how OC can support the effective use of DA in a company. This research is guided by the research question: What are facilitating factors and underlying values of OC for DA Practices? By exploring and linking the two concepts of DA and OC, the study aims to provide a greater understanding of OC for DA practices. This offers insights for DA practitioners and managers to handle their specific OC and guide DA more targeted. Method: Following an inductive, qualitative study with an exploratory research design, the authors conducted 12 semi-structured interviews. The interviewees were selected through purposive sampling and represent two different perspectives: DA experts and DA collaborators. By conducting a Grounded Analysis, a deeper understanding of OC factors and values was created, leading to a final framework of an OC for DA practices.  Conclusion: The study results illustrate various OC factors that facilitate DA practices. These factors differ between subcultures, which are represented by four groups of actors. Three factors were identified as superior, as they had an enabling effect on DA practices in the investigated OCs. Finally, the study derived five underlying values, representing a shared cultural mindset among organizational members, that facilitate DA practice.

BACKGROUND: Big data offers new opportunities to enhance healthcare practice. While researchers have shown increasing interest to use them, little is known about what drives research impact. We explored predictors of research impact, across three major sources of healthcare big data derived from the government and the private sector. METHODS: This study was based on a mixed methods approach. Using quantitative analysis, we first clustered peer-reviewed original research that used data from government sources derived through the Veterans Health Administration (VHA), and private sources of data from IBM MarketScan and Optum, using social network analysis. We analyzed a battery of research impact measures as a function of the data sources. Other main predictors were topic clusters and authors’ social influence. Additionally, we conducted key informant interviews (KII) with a purposive sample of high impact researchers who have knowledge of the data. We then compiled findings of KIIs into two case studies to provide a rich understanding of drivers of research impact. RESULTS: Analysis of 1,907 peer-reviewed publications using VHA, IBM MarketScan and Optum found that the overall research enterprise was highly dynamic and growing over time. With less than 4 years of observation, research productivity, use of machine learning (ML), natural language processing (NLP), and the Journal Impact Factor showed substantial growth. Studies that used ML and NLP, however, showed limited visibility. After adjustments, VHA studies had generally higher impact (10% and 27% higher annualized Google citation rates) compared to MarketScan and Optum (p<0.001 for both). Analysis of co-authorship networks showed that no single social actor, either a community of scientists or institutions, was dominating. Other key opportunities to achieve high impact based on KIIs include methodological innovations, under-studied populations and predictive modeling based on rich clinical data. CONCLUSIONS: Big data for purposes of research analytics has grown within the three data sources studied between 2013 and 2016. Despite important challenges, the research community is reacting favorably to the opportunities offered both by big data and advanced analytic methods. Big data may be a logical and cost-efficient choice to emulate research initiatives where RCTs are not possible.

yes / A large number of cloud middleware platforms and tools are deployed to support a variety of Internet of Things (IoT) data analytics tasks. It is a common practice that such cloud platforms are only used by its owners to achieve their primary and predefined objectives, where raw and processed data are only consumed by them. However, allowing third parties to access processed data to achieve their own objectives significantly increases intergation, cooperation, and can also lead to innovative use of the data. Multicloud, privacy-aware environments facilitate such data access, allowing different parties to share processed data to reduce computation resource consumption collectively. However, there are interoperability issues in such environments that involve heterogeneous data and analytics-as-a-service providers. There is a lack of both - architectural blueprints that can support such diverse, multi-cloud environments, and corresponding empirical studies that show feasibility of such architectures. In this paper, we have outlined an innovative hierarchical data processing architecture that utilises semantics at all the levels of IoT stack in multicloud environments. We demonstrate the feasibility of such architecture by building a system based on this architecture using OpenIoT as a middleware, and Google Cloud and Microsoft Azure as cloud environments. The evaluation shows that the system is scalable and has no significant limitations or overheads.

With the developments of industry 4.0, data analytics solutions and their applications have become more prevalent in the manufacturing industry. Currently, the typical software architecture supporting these solutions is modular, using separate software for data collection, storage, analytics, and visualization. The integration and maintenance of such a solution requires the expertise of an information technology team, making implementation more challenging for small manufacturing enterprises. To allow small manufacturing enterprises to more easily obtain the benefits of industry 4.0 data analytics, a full-stack data analytics framework is presented and its performance evaluated as applied in the common industrial analytics scenario of predictive maintenance. The predictive maintenance approach was achieved by using a full-stack data analytics framework, comprised of the PTC Thingworx software suite. When deployed on a lab-scale factory, there was a significant increase in factory uptime in comparison with both preventative and reactive maintenance approaches. The predictive maintenance approach simultaneously eliminated unexpected breakdowns and extended the uptime periods of the lab-scale factory. This research concluded that similar or better results may be obtained in actual factory settings, since the only source of error on predictions would not be present in real world scenarios.

Well-detailed reinforced concrete (RC) shear walls did not achieve the expected seismic performance in the 2011 Christchurch earthquake as per the Canterbury earthquake royal commission report. Similarly, RC shear walls showed low seismic performance in the 2010 Maule earthquake. The two major seismic events intrigued this research dissertation, where six half-scaled RC shear walls were constructed and tested. The six walls were split into two phases, each phase had different end configurations (i.e., rectangular, flanged, and boundary elements). Phase II RC walls had 2.4 times the vertical reinforcement ratio of Phase I walls. The walls were detailed as per CSA A23.3-19, and they were tested laterally under a quasi-static cyclic fully-reversed loading while maintaining a constant axial load through the full test of the walls. The overall seismic performance of the six walls is evaluated in Chapters 2 and 3 in terms of their load-displacement relationships, crack patterns, displacement ductility capacities, stiffness degradation trends, curvature profiles, end strains, energy dissipation capabilities, and equivalent viscous damping ratios. In addition, damage states are specified according to the Federal Emergency Management Assessment (FEMA P58) guidelines. The results came in agreement with the Canterbury earthquake royal commission report, where the test walls with low vertical reinforcement ratios showed lower-than-expected seismic performance due to the concentration of their plastic hinges at the primary crack locations. Moreover, the results validated the Christchurch (2011) and Maule (2010) earthquake findings as concentrating the rebars at the end zones and providing adequate confinement enhanced the seismic performance of the test walls, which was the case for Phase II flanged and boundary element walls. The displacement ductility variations of the test walls inspired the work of Chapter 4, where the objective is to develop a data-driven expression for RC shear walls to better quantify their displacement ductility capacities. In this respect, an analytical model is developed and experimentally validated using several RC walls. The analytical model is then used to generate a dataset of RC walls with a wide range of geometrical configurations and design parameters, including cross-sectional properties, aspect ratios, axial loads, vertical reinforcement ratio, and concrete compressive strengths. This dataset is utilized to develop two data-driven prediction expressions for the displacement ductility of RC walls with rectangular and flanged/boundary element end configurations. The developed data-driven expressions accurately predicted the displacement ductility of such walls and they should be adopted by relevant building codes and design standards, instead of assigning a single ductility-related modification factor for all ductile RC shear walls, as per the 2020 National Building Code of Canada. Several researchers tested well-detailed Reinforced Masonry (RM) shear walls and the results concluded that RM shear walls showed high seismic performance similar to that of RC shear walls. This intrigued the research efforts presented in Chapter 5, where a comparative analysis is performed between the six RC walls tested in this dissertation and three RM walls tested in a previous experimental program. The analysis focuses on comparing the seismic performance of both wall systems in terms of their crack patterns, load-displacement envelopes, curvature profiles, displacement ductility, normalized periods, and equivalent viscous damping ratios. In addition, an economic assessment is performed to compare such RC and RM shear walls using their total rebar weights and the total construction costs. Overall, RM shear walls achieved an acceptable seismic performance coupled with low rebar weights and low construction costs when compared to their RC counterparts. / Thesis / Doctor of Philosophy (PhD)

Modern retailers have been collecting more and more data over the past decades. The increased sizes of collected data have led to higher demand for data analytics expertise tools, which the Umeå-founded company Infobaleen provides. A recurring challenge when developing such tools is the data itself. Difficulties in finding relevant open data sets have led to a rise in the popularity of using synthetic data. By using artificially generated data, developers gain more control over the input when testing and presenting their work. However, most methods that exist today either depend on real-world data as input or produce results that look synthetic and are difficult to extend. In this thesis, I introduce a method specifically designed to generate synthetic transactional data stochastically. I first examined real-world data provided by Infobaleen to determine suitable statistical distributions to use in my algorithm empirically. I then modelled individual decision-making using points in an embedding space, where the distance between the points serves as a basis for individually unique probability weights. This solution creates data distributed similarly to real-world data and enables retroactive data enrichment using the same embeddings. The result is a data set that looks genuine to the human eye but is entirely synthetic. Infobaleen already generates data with this model when presenting its product to new potential customers or partners.

Choosing a data collection and analytics tool is no easy task. There are numerous tools available and no tool fits all. The choice of the tool depends on various factors and especially what type of organization and product the tool will be used for. Startups generally have less resources than established companies and the set of tools they can select therefore is limited. Therefore, the study has focused on finding and determining which tool is best suited for a startup. The study has chosen five criterias to evaluate a selected number of tools; setup, features, usability, privacy and cost. The criterias have been selected based on interviews, research and theoretical background. The study’s practical work has been performed at a startup in order to get real-world insight in how the tools operate and function. The tools evaluated are; Mixpanel, Amazon Pinpoint, and Google Sheets. The tools were selected based on answers from questionnaires, interviews and research. The result of the study shows that out of the three tools evaluated, Mixpanel is regarded as the most optimal tool for startups. Mixpanel is cost-effective, easy to set up, offers multiple features and is overall a user-friendly tool. It especially allows the company to have a transparent and strong privacy aspect. Allowing full control of the data collected and analyzed. However, what the study also concluded is that no specific tool is always best for everyone and the organizations should always understand their needs in order to pick the most suitable tool. / Att välja ett datainsamlings- och analysverktyg är ingen lätt uppgift. Det finns många verktyg men inget verktyg passar alla. Val av verktyget beror på olika faktorer särskilt vilken typ av organisation och produkt, verktyget ska användas till. Nystartade företag har i regel mindre resurser än etablerade företag vilket gör att verktyg de kan välja är begränsad. Därför har studien fokuserat på att hitta och avgöra vilket verktyg är bäst lämpat för en startup. Studien har valt fem kriterier för att utvärdera ett utvalt antal verktyg; installation, funktioner, användbarhet, integritet och kostnad. Kriterierna har valts ut utifrån intervjuer, forskning och teoretisk bakgrund. Studiens praktiska arbete har utförts vid en startup för att få verklig insikt i hur verktygen fungerar. De verktyg som utvärderas är; Mixpanel, Amazon Pinpoint, och Google Kalkylark. Verktygen valdes ut utifrån svar från enkäter, intervjuer och forskning. Resultatet av studien visar att av de tre utvärderade verktygen så anses Mixpanel vara det mest optimala verktyget för startups. Mixpanel är kostnadseffektivt, lätt att installera, erbjuder många funktioner och är generellt användarvänligt verktyg. Det tillåter företaget att ha transparent och stark integritet. Utöver det, tillåter Mixpanel full kontroll över data som samlas i och analyseras. Däremot, slutsatsen som kan också dras av studien är att inget specifikt verktyg alltid är bäst för alla och organisationerna bör alltid första sina behov för att välja det mest lämpliga verktyget.

Due to increasing data complexity, researchers are struggling to visually explore data using traditional methods. Research has shown that a variety of analytics-related tasks can be enhanced using new emerging technologies (e.g., virtual reality [VR] and augmented reality [AR]). This paper provides a detailed design approach for developing an immersive virtual engine system for conducting exploratory and descriptive data analytics. This system is called the Immersive Virtual Exploratory Engine (IVEE), is a VR system that allows users to experience full intractability with its graphical elements and examines various datasets within the same immersive environment. Basic plots—such as histograms, line plots, and other exploratory analytics—can be created in both 2D and 3D to visualize datasets. Furthermore, the proposed module provides a data sub-setting system, entitled “Lasso,” as well as a visual element merge system that allows users to subset and merge data using natural interactions only. The system also allows for the simultaneous visualization of multiple representations, thus supporting decisions that require numerous plots. Furthermore, the system supports remote collaboration, allowing users from different locations to come together in a virtual space and work collaboratively as they would in a real-life setting.