DB FPX 9801 Assessment 4 Pre-Implementation Conference Call Presentation

Student Name

DB-FPX9801

Capella University

Professor Name

submission Date

Pre-Implementation Conference Call Presentation

Slide 01

My name is ____ and I’m here today to do a pre-implementation conference call on leadership strategies for implementing machine learning-based fraud detection systems in U.S. financial institutions.

Slide 02

Introduction

With the increasing automation of transactions and fraudsters constantly bothering with new schemes, financial institutions have made fraud detection one of their top priorities toward both business and cyber security (Joseph & Eaw, 2023). Technological development results have been highly predictive, yet leadership decision making and organizational integration continue to be problematic for organizations in terms of implementation and adoption. The second project is to understand the nature of the interpretation, implementation and operationalization of these ML-based fraud detection systems by the organizational leaders as these are not yet widely in practice.

Slide 03

Statement of the Problem

• Problem Statement

The root of the issue is that the financial industry is grappling with an escalating number of challenges from complex fraud schemes that make backscattering money and subject the institutions to regulatory backlash. However for some time there have been also more advanced machine learning technologies for fraud detection, and many companies have yet to achieve effective implementation (Stratton, 2024).

The challenge here is that there is the lack of working practices or algorithms to fully incorporate machine learning fraud detection into practices of the financial institution, and in the cases where it was adopted you still find the effectiveness of the systems to have dropped and opportunities missed for proactive fraud mitigation (Bouncken et al., 2025).

• Purpose Statement

The application of the qualitative study is to gain insights into how fraud detection systems with machine learning are implemented in financial institutions.

• Gap in Practice

Most studies on the influence of leadership decision-making and organizational strategies in these systems have only looked at the successful application of high-tech C2 fraud detection technologies (Joseph & Eaw, 2023).

Slide 04

Theory(s) Supporting Research

Technology acceptance model (TAM) is a framework that is used to elicit how technology will be accepted by understanding cognitive and behavioural perceptions that will be used for decision making processes. TAM assumes that both attitude toward using the technology system (AU) and intentions to use (IU) are functions of perceived usefulness (PU) and perceived ease of use (PEU), and that these two factors also influence actual use (U) of the technology system (Fred Davis, 1989). In the context of financial institutions, perceived usefulness involves leadership beliefs about the attainment of increased accuracy in fraud detection, improved efficiency in operations and increased success in strategic outcomes while perceived ease of use involves the leadership beliefs related to the complexity of implementation and organizational effort (Joseph & Eaw, 2023). The use of the TAM in the management and information systems literature has been well established as it has been a determinant in understanding the adoption of new technologies from organizations, particularly the adoption of new products and services involving artificial intelligence and machine learning (Pajany, 2021). Adoption analysis can be further extended by a complementary theory viewpoints outlined in Unified Theory of Acceptance and Use of Technology (UTAUT) to call attention to performance expectancy, social influence, and facilitating conditions at the organizational level that affect the manager’s decision (Borhani et al., 2021). The adoption of the machine learning-based fraud detection in financial institutions can be explored using the TAM constructs (Masumbuko & Phiri, 2024). Thus, TAM provides the suitable conceptual base to analyze the managerial decision making procedures that result in successful technology adoption in complex monetary contexts.

Slide 05

Methodology

The study will take a generic qualitative inquiry approach to examine the leadership strategies that are being applied in implementing machine learning–based fraud detection systems in financial institutions in the U.S. It is a qualitative approach as in the research, managerial views, decision-making processes, and organisational experiences, but does not measure any causal procedures and statistical outcomes. A generic qualitative study can be useful to explore participants’ mindsets and perceptions of leadership in real-world settings in complex organisational contexts (Jafri et al., 2024). The main research method will be semi-structured interviews, which will allow flexibility while ensuring that the research objectives are met. To ensure credibility and confirmability methodological rigour will be achieved using reflexive notes, triangulation, and an audit trail. Data collection will be done to achieve data saturation for deep and comprehensive data insights of the themes.

Slide 06

Sample

Purposive sampling will be used in the study for the participants who have direct knowledge to detect fraud and those who are technology leaders. Purposive sampling enables the qualitative approach, since it identifies which given, so-called, information-rich cases have to offer as information for the study that is congruent with the research focus and questions (Stratton, 2024). Performing qualitative research guidelines that state saturation of the data is usually reached with 9-17 participants, the proposed number of leaders from the U.S. financial institutions is approximately 12. Qualitative approaches are specifically useful in the case of managerial perspectives, decision-making processes, organizational experiences, and where there is a lack of measurement through numbers (Jafri et al., 2024). To be eligible for inclusion, respondents must be bank, credit union, or fintech leaders with first-hand experience in machine learning fraud programs. It will make sure that individuals who are not responsible for leadership or those who have no relevant experience in implementation will be excluded to ensure that data is relevant. To reduce participant attrition, communication, flexible schedules, and transcript review will be provided.

Slide 07

Data Collection

Before recruitment of the participants, data collection will be done in a structured and ethical manner with a multi-step process with the help of institutional approval. Participants’ credibility and relevance will be ensured through their professional directories, industry associations, and professional networks to identify eligible leaders. Evaluation of methodological rigour and data quality will be ensured through a leadership responsibility screening process, which evaluates experience in using machine learning. After the informed consent, one-to-one virtual, semi-structured interviews using secure video conference platforms will be performed. Semi-structured interviewing is a good method to have a free and open narrative and remain close to the research questions. Interviews will be supplemented by reflexive field notes, while triangulation will be done between the transcripts, notes, and literature to ensure trustworthiness and confirmability. Once attrition is reached, data collection will end.

Slide 08

Instrumentation

A semi-structured interview protocol will be used as the primary research instrument, where the interview protocol will be developed by the researcher to investigate different facets of machine learning fraud detection adoption as seen by the leaders. The interview guide will include open-ended questions related to the research question and theoretical approach that, to some extent, will provide consistency in the interviewing process, while at the same time providing a space for exploring emerging insights. Semi-structured instruments are popular instruments in qualitative leadership research as they provide flexibility in research, allowing participants to share their experiences and are also structured in a way that ensures the interview is focused (Jordan et al., 2021). Questions of interviews will be content validated by expert review of faculty advisors and qualitative research specialists. Reflexive journaling and field notes will serve as “connected” tools to help in the context interpretation. Expert feedback on instruments refines the study and ensures that the study and the collection are aligned with each other.

Slide 09

Sample Interview Questions

• What is the level of satisfaction of the financial institution leaders about the value of machine learning systems in enhancing fraud detection results?
• What are the organisational factors that impacted adoption decisions for machine learning (ML) fraud detection technology?
• What is the impact of perceived implementation complexity on leadership attitudes to adoption?
• What obstacles were encountered when machine learning systems were integrated into the existing fraud management process?
• What is the relationship between regulatory needs and fraud detection technology adoption leadership strategies?
• What organizational resources or supports occurred that helped the implementation to be successful?
• What is driving the sustainability in machine learning investments for fraud detection?

Slide 10

Data Collection Plan

The study will use a systematic, ethical, and rigorous methodology to gather information from top financial institution leaders in the U.S. who are using machine learning–based fraud detection methodologies. The participants will be identified according to the relevant and credible professional directories and networks, as recommended in the sampling of participants in qualitative research. To align with best practices and ensure good quality and ethical compliance of data, screening will be used for the verification of the position held in the team of leadership, previous experience in the adoption of machine learning, and decision-making power. Previous studies showed that proper screening of participants according to their professional experience and decision-making role in technology adoption would lead to more robust and valid qualitative data that would be collected during research. (Ahmed, 2024). The one-on-one interviews will be semi-structured and designed to enable both deep exploration of their thoughts and easy access to geographically diverse participants via virtual methods. The context of the observations, any dynamics between the participants, as well as initial impressions and analysis, will be recorded in the reflexive field notes. Thematic saturation will be used to confirm data completeness, and triangulation of interview transcriptions, field notes, and literature will improve confirmability.

Slide 11

Data Analysis Plan

Study data analysis will be organized systematically as focused qualitative data analysis to find the pattern and theme in the persistence of machine learning–based fraud detection technology adoption through leadership perspectives. Thematic analysis will be used to code the interview data and reflexive field notes to classify the findings by the TAM constructs, including perceived usefulness and perceived ease of use. Thematic analysis can be supplemented by triangulation, which helps enhance the rigor and validity of the results in technology adoption studies, as confirmed by the findings of previous qualitative research (Beg, 2025). Confirmability and the minimization of researcher bias will be achieved through triangulation of data sources such as transcripts, field notes, and pertinent literature. To ensure transparency and reproducibility, an audit trail will be kept of coding decisions, thematic matrices, and reflexive log by the research team.

Slide 12

Conclusion

The project examines leadership views of financial institutions in the U.S. on the adoption of machine learning in fraud detection. The Technology Acceptance Model will be utilized to gain insight into the decision-making and barrier to adoption. To gain rich qualitative data for understanding in context, purposive sampling and semi-structured interviews will be used. The data will be collected using rigorous processes and themed to ensure the results are trustworthy and credible, and the data will be triangulated. The study can provide guidelines on how to be effective in linking managerial perspectives with technology in the financial context.

References

Ahmed, S. K. (2024). The pillars of trustworthiness in qualitative research. Journal of Medicine, Surgery, and Public Health, 2(1), 1–4. https://doi.org/10.1016/j.glmedi.2024.100051 

Ames, H., France, E. F., Cooper, S., Bianchim, M. S., Lewin, S., Schmidt, B., Uny, I., & Noyes, J. (2024). Assessing qualitative data richness and thickness: Development of an evidence‐based tool for use in qualitative evidence synthesis. Cochrane Evidence Synthesis and Methods, 2(7), e12059. https://doi.org/10.1002/cesm.12059 

Beg, M. J. (2025). Qualitative methods in mental health research: Standards for ethical inquiry, research practice, and peer review. Indian Journal of Psychological Medicine, 0(0). https://doi.org/10.1177/02537176251363817 

Borhani, S. A., Babajani, J., Vanani, I., Anaqiz, S., & Jamaliyanpour, M. (2021). Adopting blockchain technology to improve financial reporting by using the technology acceptance (TAM). International Journal of Finance & Managerial Accounting, 6(22), 155-171. http://www.ijfma.ir/article_17481.htmachine learning

Bouncken, R. B., Czakon, W., & Schmitt, F. (2025). Purposeful sampling and saturation in qualitative research methodologies: Recommendations and review. Review of Managerial Science, 46, 1–37. https://doi.org/10.1007/s11846-025-00881-2 

Dahal, N., Neupane, B. P., Pant, B. P., Dhakal, R. K., Giri, D. R., Ghimire, P. R., & Bhandari, L. P. (2024). Participant selection procedures in qualitative research: Experiences and some points for consideration. Frontiers in Research Metrics and Analytics, 9(1), e1512747. https://doi.org/10.3389/frma.2024.1512747 

Ebot, A. (2024). Technology acceptance model for adopting cybersecurity technology in small and medium business/enterprise: A generic qualitative study. https://www.proquest.com/openview/821ddca62bfb9689de0e377a43f7dfba/1?cbl=18750&diss=y&pq-origsite=gscholar

Granić, A. (2024). User acceptance of interactive technologies. Foundations and Fundamentals in Human-Computer Interaction, 356-389. https://doi.org/10.1201/9781003495109-12

Jafri, L., Farooqui, A. J., Grant, J., Omer, U., Gale, R., Ahmed, S., Khan, A. H., Siddiqui, I., Ghani, F., & Majid, H. (2024). Insights from semi-structured interviews on integrating artificial intelligence in clinical chemistry laboratory practices. BioMed Central Medical Education, 24(1), e170. https://doi.org/10.1186/s12909-024-05078-x 

Jordan, J., Clarke, S. O., & Coates, W. C. (2021). A practical guide for conducting qualitative research in medical education: Part 1—How to interview. Academic Emergency Medicine Education and Training, 5(3), e10646. https://doi.org/10.1002/aet2.10646 

Joseph, P. S., & Eaw, H. C. (2023). Still technology acceptance model Reborn with exostructure as a service model. International Journal of Business Information Systems, 44(3), 404-421. https://doi.org/10.1504/ijbis.2023.134949

Masumbuko, C., & Phiri, J. (2024). Technology adoption as a factor for financial performance in the banking sector using UTAUT model. African Journal of Commercial Studies, 4(2), 121-130. https://doi.org/10.59413/ajocs/v4.i2.5

Pajany, P. (2021). Ai transformative influence: Extending the tram to management student’s AI’s machine learning adoption – ProQuest. Proquest.com. https://search.proquest.com/openview/317b9e11c91c8b592822e9cb42b758ba/1?pq-origsite=gscholar&cbl=18750&diss=y

Potthoff, S., Hempeler, C., Gather, J., Gieselmann, A., Vollmann, J., & Scholten, M. (2023). Research ethics in practice: An analysis of ethical issues encountered in qualitative health research with mental health service users and relatives. Medicine Health Care and Philosophy, 26(4), 517–527. https://doi.org/10.1007/s11019-023-10169-5 

Stratton, S. J. (2024). Purposeful sampling: Advantages and pitfalls. Prehospital and Disaster Medicine, 39(2), 121–122. https://doi.org/10.1017/S1049023X24000281 

Tenny, S., Brannan, J., & Brannan, G. (2022, September 18). Qualitative study. National Library of Medicine; StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK470395/