Research Proposal

Michael Cran

November 21, 2023

Writing for Technical Professionals

Research Proposal

Unveiling Bias: Understanding Media Algorithm Influences

Overview

        The objective of this research is to investigate the issue of algorithmic bias across various social media, news, and employment platforms. The study will focus on understanding how algorithms contribute to inequalities, reinforce stereotypes, and perpetuate bias in various contexts. Our research aims to address the following questions: "How do algorithms impact gender norms and stereotypes in social media and advertising?", "What measures can be implemented to detect and mitigate biases in social media data collection?", and "Can computation algorithms be effectively used to reduce perceived bias in news articles?" Through an extensive review of existing research, this study seeks to provide insights into the consequences of algorithmic bias and offer recommendations for creating a more equitable and transparent algorithmic system in these areas.

Project Background

Significance and Importance

        This research project holds significant importance as it aims to illuminate the widespread issue of algorithmic bias and its implications it has on various aspects of our lives. In an era where technology and algorithms are advancing rapidly, incorporating artificial intelligence and learning capabilities that surpass human understanding, it becomes essential to investigate and address the biases that threaten our lives. By delving into the bias present in media algorithms, we can confront an issue that directly prohibits individuals' access to opportunities, resources, and information. This discrimination perpetuates social injustice. By gaining a deeper understanding of this problem, we can work towards a fairer and more equitable society. To foster progress, technology must prioritize transparency, starting with machine learning, where addressing bias in decision-making processes (algorithms) can identify and rectify these issues. This endeavor can serve as a catalyst for demanding greater transparency and accountability from big tech companies, policymakers, and other stakeholders. Through this understanding, we can shape the development of strategies to mitigate the detrimental effects of algorithmic bias on society.

Critical Frameworks

        In order to examine the role of algorithms in perpetuating gender stereotypes and norms, we will incorporate feminist and gender studies frameworks. These perspectives offer valuable insights into how algorithms affect diverse gender identities and how they can be made more inclusive.

        This research project is very important as it has the potential to address social injustices, promote transparency, and ensure accountability, while simultaneously mitigating the harmful effects of algorithmic bias. By delving deeper into the complexities of algorithmic bias and its implications, we can steer society toward a future that is more fair and just, with technology serving as a force for good and progress.

Research Methods

        Our research will begin by conducting an extensive review of the pre-existing research and information on algorithmic bias. This step is crucial in identifying gaps in past and current research so that we can use this information to refine our research questions to address what has not been addressed before. We will then collect data from various sources such as academic papers, journals, online content, and other documents, which will be vital in supporting our research questions. We will use this data for extensive analysis and apply analytical techniques to identify patterns, biases, and trends in the data to see where machine decision-making has flaws. This will uncover specific instances where algorithmic bias is occurring; we can assess its impact and inform others of our findings as well as use the Critical Frameworks I talked about in the previous section to interpret our findings and provide a deeper understanding of the bias. Through these critical frameworks, we are given the ability to understand other perspectives which will give us deeper insights into these biases. Based on all of our research, data, and analysis, we can formulate a comprehensive discussion to help formulate a conclusion and decide the best course of action when it comes to counteracting and mitigating the negative effects of algorithmic bias in the media. We will propose practical and actionable recommendations to foster transparency and protect equity in the digital landscape.

Timeline

• Weeks 1 through 2 will consist of an Extensive Review of Research.
• Weeks 3 through 6 will consist of Data Collection.
• Weeks 7 through 10 will consist of Data Analysis.
• Weeks 11 through 12 will consist of the Application of Critical Frameworks.
• Weeks 13 through 14 will consist of a Discussion.
• Week 15 through 16 will consist of a Conclusion and recommendations, as well as a presentation of findings and suggestions.

Expected Outcomes

        Upon completion of our research on algorithmic bias, we will inform policy and raise awareness within the UCF community and beyond. We plan to submit our research findings and insights to a reputable scholarly journal in the fields of computer science, social sciences, and technology. This article will provide an in-depth analysis of our findings on algorithmic bias, incorporating critical frameworks, evidence, and recommendations for solutions to this matter. To make our research accessible to those outside of being in this field of research or a UCF student or teacher, we will create a white paper that summarizes our key findings and recommendations, which can be sent to policymakers and industry leaders so that they can easily understand the implications of this issue and its importance in addressing it. We will also present our research at an academic conference focused on technology ethics, social injustice, and discrimination to amass a wider audience and inform more people. We plan to provide workshops and seminars at UCF to engage with students and faculty and stimulate discussions on algorithmic bias. We will share our research products by publishing them in online repositories available globally, actively presenting them in conferences, workshops, and seminars, and engaging with students around campus to educate them on this matter. This project contributes to the fields of technology, computer science, and social sciences by offering a comprehensive analysis of finding algorithmic bias and proposing a way to mitigate it, as well as highlighting the pressing need for transparency. The UCF community will gain increased awareness of the social implications of algorithmic bias, and they will be empowered to critically assess and acknowledge that the technology they use daily may lack fairness and equity. We will also offer recommendations and initiatives that students and faculty can rally their voice their support for.

Sources

Anxiety, Panic and Self-Optimization: Inequalities and the YouTube ...,
                        journals.sagepub.com/doi/abs/10.1177/1354856517736978.

Fred Morstatter, et al. “Discovering, Assessing, and Mitigating Data Bias in Social
                        Media.” Online Social Networks and Media, Elsevier, 12 Apr. 2017,
                        www.sciencedirect.com/science/article/abs/pii/S24686964 16300040.

Can an Algorithm Reduce the Perceived Bias of News ... - Sage Journals,
                        journals.sagepub.com/doi/full/10.1177/1077699018815891.

D’Alonzo, Samantha, and Max Tegmark. “Machine-Learning Media Bias.” PLOS
                        ONE, Public Library of Science, journals.plos.org/plosone/article?id=10.1371%2Fjo
                        urnal.pone.0271947.

Reinscribing Gender: Social Media, Algorithms, Bias, www.tandfonline.com/doi/full/10.1
                        080/0267257X.2020.1832378.

Williams, Betsy Anne, et al. “How Algorithms Discriminate Based on Data They
                        Lack: Challenges, Solutions, and Policy Implications.” Scholarly Publishing Collective, Duke
                        University Press, 1 Mar. 2018, scholarlypublishingcollective.org/p
                        sup/information-policy/article/doi/10.5325/jinfopoli.8.2018.0078/314474/How-Algorithms-Discr
                        iminate-Based-on-Data-They.

Preliminary Work and Experience

        I have a strong foundation and passion for the areas relevant to this research project. With my major being computer science, my coursework is very closely related to computer science, data ethics, and data analysis which are all relevant when conducting this research. These courses have equipped me to address the complex issue of algorithmic bias. Additionally, my experience in social media management for content creators on YouTube has led to my in-depth analysis and research on the YouTube algorithm. The research I have conducted on the YouTube algorithm has enabled me to understand what the algorithm favors and dislikes in making a video "trend" or become "viral." This also relates to gender stereotypes and gender norms as well as the biases the youtube algorithm has when it comes to showcasing creators on the "Recommended / For You homepage." These valuable academic and work experiences showcase my research skills and expertise in this area, demonstrating my passion and ability to complete this research project.

IRB/IACUC Statement

I believe this research project does not require IRB or IACUC approval as it does not involve contact with human subjects or animal subjects.

Budget

In order to conduct the research required for this project we will be seaking $1,100 dollars worth of expenses

Budget Breakdown

• Access to Academic Journals and Databases: $300 (Includes Social Media & News API Fees)
• Software and Data Analysis Tools: $500 (Includes Online and Desktop Software and Applications)
• Travel and Conference Fees: $200 (Sharing our findings with industry leaders.)
• Workshop and Seminar Products Creation: $100 (Sharing our findings with our generation.)

Written By,

Michael D. Cran