Methodology
Selecting Our Sources
Our team used the data set by Oxford Insights titled, “Government AI Readiness Index 2024,” a database that includes 180+ countries and their rating on different pillars that contribute to their AI readiness. Things like human capital, digital infrastructure, innovation, and more allow us to see beyond technical scores to social, political, and historical conditions of countries. There are several dimensions to how countries prepare to develop and accept AI and that is captured with our dataset. The data set gives us numbers but our goal was to tell a story of those numbers through explanation, patterns, and visuals.
In addition to the main dataset, we incorporated peer-reviewed academic papers that explain why projects with similar economic resources can perform so differently. We used this evidence to prove that the dataset was more than a neutral ranking but rather what history, governance structures, and humanistic frameworks do to a country. We did also confirm how there could be limitations in the Index, more specifically how countries may have incomplete reporting systems or how some countries may be underrepresented but we used Oxford’s report to try to gain the most information about these possible flaws. By thinking of the dataset within a more social dynamic, we aimed to understand AI readiness as more than an indicator of technological advancement. We analyzed it as a product of national policies, cultural norms, inequities in education, and long-standing political histories. More information about our specific sources can be found in our narrative and annotated bibliography.
We would like to remind readers that for our timeline, the dates and information was compiled with the assistance AI (Chat GPT) but was processed and displayed by our team members. This can be found in our acknowledgement.
Ultimately, our project approached the dataset through a humanistic lens, combining quantitative score with qualitative research, to show how it is not a simple race between nations, but a reflection of how communities are centered through this technological process.
Processing Our Data:
The dataset we used was relatively clean and had a few null values that we were able to clean using R. To create the data visualizations we used Tableau for all the different styles of graphs. Since most of the team is majoring in Statistics and Data Science all the analysis was done in R as that is where we are experienced.
The visualizations were done in Tableau since it is a platform to get good practice from since it has a wide range of graphing capacities that were useful to help visualize the answer to our research questions. Our data was mostly quantitative, therefore, that drove most of our decisions when deciding which style of graph to use that best visually explained our narrative. Due to our experience, using two platforms worked best for our team. R was helpful for our research to get summaries and get statistical analysis while Tableau required little coding for sophisticated visualizations. The minimal coding involved creating new variables in the Data section to make things like slope, residuals, and predicted totals in the calculated fields.
Presenting Our Narrative:
We built our website using WordPress, hosted on UCLA’s Humspace portal. We experimented with a variety of templates before deciding on the current structure, as it allowed us to organize our content in the clearest and most professional way. For the theme, we chose a beige background with black text to ensure readability for color-blind users. We also considered accessibility, making sure the text size is large enough for those with visual impairments and intentionally avoiding audio elements.
To create a visually consistent experience, we used a simple color palette across the website. The visualizations, however, used more color to differentiate data points and make patterns easier to see. The colors chosen are visible to those with color blindness. All charts and graphs are clearly labeled for easy interpretation. Basic CSS and HTML enhancements were also added to improve the site’s appearance and performance.
WordPress gave us the flexibility to customize the layout and formatting so the site could be both visually appealing and accessible. Guided by principles of accessible web design, we made deliberate color choices, used appropriately sized fonts, and bolded important text to enhance readability for all users. We also kept the visualizations and site layout simple to avoid overwhelming visitors.
We cleaned and analyzed our quantitative data in R, applying our team’s background in Statistics and Data Science, and used Tableau to create a variety of interactive graphs that allow readers to hover over data points for more information, with minimal coding required for the visualizations.