This week I learned how to build a webpage. I experimented with opening a website and using “Inspect Element” to view its underlying code. I even discovered that I could modify the text or layout within the inspector, but the changes only appear on my own screen rather than for all users, which I found really fascinating. Before class, I downloaded Phoenix and FileZilla so I could create and upload webpages more smoothly. I activated my account, created an HTML file, opened it through Phoenix, and named it index.html. I successfully created and uploaded my own webpage. Since this was my first time encountering web design, I felt unsure about where to begin. With the instructor’s step-by-step guidance, I gained a basic understanding of how a webpage is created, and I also learned that I could download templates to decorate and customise my site.
Data from online platforms can be scraped and captured for analysis, which helps us better understand how platforms work. The process involves clicking “Inspect,” which opens a new window in the browser. I then located the “Web Scraper” tab and created a new scraper. Under “Create new sitemap,” I entered a name and the URL of the website I wanted to scrape. After that, I clicked “Add new selector” to define the page elements I wanted to extract. By clicking “Select,” a small box appeared on the page, and the tool automatically attempted to identify the correct code for a functioning scrape.
This week we designed a questionnaire. Our group was assigned Scenario 3, where we took on the role of academic researchers at a UK university. The aim was to understand how students use generative artificial intelligence in their daily lives. Specifically, our study focused on how Chinese international students in the UK use AI tools, examining their motivations, usage patterns, perceived benefits and challenges. By focusing on the “how” aspect, the research aimed to provide a nuanced understanding of their interactions with AI technologies. We designed five questions, created the questionnaire, and distributed it.
We collected and analysed the questionnaire responses. There were several issues: First, the sample size was too small to produce representative data or reliable conclusions. Second, the number of questions was limited, which reduced reliability and validity and increased the likelihood of bias. Since all questions and wording in a questionnaire are created by the researchers, the instrument determines what can and cannot be expressed. The data produced is filtered through power structures and reflects the interests and values of the institution behind it. The act of quantifying people’s experiences is, in itself, a form of power.
This week we analysed advertising data across different platforms to understand how platforms interpret and profile us. I chose Instagram to view my data. Although there are options suggesting I can delete my history, I still do not trust the platform and believe it continues to store my data for personalised recommendations. I prefer social media apps that emphasise user privacy protection. On TikTok, I can only clear the cache, but I cannot view all the data the app holds about me. Even though I feel my privacy is being violated, I cannot avoid using it because I rely on algorithmic recommendations to show me content I like. Personalised recommendation is a major reason why I easily become addicted to social media. Algorithms analyse my likes, shares and viewing duration to infer my preferences and push tailored videos. Xiaohongshu’s algorithm feels even stronger. After browsing a topic only two or three times, the platform floods my homepage with similar content. Sometimes, conversations I have with friends seem to trigger related recommendations when I later open the app. This feels quite alarming; even though there’s no evidence, I feel as if I’m constantly being monitored.