Although we have discussed the power imbalances caused by data, can data also have positive impacts？

This is an interactive program made with processing, demonstrating the research results of Watkins Allen et al. (2007). The study points out that surveillance plays both a controlling and a protective role in the workplace. While surveillance is often seen as a control mechanism, it can also protect employees from unfair job assignments or false accusations while objectively demonstrating job performance.

The visualization has 20 keywords spread across a black background, divided into three main categories:

•Surveillance is used for control

•Surveillance is caring

•Surveillance is good for the company

The high-frequency words are displayed in dark, large font, while the other words are displayed in light, small font. The words move slowly across the screen, when the mouse hovers over a word, the word pauses and pops up with its exact frequency of occurrence. This suggests that employees perceive monitoring as a phenomenon that blends control and caring.

Collected data usage can go far beyond business management. As Mayer-Schönberger and Cukier (2013) demonstrated, data analysis has proven particularly valuable in public health. By analyzing data on disease transmission, health authorities can predict outbreaks earlier and implement timely interventions. The same technology can be used to monitor laborers as well as to help society respond to public health emergencies. The key is who controls the data and how it is used.

When reflecting on our research, we realized the following shortcomings:

Sample Limitations: We collected 50 posts each from Facebook and X, which represent only a small fraction of the content available on social media. User behavior and communication styles could be different across platforms. This limited sample may not fully capture the complexity and variety of social media interactions.

Information Bias: Social media posts are often influenced by users' self-presentation, which can introduce biases in the data we collect. As a result, the data may not accurately reflect social behaviors or opinions.

Time Constraints: Due to the unavailability of web crawling tools, our data collection was done manually, which may have resulted in a small dataset. Then, our study focuses on a relatively short timeframe and may overlook long-term trends or seasonal variations.

Limitations of Secondary Data: Some of the data we cite from the literature are old. Current data and contexts may have changed, so these data may not accurately reflect the current situation.