Employees' Black Box
An information visualization of reviews from Glassdoor
The motivation for this research involved my interests in learning techniques to create visualizations based on text and the understanding of how we, designers, can better extract information when promoting participatory workshops or any other method that collects employees’ words, such as qualitative surveys or interviews.
I aimed to investigate some methods and tools to analyze text and create visualizations based on text with hands-on experience. This research involved some techniques to transform unstructured data, with no predefined data model, into structured data. I used data obtained from Glassdoor, treated it on Excel, analyzed it with the Voyant tool, and utilized Gephi to generate network visualizations.
Since the reviews from Glassdoor already bring the idea of positive or negative experiences, it was interesting how positive words appeared in a negative context. For example, the term 'good' was present in the cons reviews. Through the visualization, it was possible to see that the key message was the lack of opportunities to grow a career, and there were complaints about the compensation. The word ‘good’ was usually linked with a positive point of the organization, and then the reviewer mentioned what the bad point was. Additionally, this word was linked with people delivering the idea that the organization might lose good people because of its compensation.
Take a look at some of the information visualizations:
Take a look at this research: