khaukhau
commented
1 month ago - Clarity of data visualization: We meant to use a diverging bar chart instead of a pyramid. Here is the intended graph use:
- Problem with dynamic interaction: Group J misunderstood our proposal as there is only 1 drop-down filter (by sector) that applies to many graphs, instead of each graph having many interactive components, so the site would be much less overwhelming than Group J thought it for.
- Suggestion: We acknowledge that the added guide and annotations for users to increase ease of use is a good idea and we will implement that in our project.
tienvu95
First, there are some impressive points your team has made:
Clear objective: When I read the high-level goal of the proposal, I can see it is well articulated and indeed addresses a huge social issue, gender pay gap, specifically in the United Kingdom. Since your team has a focused and meaningful purpose, this project will not only serve as a mere "course project", but also a chance to educate and raise awareness. Data preprocessing: Your team has a thorough plan for datapreprocessing, like filtering and categorisation which is of paramount importance to show that you care about maintaining data integrity and relevance for a meaningful analysis. Detailed plan and timeline: the proposal includes a well-structured weekly plan with specific milestones and clear responsibilities assigned to team members. In my point of view, having such project management principles will help your team execute the project successfully.
Besides that, I have some questions: Clarity of data visualisation: The use of pyramid bar charts to display the gender pay gap can be misleading as such visualisations might imply a hierarchical relationship where none exists. Do you think of any other alternative visualisation formats could better represent equality and comparability between data points without implying a hierarchy? For this, I think a dual-axis horizontal bar chart for comparing the mean percentage difference in hourly pay between genders within each sector may help. Potential overwhelm due to dynamic interaction: As much as I am impressed with the idea of using interactive elements, I am concerned with the possible drawback of them. If every graph on the dashboard has interactive components like dropdowns and sliders, do you think they might overwhelm your users with too many options? Here, what might be the optimal number of interactive elements on a single visualisation? I would suggest simplyfing user interactions by limiting the number of controls required to manipulate the data. Your team may think of implementing a guided analytic flow where users are directed through a story about the gender pay gap, with interactive elements introduced gradually, one-by-one.
Thank you and good luck with your last course project at VinUni.