Global Data Science Salary Dashboard

Overview

When looking through a dataset to analyze, I tried to find a topic that would be interesting to whomever is taking the MSDS program. Therefore, I decided to pick a dataset that focuses on the job market for data scientists. I found the dataset on Kaggle, which contains records of global job postings, specifically in data science. The dataset has cleaned and organized data from around the globe.

Information about the dataset:

URL Link: Kaggle Dataset
Data Source: Kaggle
Dataset Author: Adil Shamim
Data Sources: aijobs.net Salary Dataset (CC0)
- 365datascience.com
- Payscale
- KDnuggets
- ZipRecruiter
- Wellfound (AngelList)
Scope: Global data science salaries from 2020 to 2025
Data Size: Approximately 10,000 records

Data required some manipulation and preparation:

Renamed abbreviations: For example, 'SE' to 'Senior', 'EN' to 'Entry Level', etc. this was required for employment type, countries, and experience level.
Added longitude and latitude: I had to add longitude and latitude for the countries to plot them on a map.
Dropped NaN values: to ensure clean data for visualization.
Outliers: I removed outliers so it does not skew the data.
Made new tables for the counts of each country: this was essential for the map and the bar chart.

Goals of the Project:

The main goal here is to create an interactive visualizations that allows users to explore the data science job market globally.

Goals:

Understanding salary distribution globally and across different job dimensions.
Geographical insights to get a global perspective.
Temporal patterns with salaries and job demand over the years.
Exploration features like filtering to allow for personalized comparative analysis.

Primary Research Questions:

How do the salaries of data scientists vary by country, experience level, employment type?
Which countries have the highest demand for data scientists?
Is there an increase in the number of data science jobs over the years?
How do salaries change over time?

Target Audience:

Job Seekers in data science
Employees looking to move abroad
Employers looking to understand the job market to hire data scientists
Researchers studying job trends in data science

Interactive Visualizations

1. Interactive Dashboard

The dashboard is multi-panel that offers a wide variety of information. It offers a histogram, a bar chart, and a statistical summary. There is also a filtering feature so the user can explore the data and the numbers that are most relevant to them.

Design elements and justification:

3 panels displaying different aspects of the data; provide comprehensive overview where user can explore by filtering.
Histogram for salary distributions; job titles vary greatly in data science, so a histogram allows for easy comparison.
Bar chart for top 10 job titles; allow to pinpoint which job titles are more common than others.
Statistical summary; allows users to have a deeper understanding of the data that is filtered in and out.
Shared Filtering options; allow user to look up specific jobs characteristics to focus on.
Color scheme consistency; colors are consistent to each job title among the 2 graph panels to allow for easy comparison.
Job title angled labels; to reduce white space and make the chart more readable.

2. Global Perspective

This geographical map offers a global perspective of the job market in data science. I added a bubble overlay with proportional sizing to show where the job demand is highest. I also added a tooltip to provide more information on the map with a custom encoding to show "on-site | remote | hybrid" job types. I also added the job counts on the map for top countries.

Design elements and justification:

Geographical global map; allows for a global perspective of the job market.
Bubble overlay sized proportionally; shows where the job demand is highest.
Map colors; the light blue color is softer on the eye with white strokes to mark the borders of the countries.
Tooltip; provides more information
Countries bubble colors; Custom colors to ensure it does not clash with the background and the map, but are still soft and muted for better readability.
Bubble legend on the right; showing the colors of the countries in a descending order of job counts to emphasize the most in-demand countries.
Bubble legend on the bottom; showing proportional size difference with relation to job counts for reference
Count labels on the map for top countries; to quickly locate the countries with the highest demand on the map.

3. Time Series Salary Trends

The graph is a line chart, this graph offers a view of how the salaries in the top countries for data scientists have changed over the five years. I also added a detailed tooltip to provide statistical summaries. There is also filters to allow for specific salary trends.

Design elements and justification:

Line chart for salary over time; will show how salaries have changed over the years.
Line chart for top 10 countries; this ensures that there is no clutter while showing significant trends.
Detailed tooltip with statistical summary; shows more information on the data points.
Filtering options; allows user to look for trends among specific experience levels.
Color for each country; custom colors for softer and muted view, and allows user to easily follow a country trend of interest.

4. Job Demand Over Time

The bar chart shows how the number of data science jobs has changed over the years, allowing us to see the job demand growth. I also added filters and a tooltip providing more information. I overlayed the bars with a line to the trajectory of the job demand. There are some filtering features here if required as well.

Design elements and justification:

Bar chart with job counts each year; allows to see the growth of data science jobs over the years.
Overlayed line chart for trajectory; to show the direction the bar chart is going.
Filtering options; allows users to focus on specific countries or experience levels.
Tooltip; provides count and year information for each bar.
Color for each country: colors change with each country to emphasize the data viewed is different after filtering.
Color scheme; custom colors for softer and muted view.

Evaluation

A mix of Insight-Based and Usability Evaluation Method

I asked 8 people (family and friends) to use the visualization graphs and provide feedback. The evaluation sessions were 15-30 minutes long with a 1 week follow-up.

Insights Discovered:

Some Complex insights:

"Remote jobs correlates with higher salaries, especially in the US"
"At entry level, most common job title is 'data analyst', while at senior level, most common job title is 'data scientist'"

Some Unexpected insights:

"The job demand is not as growing from 2024 to 2025, contrary to what the media hype about AI"

Evaluation Metrics and Results:

Time to insight:
- average 3.4 minutes to insight
- average 13 minutes to deeper or more complex insights
Number of insights:
- average 5.6 insights per session
- range 2-12 insights per session
Average User rating:
- for career decision importance: 4.5/5
- for visualization clarity: 4.7/5
- for job market understanding: 4.6/5
- for geographical patterns: 3.1/5
Deeper Assessment:
- 6 out of 8 users asked follow up questions
- 5 out of 8 users asked for more data
- 4 out of 8 mentioned sharing their interesting insights with colleagues

Conclusion:

What went well:

The filtering system allowed users to build insights and to test hypotheses across multiple dimensions
The geographical map prompted random insights and questions.
The tooltips provided information the user almost always asked or wondered about.
The colors were well received, no complaints about brightness or contrast.

What could be improved:

The geographical map was not as intuitive as I hoped; requires more data points globally.
The map needed a select feature to allow users to zoom in on specific areas of the map where countries are close together.
The job demand growth graph had a lot of whitespace.
Living expenses and cost of living play a role in comparing salaries from different countries, this could be added to the dashboard.(suggested by 1 user in the follow-up questionnaire, I didn't have enough time to implement)

Data insights and conclusions:

The data gave us multiple insights and trends about the data science job market. The following are insights were found by myself and users from the evaluation sessions:

Global Distribution:

The United States dominates the data science job market. Compared to the rest of the world, the US has the highest number of job postings and salaries.
European countries (UK, Germany, France) show a significant amount of job postings in the global market.
Asia and South America have a growing presence and new opportunities.

Job Titles:

Job titles vary exponentially in the field of data science.
Common job titles include "Data Analyst", "Data Scientist", "Data Engineer", and "Software Engineer"
Common job titles vary among countries, some countries have titles not present in the rest of the countries.
Most common entry level job title is "Data Analyst", while the most common senior level job title is "Data Scientist".

Salary Trends:

Salaries vary significantly by country and experience level.
US has the highest salaries, however, living cost should be considered for true comparison.
European countries have lower salaries but still competitive for the lower cost of living.
Remote jobs are generally paid lower than on-site jobs, especially in the US.
Salary ranges are plateauing since last year.

Market Growth:

There has been consistent job postings growth from 2020 to 2024 which indicated that the field is growing and new.
2025 only shows a slight increase in job demand; however, the year is not over yet and there may be more job postings.
Job demand pattern is different by experience level.
Job demand patterns are different by country which may indicate the difference in the field maturity level regionally.

Work Arrangement:

On-site, remote, and hybrid work arrangement are well represented across the regions.
On-site jobs are the most common overall.

Note: The insights above were seen from exploring data from 2020 to 2025. however, i believe there should be more data points globally than the data collected above. each country with different languages may have other job postings websites that were not taken into account. The data above is insightful, but in my opinion, not comprehensive.