💼 Ask A Manager — Salary Survey Analysis

Uncovering the truth behind 28,000+ real-world salaries

📊 Key Findings at a Glance

🏆 Top-Paying Departments Compensation & Benefits, Technology, and Law rank at the top — averaging $30–50K more than Education or Non-profit.	🌍 Same Role, Different Country The US tops most departments. Australia and the UK compete closely in Finance and Law.
⚧ The Gender Pay Gap Over Time Modest at entry level. By 15+ years, Male salaries nearly double while Female growth stays gradual.	🎓 Race × Education → Salary PhD and Professional degrees yield the highest returns, but the premium varies significantly by race.

---

What does a Software Engineer in Australia really earn compared to one in Canada?
Does a PhD actually pay off? Does the gender pay gap widen after 3 years?
This project answers all that — and more — with data.

---

📖 The Story Behind This Project

Every year, thousands of professionals anonymously share their salaries on Ask A Manager — one of the most trusted workplace advice platforms on the internet. The 2021 edition collected over 28,000 responses spanning dozens of countries, industries, genders, and education levels.

This project takes that raw, messy, human-reported data and transforms it into clear, visual answers to the questions professionals actually ask:

• Am I being underpaid for my experience?
• Which field should I move into for better pay?
• Does where I live really matter that much?

No fluff. No guesswork. Just honest analysis backed by real numbers.

---

✨ What This Analysis Covers

#	Question Explored	What You'll Discover
🏭	Which industry pays the most?	Top 20 highest-paying fields, ranked by average USD salary
📈	Does experience always mean more money?	How salaries climb — or plateau — across career stages
🌍	Same job, different country — how big is the gap?	Side-by-side salary comparison for the same roles across 10 countries
⚧	Is the gender pay gap real, and when does it kick in?	How the gap between Male, Female, and Non-binary salaries evolves with experience
🎓	Does education + race play a role in earnings?	A heatmap breakdown of average salaries by race and education level

---

🗂️ Project Structure

survey-analysis/
│
├── 📓 survey-analysis.ipynb     ← Main notebook: cleaning, EDA, and all 5 analyses
│
├── 📁 data/
│   └── salary_survey.xlsx       ← Raw survey data (28,000+ rows, as downloaded)
│
├── 📁 output/                   ← Auto-generated after running the notebook
│   └── clean_survey.xlsx        ← Cleaned dataset (two sheets: Clean + Dirty)
│
├── 📁 plots/                    ← Auto-generated visualizations (PNG, 300 DPI)
│   ├── salary_distribution_in_industries.png
│   ├── top_dept_avg_salary_barh.png
│   ├── salary_experience_plot.png
│   ├── avg_salary_dept_country.png
│   ├── salary_gender_experience_plot.png
│   ├── avg_salary_race_education.png
│   └── salary_race_education_heatmap.png
│
├── 📁 report/                   ← Summary report and findings
├── 📄 requirements.txt          ← Python dependencies
├── 📄 .gitignore
└── 📄 LICENSE

Note: The output/ and plots/ folders are created automatically when you run the notebook — no manual setup needed.

---

🔬 How the Data Was Cleaned

Raw survey data is notoriously messy — especially when 28,000 people are typing their own answers. Here's what was tackled before a single chart was drawn:

• 🗑️ Dropped irrelevant columns — Removed job_context, other_currency, income_context, and overall_experience to keep the dataset lean and focused.
• 🌐 Standardised country names — "u.s.a", "Unites States", "🇺🇸", and 25+ other variations were all mapped back to "United States".
• 🏭 Simplified industry & department labels — Long, inconsistent entries were trimmed to their first three meaningful words and title-cased.
• ⚧ Cleaned gender categories — "Man" → "Male", "Woman" → "Female", ambiguous responses → "Other".
• 🎭 Fixed race entries — Extracted the primary racial identity from mixed/lengthy responses.
• 📅 Converted experience brackets to numbers — "5 - 7 years" became 6.0 (the average), enabling proper numerical analysis.
• 💰 Currency normalisation — All 11 currencies (GBP, EUR, CAD, AUD, CHF, ZAR, SEK, HKD, JPY, AUD/NZD) converted to USD using 2021 exchange rates.
• 📊 Outlier removal — Extreme values filtered using the IQR method to prevent skewed visuals.
• 🧹 Missing values handled — Rows with critical nulls dropped; less critical fields filled with sensible defaults ("No Bonus", "N/A").

---

📊 Key Findings at a Glance

🏆 Top-Paying Industries Compensation & Benefits, Technology, and Law consistently rank at the top. Respondents in these sectors average $30–50K more per year than those in Education or Non-profit.	📈 Experience vs. Salary Salaries grow steadily through mid-career, with the steepest jump occurring between 3–15 years of experience. Beyond 22 years, growth levels off for most groups.
🌍 Location Premium is Real The United States consistently tops country-level pay rankings across most departments. However, Switzerland and the UK compete closely in Finance and Law roles.	⚧ The Gender Pay Gap Widens Over Time At entry level, salary differences between Male and Female respondents are modest. By 15+ years of experience, the gap becomes substantial — Male salaries nearly double, while Female growth is more gradual.

---

🚀 Getting Started

Prerequisites

Make sure you have Python 3.8+ installed. Then clone this repo and install the required packages:

# 1. Clone the repository
git clone https://github.com/horridhaider/survey-analysis.git
cd survey-analysis

# 2. Install dependencies
pip install -r requirements.txt

# 3. Launch Jupyter
jupyter notebook survey-analysis.ipynb

That's it. Run all cells from top to bottom — the output/ and plots/ directories will be created automatically.

---

🛠️ Built With

Tool	Purpose
Python 3.8+	Core programming language
Pandas	Data loading, cleaning, and transformation
NumPy	Numerical operations and outlier calculations
Matplotlib	All chart generation and figure styling
Seaborn	Heatmap visualisation (race × education)
OpenPyXL	Reading and writing .xlsx Excel files

---

📂 Data Source

The raw dataset is sourced from the Ask A Manager 2021 Salary Survey:

📎 Ask A Manager Salary Survey 2021 (Google Sheets)

The survey is self-reported, voluntary, and anonymous. Results reflect the experiences of respondents — primarily English-speaking professionals — and may not represent global averages.

---

📄 License

This project is licensed under the MIT License — see the LICENSE file for details.

You're free to use, adapt, and build upon this work for personal or commercial purposes, with attribution.

---

Made with 🐍 Python · 📊 Data · ☕ Coffee

⭐ Star this repo if you found it useful!