|
Step
|
Description
|
Common Tools
|
Notes
|
|
1. Data Exploration
|
The data is first examined to understand structure, content, and potential quality issues.
|
R, Python, Excel
|
Quick descriptive summaries and visual checks help identify outliers, missing data, or inconsistent categories.
|
|
2. Initial Data Preparation
|
Data is cleaned, reshaped, and transformed to prepare for analysis and visualization.
|
R (tidyverse), Python (pandas), SQL
|
This includes renaming columns, handling missing values, merging datasets, and recoding variables.
|
|
3. Manual Tidying & Quality Checks
|
Data issues such as free-text responses or inconsistent category labels are corrected manually.
|
Excel, Google Sheets
|
Manual review ensures accuracy where automation is difficult or context-dependent.
|
|
4. Further Data Structuring
|
The cleaned file is re-imported for further transformation and preparation for visualization.
|
R, Python
|
More complex restructuring, grouping, or statistical summaries are applied here.
|
|
5. Sketching & Design Prototyping
|
Visualization ideas are drafted and iterated on paper (or digital sketching tools) to test layout, narrative, and flow.
|
Pen & Paper, Figma
|
Low-stakes sketching encourages experimentation and storytelling before coding.
|
|
6. Visualization Development
|
The final visuals are created, combining code-based reproducibility with design principles.
|
Tableau, R (ggplot2, patchwork), Python (Matplotlib, Seaborn, Plotly)
|
Code-based plotting allows consistency, reproducibility, and fine control over aesthetic elements.
|
|
7. Interactive Dashboard Creation (optional)
|
For projects requiring user interaction, filtering, or exploration, results are deployed as a dashboard.
|
R Shiny, Python Dash, Tableau, Power BI
|
Dashboards enable ongoing exploration and sharing with stakeholders.
|
