# Data Science 101 — Handover note

**To:** Dr Amara Chen  
**From:** Insight Education Analytics trainee (worked example)  
**Re:** A-level subject entries — first exploration  
**Data:** DfE *A level and other 16 to 18 results*, Academic year 2024/25  
(*Time-series — A level subject entries and grade by sex*)

## Two patterns

1. **Subject size is extremely uneven.** Among individual subjects in England in 2024/25,
   **Maths** leads with **101,079** entries. The
   distribution is right-skewed (median ≈ 14,092; mean ≈
   22,415), so percentages for tiny subjects will be noisy.
2. **Computer Science has grown but remains male-skewed in the wider file.** National All-students
   entries rose from **8,550** to **18,682**.
   A SQL join of Female and Male rows also shows Computer Science among the lowest
   female shares (about **18.7%** female in this extract).

## Best supporting chart

`outputs/charts/module4_top15_subjects.png` — top 15 individual subjects by entry count
(aggregates such as “Total subjects” excluded).

## Limitation

This exploration uses **entries**, not unique students, and cannot explain **why** choices
differ. Grade percentages for small subjects can mislead without entry-count context.

## Python vs SQL

**Python/pandas** felt more natural for charts and descriptive stats. **SQL** was clearer for
reproducing ranked filters and the Female/Male join. I would use SQL to shape extracts and
Python for analysis and visuals.

## Next question

Is the Computer Science female share statistically below 50% in 2024/25, and how large is the
effect size? That is the brief for Data Science 201.