[ Thinking ] [ ASU MOT Coursework · Fall 2025 ]← All work

Innovation Index Analysis: G20 vs Non-G20

Built a Python analysis on a 1,862-record innovation dataset spanning 130 countries and 13 years, merged from WIPO, World Bank, and UNESCO sources. Ran t-tests and ANOVA comparing G20 vs non-G20 innovation scores (F=19.14, p<0.001) and surfaced 15 non-G20 countries scoring above the G20 average. Tested an education-spending-to-innovation hypothesis via linear regression and reported the counter-finding that education explains only 7.3% of innovation variance (R²=0.07). Wrote a 22-page analysis paper translating statistical findings into policy-relevant recommendations for a 2-person team.

1,862

Dataset Records

130 countries · 13 years

F=19.14

G20 vs Non-G20 ANOVA

p<0.001

15

Non-G20 Overperformers

above the G20 average

R²=0.07

Education → Innovation

counter to dominant narrative

What I did

  • Built a Python analysis on a 1,862-record innovation dataset spanning 130 countries and 13 years, merged from WIPO, World Bank, and UNESCO sources
  • Ran t-tests and ANOVA comparing G20 vs non-G20 innovation scores and surfaced 15 non-G20 countries scoring above the G20 average (F=19.14, p<0.001)
  • Tested an education-spending-to-innovation hypothesis via linear regression and reported the counter-finding that education explains only 7.3% of innovation variance (R²=0.07)
  • Wrote a 22-page analysis paper translating statistical findings into policy-relevant recommendations for a 2-person team

How

  1. 01

    Data Merge

    Consolidated WIPO, World Bank, and UNESCO sources into a clean 1,862-record dataset

  2. 02

    Hypothesis Testing

    t-tests and ANOVA across G20 vs non-G20 groupings

  3. 03

    Regression

    Linear regression on education spending → innovation score

  4. 04

    Policy Translation

    22-page paper translating statistical findings into policy recommendations

← Back to all work