Background

Innovation drives the economy, and of the main contributors are those that reside in the academe.

This data story aims to seek the understand the influence of educational institutions that are part of the Times Higher Education World University Rankings in the innovation index. The Global Innovation Index of 2022 is primarily scored based on the innovation input index and the innovation output index.

The following questions will be answered throughout the exploration on data given by the Times Higher Education World University Rankings and the Global Innovation Index:

• How many universities are present in each country found in the THE World University Rankings?
• Is the median score of all of the universities in one country correlated with the country' ranking in the Global Innovation Index, focusing on the following criteria:
  • scores_teaching: Teaching score - learning environment score
  • scores_research: Research score - volume, incode and reputation of University
  • scores_citations: Citations score - research influence of University

The following scores from the dataset were chosen for data exploration since this are activities related to innovation output.

Methodology

This data story will make use of two datasets:

• uni_rank.csv - Shows the THE World University Rankings
• inno_index.csv - Shows the Global Innovation Ranking of each country observed

Universities of the same country will be also analyzed by the median so that it takes into consideration that data is asymmetric. These values will be compared against their respective country's rank to see if there is a correlation present.

Setting up the data

Import the libraries and the data that will be used

import pandas as pd
import matplotlib.pyplot as plt

wipo_index = pd.read_csv("inno_index.csv")
ranking = pd.read_csv("uni_rank.csv")

Replace headings of wipo_index.csv to 'location' instead of 'Economy

wipo_index.rename(columns={'Economy':'location'}, inplace=True)
wipo_index.head()

	rank	location	Income Group(Strength/Weakness)	Strength / Weakness	Score
0	1	Switzerland	Strength	Strength	64.6
1	2	United States of America	Strength	Strength	61.8
2	3	Sweden	Strength	Strength	61.6
3	4	United Kingdom	Strength	Strength	59.7
4	5	Netherlands	NaN	Strength	58.0

ranking.head()

	rank_order	rank	name	scores_overall	scores_overall_rank	scores_teaching	scores_teaching_rank	scores_research	scores_research_rank	scores_citations	...	scores_international_outlook_rank	location	stats_number_students	stats_student_staff_ratio	stats_pc_intl_students	stats_female_male_ratio	aliases	subjects_offered	closed	unaccredited
0	10	1	University of Oxford	96.4	10	92.3	5	99.7	1	99.0	...	28	United Kingdom	20,965	10.6	42%	48 : 52	University of Oxford	Accounting & Finance,General Engineering,Commu...	False	False
1	20	2	Harvard University	95.2	20	94.8	1	99.0	3	99.3	...	214	United States	21,887	9.6	25%	50 : 50	Harvard University	Mathematics & Statistics,Civil Engineering,Lan...	False	False
2	30	=3	University of Cambridge	94.8	30	90.9	7	99.5	2	97.0	...	34	United Kingdom	20,185	11.3	39%	47 : 53	University of Cambridge	Business & Management,General Engineering,Art,...	False	False
3	40	=3	Stanford University	94.8	40	94.2	2	96.7	7	99.8	...	223	United States	16,164	7.1	24%	46 : 54	Stanford University	Physics & Astronomy,Computer Science,Politics ...	False	False
4	50	5	Massachusetts Institute of Technology	94.2	50	90.7	8	93.6	11	99.8	...	119	United States	11,415	8.2	33%	40 : 60	Massachusetts Institute of Technology	Mathematics & Statistics,Languages, Literature...	False	False

5 rows × 24 columns

Create new dataframe wherein the ranking.csv dataset has only the following: rank, name, scores_teaching, location of all of the schools listed.

rdf = pd.DataFrame(ranking[['rank','location','scores_teaching','name']])
rdf['rank'] = rdf['rank'].str.replace('\W', '', regex=True)
rdf.tail()

	rank	location	scores_teaching	name
2340	Reporter	United Kingdom	NaN	York St John University
2341	Reporter	Nigeria	NaN	Yusuf Maitama Sule University, Kano
2342	Reporter	Ukraine	NaN	Zhytomyr Polytechnic State University
2343	Reporter	Pakistan	NaN	Ziauddin University
2344	Reporter	Jordan	NaN	Zarqa University

Ensure there are no duplicates or typos in 'location' in the ranking.csv file when compared with wipo_index.csv file

rdf = rdf.replace({'location':{"C?te d'Ivoire" : "Cote d'Ivoire",
                            'Iran (Islamic Republic of)' : 'Iran',
                            'Republic of Moldova' : 'Moldova',
                            'Republic of Korea': 'South Korea',
                            'T\x81rkiye' : 'Turkey',
                            'United States of America':'United States',
                            'Viet Nam':'Vietnam'
    
                            }})
pd.unique(rdf['location'])

rdfloc = pd.unique(rdf['location'])
rdfloc.sort()
print(rdfloc)

['Albania' 'Algeria' 'Argentina' 'Armenia' 'Australia' 'Austria'
 'Azerbaijan' 'Bangladesh' 'Belarus' 'Belgium' 'Botswana' 'Brazil'
 'Brunei Darussalam' 'Bulgaria' 'Cambodia' 'Canada' 'Chile' 'China'
 'Colombia' 'Costa Rica' 'Croatia' 'Cuba' 'Cyprus' 'Czech Republic'
 'Denmark' 'Ecuador' 'Egypt' 'Estonia' 'Ethiopia' 'Fiji' 'Finland'
 'France' 'Georgia' 'Germany' 'Ghana' 'Greece' 'Hong Kong' 'Hungary'
 'Iceland' 'India' 'Indonesia' 'Iran' 'Iraq' 'Ireland' 'Israel' 'Italy'
 'Jamaica' 'Japan' 'Jordan' 'Kazakhstan' 'Kenya' 'Kosovo' 'Kuwait'
 'Kyrgyzstan' 'Latvia' 'Lebanon' 'Libya' 'Lithuania' 'Luxembourg' 'Macao'
 'Malawi' 'Malaysia' 'Malta' 'Mauritius' 'Mexico' 'Moldova' 'Mongolia'
 'Montenegro' 'Morocco' 'Mozambique' 'Namibia' 'Nepal' 'Netherlands'
 'New Zealand' 'Nigeria' 'Northern Cyprus' 'Norway' 'Oman' 'Pakistan'
 'Palestine' 'Paraguay' 'Peru' 'Philippines' 'Poland' 'Portugal'
 'Puerto Rico' 'Qatar' 'Romania' 'Russian Federation' 'Saudi Arabia'
 'Serbia' 'Singapore' 'Slovakia' 'Slovenia' 'Somalia' 'South Africa'
 'South Korea' 'Spain' 'Sri Lanka' 'Sudan' 'Sweden' 'Switzerland' 'Taiwan'
 'Tanzania' 'Thailand' 'Tunisia' 'Turkey' 'Turkmenistan' 'Uganda'
 'Ukraine' 'United Arab Emirates' 'United Kingdom' 'United States'
 'Uruguay' 'Uzbekistan' 'Venezuela' 'Vietnam' 'Zambia' 'Zimbabwe']

Each of the university has a score for teaching environment which will contribute to the mean of their country. However, not all countries are present in the university world rankings, therefore must be removed.

Country Overview

This shows how many universities are present in THE World University Rankings per country.

country_plot = rdf['location'].value_counts()
country_plot.plot.bar(subplots=True, figsize=(20, 6))
plt.title("No. of Universities in THE World University Rankings per Country")
plt.xlabel("Countries")
plt.ylabel("No. of Universities")

Text(0, 0.5, 'No. of Universities')

Getting the median of each country in terms of teaching score

country = []
median = []

for x in rdfloc:
    ave = rdf.loc[rdf['location'] == x, 'scores_teaching'].median()
    ave = round(ave, 2)
    country.append(x)
    median.append(ave)


finaldf = pd.DataFrame(columns = ["location","median"])
finaldf['location'] = country
finaldf['median'] = median
finaldf.dropna(inplace = True)
finaldf.tail()

	location	median
112	United States	36.80
115	Venezuela	15.90
116	Vietnam	14.65
117	Zambia	15.70
118	Zimbabwe	14.80

Get the innovation ranking of each country present in THE World University Rankings by creating a dataframe of the wipo_index.csv file

wipodf = pd.DataFrame(wipo_index[['rank','location']])
wipodf = wipodf.replace({'location':{"C?te d'Ivoire" : "Cote d'Ivoire",
                            'Iran (Islamic Republic of)' : 'Iran',
                            'Republic of Moldova' : 'Moldova',
                            'Republic of Korea': 'South Korea',
                            'T\x81rkiye' : 'Turkey',
                            'United States of America':'United States',
                            'Viet Nam':'Vietnam'
    
                            }})
wipodf.head()

	rank	location
0	1	Switzerland
1	2	United States
2	3	Sweden
3	4	United Kingdom
4	5	Netherlands

pd.merge(finaldf, wipodf, on = 'location')

	location	median	rank
0	Algeria	18.90	115
1	Argentina	18.20	69
2	Australia	25.90	25
3	Austria	30.95	17
4	Azerbaijan	16.50	93
...	...	...	...
89	United Kingdom	24.30	4
90	United States	36.80	2
91	Vietnam	14.65	48
92	Zambia	15.70	118
93	Zimbabwe	14.80	107

94 rows × 3 columns

adf = pd.merge(finaldf, wipodf, on = 'location')
adf = pd.DataFrame(pd.merge(finaldf, wipodf, on = 'location'))
adf.sort_values(by=['rank'])

	location	median	rank
82	Switzerland	35.25	1
90	United States	36.80	2
81	Sweden	29.70	3
89	United Kingdom	24.30	4
59	Netherlands	43.20	5
...	...	...	...
24	Ethiopia	17.15	117
92	Zambia	15.70	118
86	Uganda	21.80	119
56	Mozambique	12.10	123
37	Iraq	18.30	131

94 rows × 3 columns

adf.corr(numeric_only = True)

	median	rank
median	1.000000	-0.585637
rank	-0.585637	1.000000

Visualization of the country's innovation ranking against the teaching score of the universities within that country

plt.scatter(adf['rank'], adf['median'])
plt.xlabel("Innovation Rank")
plt.ylabel("Median Teaching Score")
plt.title("Innovation Ranking against the Teaching Score")
plt.show()

Initial Analysis: There is an inverse correlation between the innovation rank and the median teaching score of the country. This suggests that the improvement of the teaching score will lead to a higher ranking in the Global Innovation Index.

Getting the median of each country in terms of research score

resdf = pd.DataFrame(ranking[['rank','location','scores_research','name']])
resdf['rank'] = resdf['rank'].str.replace('\W', '', regex=True)
resdf.dropna(inplace = True)
resdf.tail()

	rank	location	scores_research	name
1794	1501	Japan	12.3	Yamaguchi University
1795	1501	China	10.9	Yanshan University
1796	1501	Turkey	12.2	Yeditepe University
1797	1501	Russian Federation	8.9	Yuri Gagarin State Technical University of Sar...
1798	1501	Turkey	8.2	Zonguldak Bülent Ecevit University

resdfloc = pd.unique(resdf['location'])
resdfloc.sort()
print(resdfloc)

['Algeria' 'Argentina' 'Australia' 'Austria' 'Azerbaijan' 'Bangladesh'
 'Belarus' 'Belgium' 'Botswana' 'Brazil' 'Brunei Darussalam' 'Bulgaria'
 'Canada' 'Chile' 'China' 'Colombia' 'Costa Rica' 'Croatia' 'Cuba'
 'Cyprus' 'Czech Republic' 'Denmark' 'Ecuador' 'Egypt' 'Estonia'
 'Ethiopia' 'Fiji' 'Finland' 'France' 'Georgia' 'Germany' 'Ghana' 'Greece'
 'Hong Kong' 'Hungary' 'Iceland' 'India' 'Indonesia' 'Iran' 'Iraq'
 'Ireland' 'Israel' 'Italy' 'Jamaica' 'Japan' 'Jordan' 'Kazakhstan'
 'Kenya' 'Kuwait' 'Latvia' 'Lebanon' 'Lithuania' 'Luxembourg' 'Macao'
 'Malaysia' 'Malta' 'Mauritius' 'Mexico' 'Montenegro' 'Morocco'
 'Mozambique' 'Namibia' 'Nepal' 'Netherlands' 'New Zealand' 'Nigeria'
 'Northern Cyprus' 'Norway' 'Oman' 'Pakistan' 'Palestine' 'Peru'
 'Philippines' 'Poland' 'Portugal' 'Puerto Rico' 'Qatar' 'Romania'
 'Russian Federation' 'Saudi Arabia' 'Serbia' 'Singapore' 'Slovakia'
 'Slovenia' 'South Africa' 'South Korea' 'Spain' 'Sri Lanka' 'Sweden'
 'Switzerland' 'Taiwan' 'Tanzania' 'Thailand' 'Tunisia' 'Turkey' 'Uganda'
 'Ukraine' 'United Arab Emirates' 'United Kingdom' 'United States'
 'Venezuela' 'Vietnam' 'Zambia' 'Zimbabwe']

country2 = []
median2 = []

for x in resdfloc:
    ave = resdf.loc[resdf['location'] == x, 'scores_research'].median()
    ave = round(ave, 2)
    country2.append(x)
    median2.append(ave)


finalresdf = pd.DataFrame(columns = ["location","median"])
finalresdf['location'] = country2
finalresdf['median'] = median2
finalresdf.head()

	location	median
0	Algeria	8.30
1	Argentina	10.20
2	Australia	33.10
3	Austria	28.85
4	Azerbaijan	7.80

finalresdf.sort_values(by=['median'])

	location	median
100	Venezuela	7.80
4	Azerbaijan	7.80
62	Nepal	7.90
61	Namibia	8.00
0	Algeria	8.30
...	...	...
88	Sweden	43.60
7	Belgium	43.85
63	Netherlands	57.10
33	Hong Kong	59.05
81	Singapore	85.45

104 rows × 2 columns

a2df = pd.merge(finalresdf, wipodf, on = 'location')
a2df = pd.DataFrame(pd.merge(finalresdf, wipodf, on = 'location'))
a2df.sort_values(by=['rank'])

	location	median	rank
82	Switzerland	39.75	1
90	United States	27.00	2
81	Sweden	43.60	3
89	United Kingdom	24.70	4
59	Netherlands	57.10	5
...	...	...	...
24	Ethiopia	10.95	117
92	Zambia	9.20	118
86	Uganda	27.70	119
56	Mozambique	8.40	123
37	Iraq	8.45	131

94 rows × 3 columns

a2df.corr(numeric_only = True)

	median	rank
median	1.000000	-0.591913
rank	-0.591913	1.000000

Visualization of the country's innovation ranking against the research score of the universities within that country

plt.scatter(a2df['rank'], a2df['median'])
plt.xlabel("Innovation Rank")
plt.ylabel("Median Teaching Score")
plt.title("Innovation Ranking against the Research Score")
plt.show()

Initial Analysis: There is an inverse correlation between the innovation rank and the median research score of the country. This suggests that the improvement of the research score will lead to a higher ranking in the Global Innovation Index.

Getting the median of each country in terms of citation score

citdf = pd.DataFrame(ranking[['rank','location','scores_citations','name']])
citdf['rank'] = resdf['rank'].str.replace('\W', '', regex=True)
citdf.reset_index(inplace=True)
citdf.dropna(inplace=True)
citdf.tail()

	index	rank	location	scores_citations	name
1794	1794	1501	Japan	13.9	Yamaguchi University
1795	1795	1501	China	18.3	Yanshan University
1796	1796	1501	Turkey	10.5	Yeditepe University
1797	1797	1501	Russian Federation	7.2	Yuri Gagarin State Technical University of Sar...
1798	1798	1501	Turkey	23.5	Zonguldak Bülent Ecevit University

citdfloc = pd.unique(citdf['location'])
citdfloc.sort()
print(citdfloc)

['Algeria' 'Argentina' 'Australia' 'Austria' 'Azerbaijan' 'Bangladesh'
 'Belarus' 'Belgium' 'Botswana' 'Brazil' 'Brunei Darussalam' 'Bulgaria'
 'Canada' 'Chile' 'China' 'Colombia' 'Costa Rica' 'Croatia' 'Cuba'
 'Cyprus' 'Czech Republic' 'Denmark' 'Ecuador' 'Egypt' 'Estonia'
 'Ethiopia' 'Fiji' 'Finland' 'France' 'Georgia' 'Germany' 'Ghana' 'Greece'
 'Hong Kong' 'Hungary' 'Iceland' 'India' 'Indonesia' 'Iran' 'Iraq'
 'Ireland' 'Israel' 'Italy' 'Jamaica' 'Japan' 'Jordan' 'Kazakhstan'
 'Kenya' 'Kuwait' 'Latvia' 'Lebanon' 'Lithuania' 'Luxembourg' 'Macao'
 'Malaysia' 'Malta' 'Mauritius' 'Mexico' 'Montenegro' 'Morocco'
 'Mozambique' 'Namibia' 'Nepal' 'Netherlands' 'New Zealand' 'Nigeria'
 'Northern Cyprus' 'Norway' 'Oman' 'Pakistan' 'Palestine' 'Peru'
 'Philippines' 'Poland' 'Portugal' 'Puerto Rico' 'Qatar' 'Romania'
 'Russian Federation' 'Saudi Arabia' 'Serbia' 'Singapore' 'Slovakia'
 'Slovenia' 'South Africa' 'South Korea' 'Spain' 'Sri Lanka' 'Sweden'
 'Switzerland' 'Taiwan' 'Tanzania' 'Thailand' 'Tunisia' 'Turkey' 'Uganda'
 'Ukraine' 'United Arab Emirates' 'United Kingdom' 'United States'
 'Venezuela' 'Vietnam' 'Zambia' 'Zimbabwe']

country3 = []
median3 = []

for x in citdfloc:
    ave = citdf.loc[citdf['location'] == x, 'scores_citations'].median()
    ave = round(ave, 2)
    country3.append(x)
    median3.append(ave)


finalcitdf = pd.DataFrame(columns = ["location","median"])
finalcitdf['location'] = country3
finalcitdf['median'] = median3
finalcitdf.tail()

	location	median
99	United States	69.8
100	Venezuela	8.9
101	Vietnam	33.3
102	Zambia	96.8
103	Zimbabwe	35.6

a3df = pd.merge(finalcitdf, wipodf, on = 'location')
a3df = pd.DataFrame(pd.merge(finalcitdf, wipodf, on = 'location'))
a3df.sort_values(by=['rank'])

	location	median	rank
82	Switzerland	83.75	1
90	United States	69.80	2
81	Sweden	78.00	3
89	United Kingdom	72.60	4
59	Netherlands	86.60	5
...	...	...	...
24	Ethiopia	89.75	117
92	Zambia	96.80	118
86	Uganda	34.50	119
56	Mozambique	89.40	123
37	Iraq	20.50	131

94 rows × 3 columns

a3df.corr(numeric_only = True)

	median	rank
median	1.00000	-0.16005
rank	-0.16005	1.00000

Visualization of the country's innovation ranking against the citation score of the universities within that country

plt.scatter(a3df['rank'], a3df['median'])
plt.xlabel("Innovation Rank")
plt.ylabel("Median Citation Score")
plt.title("Innovation Ranking against the Citation Score")
plt.show()

Initial Analysis: There is no correlation present between the innovation ranking and the median citation score of a country.

Conclusion

The teaching score can improve by country which can lead to an improve ranking in the Global Innovation Index. Providing students and researchers the space needed for innovation is a crucial as it fosters a culture pushing for national development as well. The same can be said with the research score of each country. Research outputs is a relevant indicator for a country's participation in innovation. Therefore, countries, especially those who reside in developing countries must look into a more active participation in research and development to improve their innovation ranking.

Moreover, the citation score shows no correlation with the innovation ranking of a country which can suggest that the participation of knowledge transfer is an activity that is already expected of each country or academic institution. Making use of that knowledge does not translate in this criteria.

A country's rate of innovation is dependent on the active participation of its education system.

Recommendations

It would be intersting to incorporate socio-economic factors such as income group in future analysis. This may uncover suggestions for developing countries to create policies that places an emphasis on research and development towards innovation.