Transcript EDUCATIONAL MEASUREMENT IN ITALIAN UNIVERSITIES and …

EDUCATIONAL MEASUREMENT IN ITALIAN
UNIVERSITIES and THE UNIVERSITY
RATING EVOLUTION
by
Raimondo Manca
SOME HISTORY
• In the second half of the 19° century Italy was unified
– There was a high rate of illitteralicy
– Big differences between the north and south
• Two big intuitions at the end of the century
– Montessori’s method
• The teacher should adequate to the scholars
– Creation of universities devoted to the preparation of teachers for
primary school
• Italian primary school was very good and the Montessori
method was adopted in many countries
CRISIS OF ITALIAN SCHOOL
• Strong increase of university students in seventies
• Many people became University professors without the
quality
• The school teachers were paid less
• Teaching was considered a second best
• More and more students that are not well prepared arrive
at university
• A vicious circle was created
ACTUAL SITUATION
• The Faculties that prepared primary school teachers
were cancelled and were transformed in something more
general (Educational Sciences)
• There are 36 faculties of Educational Sciences
– 16 statistics or psicometry professors (associate or full)
– 9 assistant professors
• Educational measurement is not yet well established in
Italy
COUNTRIES OF FIRST 100 UNIVERSITIES
Criteria
Quality of Education
Quality of Faculty
Research Output
Per Capita Performance
Total
Indicator
Code
Weight
Alumni of an institution winning Nobel Prizes and Fields Medals
Alumni
10%
Staff of an institution winning Nobel Prizes and Fields Medals
Award
20%
Highly cited researchers in 21 broad subject categories
HiCi
20%
Papers published in Nature and Science*
N&S
20%
Papers indexed in Science Citation Index-expanded and Social Science Citation Index
PUB
20%
Per capita academic performance of an institution
PCP
10%
100%
FIRST 20 WORLD UNIVERSITY
World Rank
Institution*
Region
Regional Rank
National Rank
1
Harvard University
Americas
1
1
2
University of California, Berkeley
Americas
2
2
3
Stanford University
Americas
3
3
4
Massachusetts Institute of Technology (MIT)
Americas
4
4
5
University of Cambridge
Europe
1
1
6
California Institute of Technology
Americas
5
5
7
Princeton University
Americas
6
6
8
Columbia University
Americas
7
7
9
University of Chicago
Americas
8
8
10
University of Oxford
Europe
2
2
11
Yale University
Americas
9
9
12
Cornell University
Americas
10
10
13
University of California, Los Angeles
Americas
11
11
14
University of California, San Diego
Americas
12
12
15
University of Pennsylvania
Americas
13
13
16
University of Washington
Americas
14
14
17
University of Wisconsin - Madison
Americas
15
15
18
The Johns Hopkins University
Americas
16
16
18
University of California, San Francisco
Americas
16
17
20
The University of Tokyo
Asia/Pacific
1
1
UNIVERSITY RATING TIME EVOLUTION I
• Hypotheses
– Homogeneity
– Semi-Markov process
– Discrete time (DTHSMP)
• Steps
–
–
–
–
–
State subdivision
Embedded Markov Chain
Waiting time distribution function
Model application
Results
STATE SUBDIVISION
• State set:
–
–
–
–
–
–
–
–
–
–
State 1
1-20
State 2 21-40
State 3 41-70
State 4 71-100
State 5 101-150
State 6 151-200
State 7 201-300
State 8 301-400
State 9 401-500
State 10 No Rating
INPUT CONSTRUCTION
• Embedded Markov Chain construction
– Nij Number of transitions from the state i to the state j
– N Number of transition from the state i
i
–
pij 
N ij
Ni
Probability to go from state i to state j
• Waiting time distribution function
– Nij (t ) number of transition from i to j within a time t
– Fij (t ) 
Nij (t )
Nij
probability to have a transition in a time ≤ t
HSMP 1
Xn :   I
Tn :  
 X n ,Tn 
Qij (t )  P  X n1  j,Tn1  Tn  t X n  i 
pij  limQij (t ); i, j  I , t 
t 
Hi (t )  P  Tn1  Tn  t X n  i 
m
H i (t )   Qij (t )
j 1
HSMP 2
Fij (t )  P  Tn1  Tn  t X n1  j, X n  i 

Qij (t ) / pij
Fij (t )  
1


if
if
pij  0
pij  0
0
if


bij (t )  P  X n1  j, Tn1  Tn  t X n  i   

Qij (t )  Qij (t  1) if
m
t
ij (t )   ij (1  H i (t ))   bi ( ) j (t   )
 1  1
ij (1  Hi (t ))
m
t
b  ( )

 
1 1
i
j
(t   )
t 0
t 0
CONCLUSIONS
•
•
•
•
•
Universities
Private universities
Countries
Data collection for the first 6000 world universities
Construction of input data