Italy : COVID-19, cases to 30/03/2020

https://wp.me/p7ciWq-kV

General comment

The momentum data, from 16/03 to 30/03, confirms the dramatic role of the most populous regions of Italy, i.e. Lombardia in the first place and Emilia-Romagna and Piemonte in the second and third, but much lesser extend. Other regions behave quasi similarly , from one to the other. A regression only on them would give a line with a rather low slope.

Notice that the slope of the regression line (on all regions) has increased significantly between the 16th and the 30th of March. The peak of the pandemic in Italy is to be fixed at around the mid of March (see Figura 1 and 2 of the report quoted below in the sources).

This is an analysis based on public data, and subject to revisions or errors including the processing.

Data sources: Epidemia COVID-19 30 marzo 2020 – ore 16:00 – Istituto  Superiore di Sanita (ISS), Roma, and regional data concerning demographics and areas

Analysis I – Cumulative cases to 30/03/2020

Analysis

Multiple Regression – COVID-19 Cases 30/03/20
Dependent variable: COVID-19 Cases 30/03/20
Independent variables:
Population (01-2019)
hab/km²

.                                                                    Standard             T
Parameter                       Estimate                Error         Statistic           P-Value
CONSTANT                       -3314.87               472.75         -1.34056         0.1967
Population (01-2019 ) 0.00254471     0.000633961        4.01398        0.0008

Analysis of Variance
Source                            Sum of Squares            Df     Mean Square F-Ratio P-Value
Model                                 7.90733E8                  1            7.90733E8    16.11    0.0008
Residual                             8.83389E8               18             4.90772E7
Total (Corr.)                      1.67412E9                19

R-squared = 47.2327 percent
R-squared (adjusted for d.f.) = 44.3012 percent
Standard Error of Est. = 7005.51
Mean absolute error = 4900.97
Durbin-Watson statistic = 1.8902 (P=0.4265)
Lag 1 residual autocorrelation = 0.012704

Stepwise regression
Method: forward selection
P-to-enter: 0.05
P-to-remove: 0.05

Chart

Analysis II – Cases from 16/03/2020 to 30/03/2020

Analysis

Multiple Regression – COVID-19 Cases 30/03/20-COVID-19 Cases 16/3/2020
Dependent variable: COVID-19 Cases 30/03/20-COVID-19 Cases 16/3/2020
Independent variables:
Population (01-2019)
km²

.                                                                                Standard                T
Parameter                                  Estimate                 Error            Statistic          P-Value
CONSTANT                                – 1817.49                 1708.3           -1.06391          0.3014
Population (01-2019)           0.00163341           0.000437973       3.72947          0.0015

Analysis of Variance
Source                               Sum of Squares         Df            Mean Square   F-Ratio     P-Value
Model                                    3.25794E8                1                 3.25794E8         13.91     0.0015
Residual                                  4.2162E8             18                  2.34233E7
Total (Corr.) 7.47414E8 19

R-squared = 43.5895 percent
R-squared (adjusted for d.f.) = 40.4556 percent
Standard Error of Est. = 4839.77
Mean absolute error = 3436.17
Durbin-Watson statistic = 1.88585 (P=0.4226)
Lag 1 residual autocorrelation = -0.0204147

Stepwise regression
Method: forward selection

Chart

Advertisement