Table 2. Open in new tab Predictive...

Table 2.

Predictive Performance Using Mean Squared Error and R² of the Third Pillar Time for the Total Model and the Paid Staff Model

		Mean	SD	Min	Max
Mean squared error	Total	6.1787	0.5511	5.1037	7.0633
	Paid staff	4.2037	0.5205	3.3196	4.9077
R²	Total	0.7829	0.1736	0.3873	0.9707
	Paid staff	0.7831	0.1833	0.3859	0.9877

		Mean	SD	Min	Max
Mean squared error	Total	6.1787	0.5511	5.1037	7.0633
	Paid staff	4.2037	0.5205	3.3196	4.9077
R²	Total	0.7829	0.1736	0.3873	0.9707
	Paid staff	0.7831	0.1833	0.3859	0.9877

Source: Authors’ analysis estimating third pillar size using the Random Forest model. The model includes data extracted from the international Johns Hopkins University dataset, the European Union Johns Hopkins University dataset, the International Labour Organization labor-market dataset, the World Development Indicators, and the Polity5 Project dataset.

Note: Total includes paid staff and volunteers. Paid staff includes paid staff only. The models include the 51 countries with complete data (labor-market structure, the level of economic development, the political regime, and third pillar size).

This Feature Is Available To Subscribers Only