Fig. 8

Shenzhen innovation system centres around university-company partnership and own R&D. The figure shows the results of surveys, of 167 Shenzhen electronics firms, asking about the importance of various methods for producing electronics-related innovations. These results represent the b-value coefficients for a logit regression. In other words, a ten unit increase in a Shenzhen electronics firms ow n innovation results in 8 units of "innovative performance" (whatever that is). Intergroup training and training from the customer result in harms rather than been its. In Shenzhen, unlike in other Chinese cities, the ownership structure, firm size, extent of foreign led innovation, training from a multinational’s headquarters and training done via Guanxi relations have no statistically significant effect on innovative performance