4 Figure 16 gives the results of a stepwise regression analysis for tendering costs. This model explained 75 per cent of the variations in tendering costs (adjusted R2) and in view of the extent of variance accounted for, statistically insignificant variables are not shown from the regression which entered all variables. The results show that PFI sector was associated with higher tendering costs, with tendering for PFI hospitals and 'other' projects costing more than PFI schools. There was a strong link with the capital value of the projects even allowing for PFI sector, reflecting factors such as the scale of preparation of output specification which increased with capital value.
16 | multiple regression coefficients and significance values to explain the cost of PFI projects | |||||||
| Unstandardized Coefficients | Standardized Coefficients |
|
| ||||
| B | Std. Error | Beta | t | Sig. | |||
|
|
|
|
|
| |||
(Constant) | 391,384.810 | 623,253.731 |
| 0.628 | 0.535 | |||
Hospital | 726,643.044 | 858,562.944 | 0.085 | 0.846 | 0.405 | |||
Capital value | 0.009 | 0.001 | 0.671 | 6.823 | 0.000 | |||
Other sector | 3,928,306.291 | 1,107,705.456 | 0.386 | 3.546 | 0.001 | |||
Source: National Audit Office |
|
|
|
| ||||