Table 5. Factor analysis of all GP views on clusters.
| Cluster item | Factor 1a | Factor 2a |
|---|---|---|
| Informed about cluster | 0.808 | 0.229 |
| Decisions reflect my views | 0.828 | 0.243 |
| PQL responsive to queries or concerns | 0.688 | 0.157 |
| Feel ‘ownership’ of cluster | 0.813 | 0.251 |
| Can influence cluster work | 0.805 | 0.246 |
| Quality planning | 0.299 | 0.822 |
| Quality improvement | 0.270 | 0.862 |
| Quality control | 0.236 | 0.873 |
| Local population needs | 0.292 | 0.731 |
| Quality of care | 0.238 | 0.718 |
| Shared decision making | 0.080 | 0.707 |
Bartlett's test of sphericity = P<0.001.
KMO measure of sampling adequacy = 0.908.
PQL = practice quality lead.
aHigh factor loadings identify a single construct, which means the items within it can be legitimately aggregated to give a total score.