Link to current text
MMP.16:3 - Forces
| Force | Tension |
|---|
| Visible contrast and recorded contrast | Models can disagree about an unobserved quantity while predicting the same available record. |
| Model identity and receiving result | Different models may support the same answer; a single model family may contain the important disagreement. |
| Discrimination and uncertainty | One certain separating observation is attractive, but useful statistical distinctions often retain error. |
| Informative design and model adequacy | A design can be excellent within an inadequate model family. |
| Learning and available work | A better distinction can cost more time, access or computation than its contribution warrants. |
| Immediate gain and continued inquiry | One observation may prepare a later useful test without answering the final question itself. |