G2 Science K223, K224 and K225 experimental reasoning can fail when the reference condition itself changes. A baseline is not automatically stable because it was measured first. Temperature, instruments, organisms, reagents and environments can drift over time.
This one-hundred-and-fifteenth Learner’s Guide develops baseline-drift control. It extends Vol 0099 Control-Condition Logic and Vol 0103 Model-Validity Boundaries.
The baseline-drift question
Ask what would have happened to the measured outcome over time if the treatment had never been applied. A concurrent control, repeated baseline or calibration check can reveal whether the reference process is moving.
1. temperature baseline
For temperature baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
2. light baseline
For light baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
3. humidity baseline
For humidity baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
4. airflow baseline
For airflow baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
5. background sound
For background sound, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
6. background light
For background light, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
7. instrument zero
For instrument zero, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
8. balance zero
For balance zero, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
9. thermometer calibration
For thermometer calibration, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
10. pH calibration
For pH calibration, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
11. voltage supply
For voltage supply, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
12. battery condition
For battery condition, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
13. lamp output
For lamp output, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
14. reagent strength
For reagent strength, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
15. enzyme preparation
For enzyme preparation, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
16. organism activity
For organism activity, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
17. plant water status
For plant water status, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
18. seed viability
For seed viability, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
19. room-temperature trend
For room-temperature trend, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
20. time-of-day biology
For time-of-day biology, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
21. practice or fatigue
For practice or fatigue, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
22. apparatus warming
For apparatus warming, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
23. evaporation baseline
For evaporation baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
24. natural cooling
For natural cooling, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
25. natural heating
For natural heating, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
26. control-group drift
For control-group drift, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
27. treatment pre-trend
For treatment pre-trend, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
28. different starting baselines
For different starting baselines, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
29. different baseline slopes
For different baseline slopes, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
30. historical baseline
For historical baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
31. concurrent baseline
For concurrent baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
32. interleaved control
For interleaved control, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
33. randomised order
For randomised order, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
34. counterbalanced order
For counterbalanced order, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
35. repeat baseline
For repeat baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
36. calibration before
For calibration before, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
37. calibration after
For calibration after, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
38. calibration during
For calibration during, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
39. blank over time
For blank over time, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
40. positive control over time
For positive control over time, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
41. negative control over time
For negative control over time, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
42. sensor warm-up
For sensor warm-up, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
43. sample settling
For sample settling, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
44. reaction already started
For reaction already started, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
45. delayed measurement
For delayed measurement, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
46. unequal exposure duration
For unequal exposure duration, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
47. sequential sampling
For sequential sampling, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
48. environmental trend
For environmental trend, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
49. batch effect
For batch effect, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
50. day effect
For day effect, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
51. operator effect
For operator effect, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
52. observer calibration
For observer calibration, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
53. camera exposure
For camera exposure, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
54. scale drift
For scale drift, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
55. baseline subtraction
For baseline subtraction, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
56. difference-in-change
For difference-in-change, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
57. ratio-to-baseline
For ratio-to-baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
58. percentage change
For percentage change, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
59. rolling baseline
For rolling baseline, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
60. baseline noise
For baseline noise, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
61. baseline trend
For baseline trend, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
62. baseline step change
For baseline step change, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
63. baseline outlier
For baseline outlier, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
64. drift mistaken for treatment
For drift mistaken for treatment, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
65. treatment mistaken for drift
For treatment mistaken for drift, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
66. shared drift
For shared drift, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
67. opposite drift
For opposite drift, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
68. drift plus interaction
For drift plus interaction, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
69. drift plus ceiling
For drift plus ceiling, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
70. drift plus floor
For drift plus floor, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
71. drift and rate
For drift and rate, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
72. drift and amount
For drift and amount, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
73. drift and model
For drift and model, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
74. drift and conclusion
For drift and conclusion, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
75. drift and improvement
For drift and improvement, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
76. baseline drift final rule
For baseline drift final rule, identify the reference quantity that is assumed to stay stable. Then ask what physical, biological, environmental, instrumental or procedural process could move that reference even without the treatment.
For practice, draw a baseline trajectory and a treatment trajectory across the same time axis. Decide whether the relevant treatment effect is the final difference, the change from each group’s start, or the difference between changes. State why that comparison controls the drift more effectively.
Do not automatically subtract the first reading. If the baseline continues moving, one initial correction can leave later bias. Match the reference to the time, apparatus and condition of the treatment measurement, and use concurrent controls or repeated calibration where the design allows.
Baseline drift is a time-dependent confounder
When treatments are applied sequentially, time can become correlated with condition. Randomising or counterbalancing order, interleaving controls and recalibrating can separate treatment from session drift. The correct repair depends on what is drifting.
Stable-looking data can still contain drift
A slow drift can be hidden by noisy measurements, rounding or a short observation window. Conversely, random fluctuation should not automatically be labelled drift. Look for a systematic time pattern, repeated reference measurements and a plausible process that could move the baseline.
Links
Use the Science Hub, Vol 0111 Rate Versus Amount, Vol 0083 Diagnostic-Test Evidence, the Examination Craft hub and the PSLE Learner’s Guide.
Official-source discipline
For the current 2027 SEC G2 school-candidate framework, use the official SEAB G2 syllabus directory and linked K223–K225 Science syllabuses. Baseline-drift control is an eduKateSengkang experimental-reasoning framework, not an additional SEAB syllabus topic.
Final rule
A baseline is not merely the first number. It is the reference process that would have happened without the treatment. Check whether that process is stable before attributing every later difference to the factor you changed.
