G2 Science K223, K224 and K225 often teaches variables one at a time because that makes mechanisms easier to see. Real experimental data can still show an important next idea: the effect of one factor may depend on the level of another factor. This is an interaction effect.
This one-hundred-and-seventh Learner’s Guide develops interaction reasoning without importing advanced statistical machinery. It extends Vol 0099 Control-Condition Logic and Vol 0103 Model-Validity Boundaries.
The interaction question
Ask: if factor A changes, does the size or direction of its effect stay the same at different levels of factor B? If not, describe the effect conditionally: “A increases the outcome when B is low, but has little additional effect when B is high,” for example.
1. light × carbon dioxide
Interaction structure: photosynthesis response to light depends on carbon dioxide availability.
Reasoning move: effect of one factor can shrink when another becomes limiting. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
2. light × temperature
Interaction structure: photosynthesis rate may respond differently to light at different temperatures.
Reasoning move: avoid one-factor universal claim. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
3. carbon dioxide × temperature
Interaction structure: CO2 effect can depend on enzyme-controlled temperature conditions.
Reasoning move: compare combinations. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
4. enzyme × temperature
Interaction structure: enzyme presence/activity interacts with temperature.
Reasoning move: temperature effect depends on functional enzyme. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
5. enzyme × pH
Interaction structure: pH can alter enzyme function.
Reasoning move: effect is conditional on enzyme system. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
6. substrate × enzyme amount
Interaction structure: substrate effect depends on available enzyme sites.
Reasoning move: saturation can change interaction. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
7. concentration × temperature
Interaction structure: reaction rate responds to both.
Reasoning move: combined effect may not be simple addition. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
8. surface area × concentration
Interaction structure: collision opportunities change through two routes.
Reasoning move: hold one factor fixed when estimating the other. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
9. catalyst × temperature
Interaction structure: both affect reaction rate.
Reasoning move: catalyst effect may differ across temperature range. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
10. pressure × temperature gas context
Interaction structure: gas behaviour depends on multiple state variables.
Reasoning move: one-variable prediction needs fixed conditions. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
11. volume × temperature gas context
Interaction structure: change in one may be offset by another.
Reasoning move: state what is held fixed. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
12. force × mass
Interaction structure: acceleration response to force depends on mass.
Reasoning move: same force gives different acceleration across masses. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
13. force × friction
Interaction structure: motion response to applied force depends on opposing friction.
Reasoning move: applied force alone may not predict acceleration. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
14. slope × friction
Interaction structure: motion down slope depends on gravitational component and friction.
Reasoning move: interaction determines net effect. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
15. current × resistance
Interaction structure: circuit current depends on potential difference and resistance.
Reasoning move: effect of voltage is conditional on resistance. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
16. voltage × resistance
Interaction structure: same voltage produces different current at different resistance.
Reasoning move: compare matched cases. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
17. component arrangement × resistance
Interaction structure: series/parallel arrangement changes total behaviour.
Reasoning move: component value effect depends on topology. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
18. current × time heating
Interaction structure: thermal effect depends on electrical input and duration.
Reasoning move: short exposure can hide effect. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
19. light intensity × distance
Interaction structure: measured light effect depends on geometry/distance.
Reasoning move: do not attribute all change to source strength. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
20. sound level × distance
Interaction structure: detected signal depends on source and detector distance.
Reasoning move: control geometry. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
21. magnetic strength × distance
Interaction structure: response depends on both.
Reasoning move: effect of stronger magnet can be masked by greater distance. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
22. load × material
Interaction structure: extension depends on applied load and material properties.
Reasoning move: same load yields different response. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
23. load × length
Interaction structure: extension may depend on specimen dimensions.
Reasoning move: hold geometry fixed for material comparison. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
24. load × cross-section
Interaction structure: mechanical response depends on geometry.
Reasoning move: avoid single-factor attribution. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
25. plant water × light
Interaction structure: growth/photosynthesis outcome can depend on both resources.
Reasoning move: one resource may limit effect of another. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
26. plant water × temperature
Interaction structure: water loss/growth response depends on environmental temperature.
Reasoning move: interaction can alter net outcome. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
27. humidity × airflow
Interaction structure: transpiration responds to both.
Reasoning move: effect of airflow can differ at different humidity. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
28. leaf area × airflow
Interaction structure: whole-plant water loss depends on exposed area and air movement.
Reasoning move: normalise or control area. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
29. stomatal opening × humidity
Interaction structure: local regulation and environment interact.
Reasoning move: macroscopic loss is conditional. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
30. seed water × temperature
Interaction structure: germination requires suitable combination.
Reasoning move: one adequate factor cannot compensate for another missing requirement. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
31. seed oxygen × water
Interaction structure: both conditions matter.
Reasoning move: absence of either can block germination. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
32. nutrient × water
Interaction structure: growth response to nutrients depends on water availability.
Reasoning move: limiting factor can switch. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
33. predator × prey abundance
Interaction structure: predation effect depends on both populations.
Reasoning move: simple one-direction relation may change over time. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
34. resource × population size
Interaction structure: per-capita resource depends on both.
Reasoning move: population response is conditional. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
35. pollutant × exposure time
Interaction structure: effect depends on concentration and duration.
Reasoning move: dose-like interaction. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
36. pollutant × organism type
Interaction structure: same exposure can affect species differently.
Reasoning move: population/species is interacting factor. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
37. temperature × organism type
Interaction structure: response differs across organisms.
Reasoning move: do not generalise one species. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
38. drug/treatment analogy × dose
Interaction structure: response can depend on amount.
Reasoning move: school-science reasoning should stay within given context. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
39. treatment × baseline
Interaction structure: same intervention can produce different visible change from different starting levels.
Reasoning move: ceiling/floor effects. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
40. treatment × time
Interaction structure: effect can emerge or disappear over time.
Reasoning move: single endpoint may miss interaction. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
41. treatment × age/stage
Interaction structure: response may differ by developmental stage.
Reasoning move: population scope matters. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
42. factor × sex/group where provided
Interaction structure: response can differ across defined groups.
Reasoning move: do not infer unless data include the grouping. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
43. factor × genotype where provided
Interaction structure: biological response may depend on inherited variant.
Reasoning move: use only when syllabus/context supports. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
44. independent variable × control variable drift
Interaction structure: apparent effect includes uncontrolled interaction.
Reasoning move: design must separate. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
45. two manipulated factors
Interaction structure: factorial-style comparison has four combinations.
Reasoning move: compare simple effects at each level. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
46. A absent/B absent
Interaction structure: baseline combination.
Reasoning move: anchors interaction. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
47. A present/B absent
Interaction structure: isolates A at one B level.
Reasoning move: compare with baseline. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
48. A absent/B present
Interaction structure: isolates B at one A level.
Reasoning move: compare with baseline. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
49. A present/B present
Interaction structure: combined condition.
Reasoning move: compare with expected simple addition cautiously. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
50. additive effect
Interaction structure: combined change roughly equals sum of separate changes.
Reasoning move: interaction may be small under measured scale. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
51. synergistic effect
Interaction structure: combined effect exceeds simple separate expectation.
Reasoning move: requires data; do not infer from adjectives. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
52. antagonistic effect
Interaction structure: one factor reduces effect of another.
Reasoning move: combined outcome is less than simple expectation. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
53. masking
Interaction structure: two effects oppose and net change looks small.
Reasoning move: null result may hide active mechanisms. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
54. ceiling interaction
Interaction structure: one condition already near maximum.
Reasoning move: second factor appears ineffective. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
55. floor interaction
Interaction structure: one condition near minimum.
Reasoning move: further decrease cannot be observed. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
56. threshold interaction
Interaction structure: factor matters only after another crosses boundary.
Reasoning move: simple average can hide condition. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
57. saturation interaction
Interaction structure: response to A flattens at high B or vice versa.
Reasoning move: limiting factor shifts. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
58. limiting-factor switch
Interaction structure: A limits at one condition, B at another.
Reasoning move: state range. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
59. time-lag interaction
Interaction structure: one factor changes timing of another’s effect.
Reasoning move: compare time courses. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
60. feedback interaction
Interaction structure: response changes the factor driving it.
Reasoning move: one-way model may fail. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
61. competition interaction
Interaction structure: two processes use same resource.
Reasoning move: effect of one depends on strength of other. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
62. compensation
Interaction structure: one process offsets another.
Reasoning move: stable output can hide internal change. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
63. control × treatment
Interaction structure: treatment effect is defined relative to control.
Reasoning move: control condition is part of causal interpretation. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
64. measurement × range
Interaction structure: instrument sensitivity interacts with effect size.
Reasoning move: small effects can disappear below resolution. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
65. measurement × timing
Interaction structure: detectability depends on when measured.
Reasoning move: negative result can be timing-dependent. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
66. measurement × location
Interaction structure: local sensor may miss spatial variation.
Reasoning move: site interacts with system gradient. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
67. sample × condition
Interaction structure: different samples respond differently.
Reasoning move: randomisation/matching matters. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
68. sample size × variability
Interaction structure: ability to see pattern depends on variation and n.
Reasoning move: small sample can hide interaction. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
69. normalisation × group size
Interaction structure: rate comparison depends on denominator.
Reasoning move: raw count interaction may differ from rate. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
70. graph interaction
Interaction structure: lines for two conditions are non-parallel or cross.
Reasoning move: effect of x differs by group/condition. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
71. parallel trends
Interaction structure: difference stays similar across x.
Reasoning move: interaction may be limited under plotted range. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
72. crossing trends
Interaction structure: direction reverses across conditions.
Reasoning move: one global effect statement is misleading. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
73. separate slopes
Interaction structure: rate of change differs by condition.
Reasoning move: compare slopes. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
74. same endpoint different path
Interaction structure: final values match but time courses differ.
Reasoning move: endpoint alone hides interaction. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
75. different endpoint same early path
Interaction structure: interaction emerges later.
Reasoning move: time window matters. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
76. one-factor-at-a-time experiment
Interaction structure: easy to interpret main effect.
Reasoning move: cannot reveal all interactions efficiently. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
77. multi-factor experiment
Interaction structure: can expose interaction.
Reasoning move: requires careful condition ownership. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
78. interaction mistaken for noise
Interaction structure: variation follows a second variable.
Reasoning move: stratify/compare conditions before dismissing. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
79. interaction mistaken for causation
Interaction structure: groups differ on two factors.
Reasoning move: cannot attribute to one factor cleanly. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
80. interaction and confounding
Interaction structure: uncontrolled factor changes with treatment.
Reasoning move: effect is not separable. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
81. interaction and control
Interaction structure: hold B fixed to estimate A effect.
Reasoning move: then repeat at another B level to test interaction. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
82. interaction and prediction
Interaction structure: model predicts same A effect everywhere.
Reasoning move: data show different A effects. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
83. interaction and mechanism
Interaction structure: mechanism explains why B changes A effect.
Reasoning move: stronger than pattern alone. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
84. interaction and negative evidence
Interaction structure: A shows no effect at one B level.
Reasoning move: do not conclude A never matters. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
85. interaction and generalisation
Interaction structure: effect measured in one condition.
Reasoning move: scope conclusion to that condition. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
86. interaction and exception
Interaction structure: counterexample occurs under different B.
Reasoning move: boundary may explain exception. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
87. interaction and model validity
Interaction structure: simple model works only when B fixed.
Reasoning move: state assumption. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
88. interaction final rule
Interaction structure: ask whether the effect of A stays the same when B changes.
Reasoning move: if not, describe the conditional effect rather than one universal main effect. For practice, build a four-cell comparison where possible: A-/B-, A+/B-, A-/B+, A+/B+. Compare the effect of A at each B level rather than looking only at the grand average.
Do not claim interaction merely because two factors both matter. Interaction means the effect of one changes with the other. If both effects are stable and simply add, a simpler model may be sufficient.
Interactions explain many apparent exceptions
A rule can appear to fail because a second factor changed. Light may stop increasing photosynthesis when another factor becomes limiting; a treatment may appear ineffective at a ceiling; a force may not accelerate an object when opposing forces change. Interaction reasoning turns exceptions into testable boundary conditions.
Links
Use the Science Hub, Vol 0071 Competing Mechanisms, Vol 0095 Scale Translation, 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. Interaction-effect reasoning here is an eduKateSengkang framework for interpreting multi-factor situations, not an additional SEAB statistical syllabus topic.
Final rule
Do not ask only whether A matters and whether B matters. Ask whether A matters in the same way when B changes. When the answer is no, state the condition-dependent effect instead of forcing one universal rule.
