G2 Science Model-Validity Boundaries | Learner’s Guide Vol 0103

G2 Science K223, K224 and K225 uses models everywhere: particles, rays, circuits, forces, cells, populations, rates, graphs and experimental comparisons. A model is useful because it simplifies. The same simplification also creates a boundary.

This one-hundred-and-third Learner’s Guide develops model-validity boundaries: know what a model explains, what assumptions it needs, where it applies and what observation would force revision. It extends Vol 0095 Scale Translation and Vol 0087 Converging Evidence.

The model-validity audit

For any model, ask four questions: what does it represent, what does it deliberately ignore, under what conditions should it work, and what observation would count against it? This prevents learners from treating a useful model as literal reality or universal law.

1. particle model

Model: particles represent matter at microscopic scale.

Boundary: use for spacing/motion explanations, not literal visible paths of every particle. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

When data disagree, do not immediately choose between “the model is wrong” and “the experiment is wrong”. Check assumptions, measurement validity, range, scale and controls. Model evaluation is a comparison between a prediction and a fair test of that prediction.

2. idealised diagram

Model: diagram simplifies shape or arrangement.

Boundary: do not infer unlabelled scale or exact geometry. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

3. ray model

Model: light paths represented by rays.

Boundary: use for geometric propagation, not every wave property. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

4. circuit model

Model: components represented ideally.

Boundary: real internal resistance or heating may be omitted unless included. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

5. force diagram

Model: selected forces represented.

Boundary: valid only if all relevant forces for the chosen object are included. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

6. energy model

Model: energy stores/transfers simplified.

Boundary: do not invent untracked energy loss without model/context. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

7. collision model

Model: reaction rate explained through collisions.

Boundary: valid when factors change collision frequency/energy as described. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

8. enzyme lock/key-style model

Model: shape compatibility simplified.

Boundary: use within syllabus explanation without treating model as literal rigid machinery. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

9. cell diagram

Model: organelles shown schematically.

Boundary: relative sizes/positions may be simplified. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

10. organ-system diagram

Model: connections simplified.

Boundary: do not infer exact anatomy from schematic layout. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

11. food chain

Model: linear feeding relation shown.

Boundary: real ecosystems may form networks. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

12. food web

Model: multiple feeding links represented.

Boundary: still a selected subset of ecosystem interactions. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

13. population model

Model: counts/rates represent group behaviour.

Boundary: individual variation may be hidden. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

14. graph model

Model: relationship represented over measured range.

Boundary: do not extrapolate automatically. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

15. straight-line fit

Model: linear approximation used.

Boundary: valid only where data support near-linear relation. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

16. proportional model

Model: constant ratio assumed.

Boundary: test whether range/conditions preserve proportionality. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

17. inverse model

Model: constant product assumed.

Boundary: check positive domain and range. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

18. constant-rate model

Model: rate assumed stable.

Boundary: fails if rate changes with time or quantity. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

19. uniform motion

Model: speed assumed constant.

Boundary: do not use across acceleration interval. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

20. no-friction model

Model: friction neglected.

Boundary: valid only when explicitly simplified or negligible for task. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

21. closed-system model

Model: matter/energy boundary specified.

Boundary: conservation inference depends on boundary. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

22. open-system model

Model: matter can enter/leave.

Boundary: mass change may reflect transfer rather than creation/destruction. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

23. fixed-temperature model

Model: temperature held constant.

Boundary: do not apply if process changes temperature materially. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

24. fixed-pressure model

Model: pressure assumed constant.

Boundary: gas relation may change if pressure varies. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

25. fixed-volume model

Model: volume constant.

Boundary: do not apply when container/system expands. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

26. dilute-solution approximation

Model: concentration relations simplified.

Boundary: range matters. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

27. same-density assumption

Model: density treated constant.

Boundary: may fail across state/temperature changes. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

28. same-gravity assumption

Model: g treated constant locally.

Boundary: appropriate for ordinary school-scale contexts. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

29. negligible-air-resistance assumption

Model: drag ignored.

Boundary: model becomes weaker when drag is important. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

30. point-object assumption

Model: size ignored.

Boundary: invalid if dimensions/orientation matter. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

31. rigid-body assumption

Model: shape treated fixed.

Boundary: fails for deformation problems. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

32. perfect-conductor simplification

Model: wire resistance ignored.

Boundary: real circuits may differ. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

33. ideal-meter assumption

Model: meter does not alter circuit materially.

Boundary: measurement device effects usually neglected. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

34. uniform-material assumption

Model: properties treated same throughout.

Boundary: fails if material is layered/heterogeneous. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

35. representative-sample assumption

Model: sample reflects population.

Boundary: requires appropriate sampling. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

36. independent-trial assumption

Model: one trial does not affect another.

Boundary: fails without replacement or shared changing conditions. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

37. random-error model

Model: variation assumed unsystematic.

Boundary: systematic bias is different. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

38. mean-as-centre model

Model: average summarises group.

Boundary: can hide skew, subgroups or outliers. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

39. single-variable causal model

Model: one factor manipulated.

Boundary: other causal factors must be controlled or acknowledged. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

40. dose-response model

Model: output changes with input.

Boundary: may plateau, threshold or reverse outside range. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

41. threshold model

Model: response appears after boundary.

Boundary: below-threshold negatives do not imply no relation. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

42. saturation model

Model: response reaches maximum.

Boundary: extra input gives little additional output. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

43. limiting-factor model

Model: one factor constrains rate.

Boundary: limiting factor can change as conditions change. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

44. feedback model

Model: system response influences its own driver.

Boundary: simple one-way cause may fail. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

45. homeostasis model

Model: system regulates internal condition.

Boundary: valid around operating range, not unlimited disturbance. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

46. adaptation model

Model: response changes over time.

Boundary: short-term and long-term predictions differ. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

47. equilibrium-style model

Model: opposing processes balance macroscopically.

Boundary: microscopic activity may continue. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

48. diffusion model

Model: net movement follows concentration difference.

Boundary: random motion exists in both directions. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

49. osmosis model

Model: water movement across partially permeable membrane.

Boundary: requires correct membrane/solution context. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

50. photosynthesis model

Model: light/carbon dioxide/water contribute to process.

Boundary: net gas observation also includes respiration. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

51. respiration model

Model: cells release energy from food.

Boundary: whole-organism gas exchange depends on transport and activity. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

52. transpiration model

Model: water loss linked to stomata/environment.

Boundary: multiple environmental factors interact. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

53. enzyme-rate model

Model: rate depends on conditions.

Boundary: optimum/range and denaturation boundaries matter. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

54. growth model

Model: size change used as outcome.

Boundary: growth can involve multiple biological processes. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

55. ecological interaction model

Model: one species affects another.

Boundary: network effects and time lags can alter simple prediction. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

56. inheritance model

Model: genetic factors relate to traits.

Boundary: environment and multiple genes may matter depending context. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

57. test-indicator model

Model: signal represents target condition.

Boundary: false positives/negatives limit inference. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

58. proxy model

Model: measured quantity stands for another.

Boundary: valid only if proxy-target relation is established. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

59. calibration model

Model: instrument reading maps to true quantity.

Boundary: valid near calibrated range and functioning state. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

60. linear scale

Model: equal visual intervals represent equal numeric increments.

Boundary: log/nonlinear scales would require different reading. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

61. diagram not to scale

Model: geometry schematic conveys relations not measurements.

Boundary: use stated values/theorems. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

62. table aggregation

Model: rows/columns summarise cases.

Boundary: grouping can hide individual variation. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

63. percentage model

Model: relative share summarises count.

Boundary: denominator determines meaning. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

64. rate model

Model: quantity per base unit.

Boundary: changing denominator can change interpretation. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

65. average-speed model

Model: total distance/total time.

Boundary: not simple mean of segment speeds unless conditions justify. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

66. probability model

Model: sample space represents possible outcomes.

Boundary: valid only if universe and weights are correctly specified. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

67. random sampling model

Model: sample intended to represent population.

Boundary: selection bias breaks generalisation. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

68. control-group model

Model: difference attributed to treatment.

Boundary: valid only if groups are otherwise comparable and measurement works. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

69. counterfactual model

Model: remove cause and predict change.

Boundary: alternative pathways can preserve effect. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

70. negative-evidence model

Model: missing expected signal weakens hypothesis.

Boundary: strength depends on detection capability. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

71. converging-evidence model

Model: independent clues strengthen conclusion.

Boundary: duplicated failure modes reduce independence. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

72. scale-translation model

Model: local mechanisms aggregate to system outcome.

Boundary: bridge may involve feedback/competition. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

73. time-window model

Model: relationship described in selected interval.

Boundary: do not extend outside interval automatically. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

74. population-scope model

Model: claim applies to studied group.

Boundary: do not universalise. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

75. boundary-condition model

Model: rule holds under stated conditions.

Boundary: exception outside boundary may not falsify bounded rule. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

76. model-data mismatch

Model: prediction and observation differ.

Boundary: test assumption, measurement and model range separately. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

77. model success

Model: prediction matches data.

Boundary: support model within tested conditions, not universal truth. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

78. model failure once

Model: one mismatch occurs.

Boundary: investigate anomaly/method before total rejection. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

79. model failure repeatedly

Model: valid tests repeatedly disagree.

Boundary: revise mechanism, assumptions or range. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

80. two models same prediction

Model: current evidence cannot distinguish them.

Boundary: seek discriminating observation. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

81. two models different predictions

Model: design observation where they diverge.

Boundary: use result to compare models. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

82. simpler model sufficient

Model: extra complexity adds no predictive value.

Boundary: use minimum model needed for task. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

83. complex model needed

Model: simple model misses decisive condition.

Boundary: add only the mechanism required. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

84. model parameter change

Model: same structure with different parameter.

Boundary: separate model form from fitted value. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

85. model range

Model: relation works only between tested bounds.

Boundary: state range in conclusion. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

86. model scale

Model: relation works at one spatial/biological level.

Boundary: do not jump levels without bridge. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

87. model time scale

Model: short-term model differs from long-term.

Boundary: state horizon. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

88. model population

Model: relationship differs across groups.

Boundary: scope by population. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

89. model apparatus

Model: ideal behaviour assumes functioning setup.

Boundary: method failure can mimic model failure. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

90. model measurement

Model: observable is indirect proxy.

Boundary: measurement validity limits conclusion. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

91. model assumption audit

Model: list assumptions required for prediction.

Boundary: unsupported assumption is a repair target. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

92. model validity final rule

Model: state what the model explains, under which conditions, and what observation would make you revise it.

Boundary: a useful model has a domain, not unlimited authority. For practice, write one prediction inside the model’s valid range and one situation where the model would need qualification or replacement.

A model can be useful without being complete

School Science models often isolate one mechanism so learners can reason clearly. Completeness is not the criterion. The criterion is whether the simplification is appropriate for the question and whether its assumptions are respected.

Links

Use the Science Hub, Vol 0071 Competing Mechanisms, Vol 0099 Control-Condition Logic, 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. The official Science assessment objectives include scientific phenomena, facts, laws, definitions, concepts, theories, instruments, quantities and applications; this article’s model-validity framework is an eduKateSengkang reasoning method for organising that work.

Final rule

A useful model has a domain. State what it represents, what it ignores, where it should work and what evidence would make you revise it. That is stronger Science reasoning than treating every diagram or formula as unlimited reality.

Small Group Tutorials

How to Perform in the new G2 SEC Examinations | Learner’s Guide Vol 0103 | Science: Model-Validity Boundaries — Know Where a Useful Model Stops Being Reliable