4.2 Drawing Valid Conclusions

4.2 Drawing Valid Conclusions

Once the measurements are in, the skill flips from looking forward to looking back: what do the results actually allow you to say? A conclusion is a statement about what the results mean. On ICAS, concluding questions usually hand you a table or graph plus four possible conclusions, and your job is to pick the one the evidence genuinely supports — which is harder than it sounds, because the wrong options are written to sound scientific.

Results Are Not Conclusions

Keep these two ideas separate:

• Results are what you actually measured or observed — the raw numbers and facts. "The plants with fertiliser averaged 24 cm tall; the plants without averaged 19 cm."
• A conclusion is the meaning you draw from those results. "In this test, the fertiliser helped the plants grow taller."

Results say what happened; conclusions say what it means. If a so-called conclusion just repeats a number from the table, it is really a result.

The Two Tests of a Valid Conclusion

A conclusion is valid only when it passes both of these checks:

1. It is supported by the actual results. Every claim in the conclusion must match something in the data. If the table shows plant heights, a conclusion about leaf colour has no support — leaf colour was never measured.
2. It is scoped to what was actually tested. The conclusion may only cover the conditions, materials and range you investigated. Tested one brand of paper towel? Your conclusion is about that brand, not every brand ever made.

Worked example. Students grew bean plants with different amounts of daily water:

A valid conclusion: "Up to 40 mL per day, more water led to taller plants; above that, extra water made little difference." It matches every number and stays inside what was tested. An invalid one: "Water is the only thing plants need to grow" — light, soil and air were never tested, so the data says nothing about them. Another invalid one: "All plants need exactly 40 mL of water a day" — only bean plants were tested, and only for three weeks.

Overgeneralising from One Trial or One Sample

The most common concluding error is overgeneralising — stretching a small result into a giant claim. One trial can contain flukes; one sample of five snails cannot speak for every snail in Australia. Watch for words like all, always and every: they are usually a sign the conclusion has escaped its evidence. Scientists repeat trials and use larger samples precisely so their conclusions can safely be broader.

Correlation Is Not Causation

Two things can rise and fall together without one causing the other. Correlation means two variables change together; causation means one actually makes the other happen. The classic example: ice-cream sales and the number of people swimming at the beach both rise in January. Eating ice cream does not make people swim — a third factor, hot summer weather, pushes both. The age-appropriate rule for ICAS: if a conclusion says X causes Y, check whether the experiment actually changed X and measured Y while keeping everything else the same. If the two things were only observed side by side, the safest conclusion is "X and Y are linked," not "X causes Y."

A Conclusion Must Survive Every Data Point

One awkward result can sink an otherwise tidy conclusion. Suppose four of five trials show a rubber ball bouncing higher than a foam ball, but one trial shows the opposite. A conclusion like "the rubber ball bounced higher in every trial" is now false, even though it is nearly true. The honest fix is to soften the claim — "in four of our five trials, the rubber ball bounced higher" — or to investigate the odd trial before concluding anything. When you check a conclusion option on ICAS, hold it up against each row of the table or each point on the graph, not just the overall impression. A conclusion that contradicts a single data point is out, no matter how scientific its wording sounds.

How ICAS Words Conclusions

Exam writers have favourite phrasings, and they are clues:

• Careful, valid wording: "The data shows...", "In this experiment...", "These results suggest...", "Under the conditions tested..."
• Overclaiming, usually wrong wording: "This proves...", "All...", "Always...", "This will happen every time..."

Science rarely claims to prove something from a single school experiment — results support or suggest a conclusion. When two options both look plausible, choose the one with the humbler, better-scoped wording.

Choosing the Best-Supported Conclusion

When all four options sound plausible, use this routine:

1. Cross out any option that mentions something not measured in the data.
2. Cross out any option that overgeneralises (all, always, proves) beyond the test.
3. Cross out any option that contradicts even one data point.
4. From what remains, pick the option closest to the results, with the narrowest claim that still answers the question.

The best conclusion is rarely the most exciting one — it is the one that says exactly what the data says, and no more.