7. Drawing Conclusions: Evidence, Inference & Error

The final reasoning skill is judging which conclusions the evidence actually supports — and which go too far.

• An observation is what the data directly shows ('the treated plants averaged 5 cm taller').
• An inference is a reasonable interpretation that goes a step beyond ('the fertilizer probably helped them grow').
• A valid conclusion stays within what the data supports; an overreach claims more than the evidence allows ('this fertilizer makes any plant grow taller anywhere').

Strong science also values replication — a result you can repeat is far more trustworthy than a one-time finding — and a large sample size, because a few data points can mislead while many reveal the real pattern.

On the test, the best answer is usually the most cautious one that the data still backs up. Watch for trap choices that use absolute words like always, never, proves, or all — real data rarely supports them.

Example: a study finds students who ate breakfast scored higher on one test at one school. A valid conclusion: 'breakfast was associated with higher scores in this group.' An overreach: 'breakfast guarantees better grades for everyone.'

⚠️ Common misconception: a single experiment rarely 'proves' anything for all time. Science builds confidence through repeated, consistent results.

💡 Tip: match the conclusion to the data's scope. If the study tested tomatoes, the conclusion can only be about tomatoes — not 'all plants'.