Predicting & Concluding
What you must understand
A prediction states what you expect to happen, usually based on a pattern or a scientific idea; a hypothesis is a testable proposed explanation; a conclusion states what the evidence shows and whether it supports the hypothesis. Good conclusions refer back to the aim, use specific data (values, directions, comparisons), acknowledge anomalies and do not overstate. You should be able to continue a trend to predict a new value, judge whether a prediction is safe, and recognise that a hypothesis is supported or not supported rather than proven. Consider sample size, number of trials, consistency between repeats and alternative explanations when judging how much trust to place in a conclusion.
Traps to watch
- Extrapolating far beyond the measured range and presenting the result as certain.
- Using absolute words such as "proves" or "always" when evidence is limited.
- Ignoring an outlier rather than commenting on it, or letting it drive the conclusion.
- Writing a conclusion that simply restates the prediction without data.
- Concluding from one trial or a very small sample.
Worked example
A seedling's height is recorded weekly as 2, 4, 6 and 8 cm. A straight-line continuation predicts 10 cm in week five. That prediction is reasonable only if conditions stay the same; if growth is slowing as a limiting factor such as light or pot size takes effect, a straight line will overestimate. Rule of thumb: interpolate freely, extrapolate cautiously, and say what assumption your prediction depends on.