ACT Science: Make Accurate Predictions Using Data Trends

When the ACT asks "If the experiment were repeated at X value, what would the result be?", you're being tested on pattern recognition. The key is this: Only extrapolate in the direction and extent supported by the data shown. If a table shows enzyme activity increasing from 10°C to 50°C, you can predict it will continue increasing at 60°C if there's a clear upward trend. But you cannot predict it will increase indefinitely; biology tells you it'll eventually plateau or decline as the enzyme denatures. The ACT respects this: predictions must be reasonable extensions of the trend, not wild guesses. Identify the type of trend first (linear, exponential, curved, cyclical) and then extend it only one step beyond the data shown.

Example: Data shows 2, 4, 6, 8 values at increasing x. Predict the next: 10 (linear trend continues). Data shows 1, 2, 4, 8 values. Predict the next: 16 (exponential trend continues). Data shows 10, 9, 7, 4 values. Predict the next: 0 or negative (quadratic/decreasing trend continues, likely approaching zero).

Error 1: Extending too far beyond the data. If data spans 20°C to 50°C, predicting at 100°C is too far; the trend may change. Predict at 55°C or 60°C if asked. Error 2: Assuming a trend continues when biological/physical limits suggest it won't. Enzyme activity can't exceed 100%, and at very high temps, all enzymes denature. Error 3: Ignoring the scatter or variability in the data. If values bounce around rather than forming a smooth line, predictions are less reliable. Always note: "Based on the trend shown, the prediction is..." This qualifier acknowledges the limits of your extrapolation.

When you make a prediction, write the trend type: linear, exponential, or non-linear. This discipline forces you to think before answering.

Scenario 1: Plant height at days 0, 10, 20, 30 is 5cm, 8cm, 11cm, 14cm. Predict at day 40. Trend: +3cm every 10 days (linear). Prediction: 14+3=17cm. Scenario 2: Bacterial count at hours 0, 1, 2, 3 is 100, 200, 400, 800. Predict at hour 4. Trend: doubles each hour (exponential). Prediction: 800×2=1600. Scenario 3: Reaction rate at pressures 1, 2, 3, 4 atm is 10, 15, 18, 20 reactions/min. Predict at 5 atm. Trend: increases but at a decreasing rate (curved/logarithmic). Prediction: ~21-22 reactions/min (not 25, which would assume linear continuation). For each, state your reasoning: "The trend is [type], so at the next value, I predict..."

If you miss a prediction, compare your extrapolated point to the actual data and ask whether your trend identification was correct.

Prediction questions are common on ACT Science and test reading comprehension plus basic pattern reasoning. They're not hard, but students often overreach or misidentify trends and lose points. Students who identify trends clearly and extend conservatively score consistently. This single skill can add 2-3 points to your Science section without requiring advanced scientific knowledge.

This week, answer a prediction question on every data-based passage. Check your prediction against the answer key and trace back to see where your trend identification was off if needed.