Scientific Reasoning and Experimental Data Analysis
Scientific reasoning questions test how evidence supports a claim: identifying variables, judging measurement quality, and reading tables and graphs accurately. The exam rewards explicit reasoning about controls, uncertainty and alternative explanations.
Core rules
- Identify the independent variable (changed), dependent variable (measured) and controlled variables (held constant). A control group isolates the effect of the treatment.
- Accuracy is closeness to the true value; precision is consistency among repeats. Random errors scatter around the true value; systematic errors shift all readings in one direction.
- Significant figures: multiplication and division keep the least number of significant figures; addition and subtraction keep the least number of decimal places.
- Graphs: label axes with units, plot the independent variable horizontally, and interpret slope with its unit. Linearise when a relationship is expected to be proportional.
- Correlation does not establish causation; a conclusion must state the evidence and a plausible mechanism.
Worked example
A metal sample has mass 12.45 g and volume 5.0 mL. Density = 12.45/5.0 = 2.49 g/mL, but the volume has two significant figures, so report 2.5 g/mL. If the accepted value is 2.70 g/mL, percent error = |2.49 − 2.70|/2.70 × 100 ≈ 7.8%. Reporting 2.49 g/mL overstates precision; calculating from rounded 2.5 gives 7.4%, so keep unrounded values until the final error calculation.
Traps
Averaging readings without checking for an outlier, treating a systematic offset as random scatter, over-reporting digits, and extrapolating beyond the calibration range are frequent errors. Do not say an experiment 'proves' a hypothesis; state that the data support or fail to support it.