Scientific Inquiry, Experimental Design and Data Analysis
Variables, controls, conclusions
Inquiry items hand you an unfamiliar experiment and test whether you can identify variables, judge the controls honestly and draw only the conclusion the data support.
Core rules
- Hypothesis is a testable, falsifiable statement; a prediction states the expected result if it holds.
- Variables: independent (manipulated), dependent (measured), controlled (held constant). The control group receives no treatment or a standard treatment; the experimental group receives the treatment.
- Design: random assignment, adequate sample size, replication, blind or double-blind procedures. Reliability means repeatability; validity means the experiment measures what you claim.
- Data: mean, median, range, standard deviation; overlapping error bars suggest the difference may not be significant; correlation is not causation; look for outliers and measurement error.
- Presentation: labelled axes with units, sensible scale, a table for raw data and a graph for trends.
Worked example
An enzyme is tested at five temperatures with ten tubes per temperature and constant pH and substrate concentration, and activity is recorded as product per minute. Suppose the mean is highest at 40 °C but its error bars overlap those at 30 °C. The correct conclusion is that the difference between 30 °C and 40 °C is not clearly established, not that 40 °C is the optimum. Reasoning: overlapping error bars mean the observed gap may reflect random variation, so more replicates are needed before claiming a peak.
Marks lost
Changing two variables at once; calling the treatment group the control; concluding causation from correlation; reporting numbers without units; ignoring that the stated conclusion must match the measured variable.