Science Inquiry Skills: Designing, Measuring and Evaluating Investigations
What you must understand
Science Inquiry Skills is the domain that tests how you do biology rather than what you have memorised. You need to be able to state a testable hypothesis that names the independent and dependent variables, identify controlled variables, choose an appropriate control group, and describe a sampling or measurement protocol that would actually produce reliable data. You should also be able to interpret tables and graphs, describe trends using data, distinguish correlation from causation, and evaluate sources of error, including random error, systematic error and bias. Understanding the difference between accuracy, precision, reliability and validity is essential.
Traps that appear in questions
- Control group versus controlled variables. A control group receives no treatment or a placebo; controlled variables are the factors held constant. Questions often ask for both.
- Repeat versus replicate. Repeating a measurement on the same individual tests precision; replicating across many individuals tests the generality of a conclusion.
- Reliability versus validity. A balance that always reads 2 g too high is precise but not accurate, so the data are reliable but the measurement is invalid.
- Confounding variables. If light and temperature both change along a transect, you cannot attribute a result to either one.
- Describing trends. "The rate increased" is weaker than "the rate increased from 12 to 34 units as temperature rose from 10 to 30 degrees Celsius".
Worked example
A student claims a fertiliser increases plant height. A strong answer states: hypothesis - plants given fertiliser will grow taller than untreated plants; independent variable - fertiliser concentration; dependent variable - final height in millimetres measured with a ruler; controlled variables - species, soil volume, light, water, pot size; control - identical plants given the same volume of water only; replication - at least ten plants per treatment to average out individual variation; evaluation - a ruler read from a fixed angle reduces parallax error, and a random allocation of pots reduces bias.