Describing your control of variables and your accuracy techniques

Two Merit criteria live in the method

Two of the Merit criteria are earned in the method section, not the analysis: describing the control of other variables that could significantly affect the results, and using techniques to improve the accuracy of measurements. Both must be written down — doing them without saying so earns nothing.

Controlling variables: describe, do not list

The criterion says describe the control, which means three things per variable.

1. Which variable, and why it would significantly affect the result.
2. How you kept it constant, specifically.
3. How you checked that it stayed constant.

Weak	Meets the criterion
"I kept the temperature the same."	"The wire's resistivity rises with temperature, so heating would raise the measured resistance. I switched the current on only while taking each reading, kept it below $0.3$ A, and checked with a thermometer beside the wire that it stayed within $2$ °C of room temperature."
"I used the same bob."	"A heavier bob is less affected by air resistance, so changing it mid-experiment would change the damping. I used the same $200$ g brass bob throughout and left it attached to the string between runs."

• Only include variables that would have a significant effect. Listing irrelevant controlled variables does not earn the criterion, and the clarification for the standard says so directly.

Choosing which variables matter

Ask, for each candidate: if this changed by a realistic amount, would the dependent variable change by more than its uncertainty? If not, it is not significant and does not belong in the list.

Investigation	Significant	Not significant
Pendulum period vs length	release angle, string stretch	colour of the string, room lighting
Resistance vs wire length	wire temperature, wire diameter, contact points	ambient humidity
Illuminance vs distance	background light, lamp voltage, sensor alignment	time of day (if in a dark room)
Spring period vs mass	spring identity, amplitude, whether the mass swings sideways	bench height

Techniques that improve accuracy at Level 3

Level 2's techniques all still apply — timing multiple oscillations, repeats and means, fiducial markers, avoiding parallax, zeroing instruments. Level 3 adds a few worth knowing.

• Measure over the largest interval available. The instrument's absolute uncertainty is fixed, so a larger measured value has a smaller percentage uncertainty.
• Use an electronic timer — light gates, a data logger, or video at a known frame rate — to remove reaction time entirely rather than merely reducing it.
• Reverse the measurement and average. Taking a reading with the independent variable increasing, then again with it decreasing, and averaging, cancels any hysteresis or drift that builds up during the run.
• Take a background or zero reading and subtract it. This removes a systematic offset outright, which repeats and averaging cannot do.
• Use a four-terminal or differential measurement where lead resistance or a common offset would otherwise be included in every reading.

Writing it up so it counts

Each technique needs the same three parts as at Level 2, plus a number where possible:

"I timed 20 oscillations rather than one, because my reaction-time uncertainty of about $0.2$ s is the same either way; dividing by 20 reduces its contribution to the period to $0.01$ s, which is under $1\%$ of the shortest period I measured."