Controlling variables (age 11+)

Learners aged 11–12 design a full, hypothesis-led investigation (KS3 / NGSS SEP 3): a testable hypothesis with a mechanism ('because…'), named independent/dependent/≥3 control variables, planned repeats and sample size, and a hazard-with-control-measure risk assessment. The core is coordinating this design with the epistemic distinction between validity (controlling variables so the IV's effect is what is measured) and reliability (consistency across repeats). The proposed primary approach is a design-and-run studio in which poor design is intended to produce visibly untrustworthy data; supporting approaches include variable/hypothesis classification tasks and a Socratic supervisor dialogue.

• Writing a hypothesis as a bare prediction with no 'because' / mechanism, or as a non-testable statement of opinion.
• Confusing independent and dependent variables, or mislabelling a control variable as the independent variable.
• Treating 'repeat three times' as a rule without understanding that repeats and sample size exist to improve reliability and expose anomalies.
• Naming only one or two obvious control variables and missing subtler confounds that would invalidate the result.
• Seeing risk assessment as a bureaucratic add-on rather than an integral part of designing a valid, safe investigation.
• Thinking a hypothesis that turns out wrong means the investigation failed — a refuted hypothesis is a valid, informative result.

• The hypothesis frame 'I predict that [IV] will affect [DV] because [reasoning]'
• An IV / DV / control-variables table (with ≥3 named controls, each justified)
• A full planning template: hypothesis → variables → method → repeats & sample size → risk assessment → recording
• A risk-assessment grid (hazard | risk | control measure)
• A predicted results sketch-graph showing the expected IV–DV relationship before data collection

• Teach the hypothesis as prediction-plus-mechanism: the 'because…' clause carrying scientific reasoning is what separates a hypothesis from a guess and is the assessed core — model strong vs weak 'because' clauses.
• Formalise the KS2 change/measure/keep-same language into independent / dependent / control variables now, explicitly bridging from the pupils' prior vocabulary so the terms attach to an already-understood idea.
• Make control-variable identification demanding — require at least three, and justify each — because at KS3 the plausible confounds multiply and 'what else could affect the result?' is the key habit.
• Introduce sample size and repeat readings as reliability tools with a reason (spotting anomalies, reducing the effect of random error, enabling a mean) rather than as a ritual 'do it three times'.
• Embed risk assessment as routine professional practice (hazard → who/what is at risk → control measure), normalising safety planning as part of every KS3 investigation design.