Assessment
Assessment tests your ability to design, analyse and interpret assessment in physics. Items combine definitions of assessment types with practical work on item statistics, cognitive demand levels and marking.
Core rules
- Formative assessment happens during learning and feeds teaching forward; summative assessment happens at the end and reports achievement; diagnostic assessment locates specific difficulties before or during teaching. Assessment as learning involves learners judging their own work through self- and peer-assessment.
- Validity asks whether the instrument measures what it claims; reliability asks whether it gives consistent results on repetition; fairness asks whether it disadvantages any group. A test can be reliable but invalid, for example a physics test that is really a reading-comprehension test.
- Bloom's taxonomy orders cognitive demand: remember, understand, apply, analyse, evaluate, create. Command words must match the intended level - 'state', 'explain', 'calculate', 'suggest and justify'.
- A table of specifications (test blueprint) cross-tabulates topics against cognitive levels and marks, so that coverage and demand are deliberate.
- Difficulty index p = mean mark obtained ÷ maximum mark. Values around 0.3 to 0.7 discriminate best; p = 1.0 means everyone scored full marks.
- Discrimination index D = (correct in upper group − correct in lower group) / number in one group. D above 0.3 is good; near zero means the item does not separate able from weak learners.
- Marking schemes must award marks for method as well as answer, and rubrics need explicit, observable criteria.
Worked example
A class of 40 takes a 1-mark item. Taking the top 20 and bottom 20 scorers, 14 of the upper group and 6 of the lower group answered correctly. D = (14 − 6)/20 = 0.40, which is a well-discriminating item and can be retained. If all 40 had answered correctly, p = 1.0 and D = 0: the item is too easy and teaches nothing about differences in attainment, so it should be revised or replaced.
Traps that cost marks
- Using validity and reliability interchangeably.
- Writing items that test recall when the objective demands application.
- Setting p values at the extremes and concluding the item is fine because it is 'clear'.
- Giving full marks for a correct answer with no working, when the scheme requires method marks.
- Providing vague rubric criteria such as 'good understanding'.
Misconceptions and Classroom Practice in Physics
This domain tests whether you recognise the standard physics misconceptions learners bring into the classroom and can plan teaching that displaces them rather than merely stating the correct idea over the top of them.
Core rules
- Common misconceptions with their correct physics: a force is needed to keep a body moving (Newton's first law, friction); heavier bodies fall faster (in free fall all accelerate at g); current is used up around a series circuit (current is conserved); voltage flows (potential difference is across, current is through); mass and weight are the same; heat and temperature are the same; sound travels through a vacuum; cold flows into a room; seasons arise from the Earth's varying distance from the Sun; light always bends towards the normal on refraction; astronauts are weightless because there is no gravity; a moving charge always feels a magnetic force.
- Misconceptions are stable and resistant to being told. Elicit–confront–resolve is the standard route: surface the learner's idea, present a discrepant event, then help the learner rebuild the explanation.
- Predict–Observe–Explain (POE) is a practical classroom protocol: learners commit to a written prediction, watch the demonstration, then reconcile the difference.
- Analogies are useful but must be mapped explicitly, including where they break down - the water-flow model of current, if pushed, produces the 'used up' error.
- Diagnostic questioning and concept cartoons reveal ideas cheaply before teaching; the same questions after teaching show whether change occurred.
Worked example
A POE on inertia. A card carrying a coin sits on a glass tumbler. Learners predict what happens to the coin when the card is flicked horizontally. Most predict the coin moves with the card. The demonstration shows the coin dropping into the glass. The explanation: the flick acts on the card over a very short time, the friction between card and coin is small and brief, so the coin's inertia keeps it almost stationary horizontally while gravity brings it down. Recording the prediction first is what makes the learner's prior idea visible and therefore changeable.
Traps that cost marks
- Simply telling the correct answer and assuming the misconception has gone.
- Ignoring a learner's wrong answer instead of probing why they think it.
- Stretching an analogy past its limits, for example treating current as a fluid that is consumed.
- Writing a teaching sequence that never elicits prior ideas.
- Dismissing everyday language such as 'sucking' or 'cold entering' without contrast.
Mathematical and Graphical Reasoning in Physics
This domain tests the mathematical fluency a physics teacher must model: formula rearrangement, proportionality, vectors, and above all the interpretation of graphs - gradient, intercept, area and linearisation.
Core rules
- Every physical quantity carries units; a correct dimensional result rules out many wrong rearrangements before any numbers are substituted.
- Rearrange algebraically first, then substitute, and carry units through the calculation.
- On a straight-line graph y = mx + c, m is the gradient and c the intercept. Gradient of a displacement–time graph is velocity; of a velocity–time graph, acceleration; the area under a velocity–time graph is displacement; area under a force–extension graph is energy stored; area under a current–time graph is charge.
- Linearise any relationship that is not already linear before plotting. For T = 2π√(l/g), plot T² against l; the gradient is 4π²/g, so g = 4π²/gradient. For v = u + at, plot v against t: gradient a, intercept u. For pV = nRT, plot p against 1/V. For Hooke's law, plot F against x: gradient k.
- Vectors: resolve into perpendicular components, add components, then recombine with Pythagoras and tanθ = opposite/adjacent.
- Proportionality means a straight line through the origin; linearity alone does not.
- Read graph scales carefully, use the largest triangle available for the gradient, and keep the calculator in degree mode for trigonometric work.
Worked example
A velocity–time graph shows speed rising uniformly from 0 to 20 m/s in 5 s, holding at 20 m/s for 10 s, then falling uniformly to 0 in 5 s. Area under the graph gives displacement: ½ × 5 × 20 = 50 m, plus 10 × 20 = 200 m, plus ½ × 5 × 20 = 50 m, giving 300 m. Gradient during the first phase is 20/5 = 4 m/s² acceleration; during the final phase, −20/5 = −4 m/s² deceleration. The same graph answers both questions, provided you know which feature is which.
Traps that cost marks
- Confusing the gradient of a graph with the area beneath it.
- Connecting points dot to dot rather than drawing a line of best fit.
- Ignoring a false origin and reading the intercept as the true value.
- Quoting a gradient with no units, or with units derived from the wrong axes.
- Leaving the calculator in radian mode for sin, cos and tan problems.
- Substituting numbers before rearranging, and losing the required variable.