Author: eduKate Asia
-
PSLE Science Reality Lab Vol No.462 | “Surrogate Recovery = 82%” — Does That Mean 82% of the Target Chemical Was Recovered?
A massive Primary 5/6 PSLE Science Reality Lab guide to interpreting surrogate recovery in laboratory reports: why a surrogate-recovery percentage is a quality-control check on an added compound, not the recovery, removal, concentration or accuracy percentage of the target chemical.
-
PSLE Science Reality Lab Vol No.461 | “Maximum Intensity Projection” — Does One 2D Image Show Everything at the Same Depth?
A massive Primary 5/6 PSLE Science Reality Lab guide to reading maximum intensity projection images: how a 2D projection can combine bright features from different depths, what evidence it preserves, what depth information it loses, and how to avoid mistaking projected overlap for physical contact.
-
PSLE Science Reality Lab Vol No.460 | “Linearity = ±0.2%” — Does That Mean Every Reading Is Accurate to Within 0.2%?
A Primary 5/6 Reality Lab guide to instrument linearity claims: learn why ±0.2% linearity is not automatically ±0.2% total accuracy, and how to check range, offset, calibration and other evidence.
-
PSLE Science Reality Lab Vol No.459 | “Pixel Value = 65535” — Is That the Exact Brightness, or Has the Detector Saturated?
A Primary 5/6 Reality Lab guide to saturated scientific images: learn why a maximum pixel value such as 65535 may be a detector ceiling, not the exact signal, and how to test quantitative claims safely.
-
PSLE Science Reality Lab Vol No.458 | “98% Accurate on the Training Data” — Has the Model Proven It Can Predict New Cases?
A Primary 5/6 Reality Lab guide to model accuracy claims: learn why a high training score does not prove performance on new cases, and how to check independent test evidence, leakage and scope.
-
PSLE Science Reality Lab Vol No.457 | “12-Month Running Mean” — Did the Raw Measurements Really Change That Smoothly?
A Primary 5/6 Reality Lab guide to running means and moving averages: learn why a smooth scientific line is a calculated summary, not the raw measurement path, and how to test claims against the original data.
