Predicting the Effect of Modifying the Design (SIN 702)

An experiment's tables and graphs already show how the measured variable responds to the thing being changed. If the data show a steady pattern, like every doubling of enzyme cutting reaction time in half, that pattern is a rule you can apply to any trial, not just the ones actually run. So a question that changes one setting and holds everything else fixed is asking you to read the existing trend and slide along it to the new value. No new data, no new logic, just extending what the graph or table already tells you.

The trace works in one direction: find where the modified trial would sit in the original data, then read off what the pattern says happens there. If concentration climbs steeply from 10 to 30 micromol/L then flattens from 30 to 50, a trial sitting at 45 and then doubled is already past the climbing region, so the flat line predicts almost no change in rate. The shape of the curve, not the size of the change you make, decides the outcome. That is also why doubling an input doubles an output only in the region where the relationship is still linear or proportional; past a saturation point, enzymes run out of substrate to grab, receptors fill up, or some other limit takes over, and the curve goes flat or bends.

This connects directly to control conditions covered nearby. A design change can also strip out the untreated or solvent-only flask, and when that happens the experiment loses its baseline. Without it, treatments can only be ranked against each other, never shown to work at all compared to doing nothing.