Exercise 2: Why the 9 MPH Gap Shows Reconstruction, Not Poor Speed Estimation — Possible Solution ======================================================================================================= If the study had simply shown that people's speed estimates were generally inaccurate compared to the real, actual speed of the cars in the footage, that alone would only demonstrate poor estimation skill - a real but different, much less interesting finding than what Loftus and Palmer's actual design revealed. The real, crucial detail is that every participant watched the exact SAME footage - the true, actual speed of the cars never changed at all. The only thing that varied between groups was a single word in the QUESTION asked afterward. If the differing estimates were simply about people being bad at judging speed in general, there would be no real reason for the specific word used in a later question to produce a real, systematic, predictable shift in the answers - poor estimation skill would more likely just produce random, inconsistent guesses scattered around the true speed, not a real number that moved in a directly, predictable relationship to how intense the verb sounded. Because the estimates shifted in a real, consistent, directional way that tracked the verb's own intensity ("collided" lowest, "smashed" highest) despite the underlying footage never changing, the only real explanation is that the LATER question itself altered what participants reported having seen - meaning the question is influencing the actual reported memory, not merely revealing a pre-existing inaccuracy in how well people judge speed. ANSWER: The identical footage was shown to every participant, so the true speed never changed - only the verb in the follow-up question did. Since the estimates shifted in a consistent, predictable direction matching the verb's own intensity, this shows the later question itself is reshaping the reported memory, not simply revealing random inaccuracy in people's general ability to judge speed. WHY THIS WORKS AS AN ANSWER ------------------------------ This identifies the real controlled variable (identical footage, only the verb changed) and explains why a directional, predictable shift specifically points to reconstruction rather than generic estimation error.