An Anchoring Experiment: Results — LessWrong