Were your errors equally distributed between positive and negative errors? Was your PSE more accurate for the second setof 20 trials? What might account for this?
Why do you think people make errors judging linelength? After all, the two lines are right there in front ofyou, in plain view…….