Gamble A: Win $1000 with 0

Multiple Choice

Gamble A: Win $1000 with 0.50 probability or Gamble B: Win $500 for sure.
On the other hand, suppose you have won $2,000 on a game show and are then asked to choose between:
Gamble C: Lose $1,000 with 0.50 probability or Gamble D: Lose $500 for sure.
These two gambles are obviously identical in terms of final wealth states and probabilities. However, subjects are much more likely to choose the risk averse B and the risk seeking C. This suggests that participants:

A) Are making their decisions over changes in wealth and are anchoring their choices on the basis of an initial reference point, rather than the final asset positions and wealth levels.
B) Underweight the 0.5 probability after they win $1000 but overweight that same probability after they win $2000.
C) Behave in accordance with expected utility theory since Gambles A and C yield higher expected value compared to Gambles B and D respectively.
D) Overweight the 0.5 probability after they win $1000 but underweight that same probability after they win $2000.