The only thing you missed on the test was on the second problem, where it specified that the oscillator was at equilibrium and moving downward at t = 0. This implies that it was released prior to t = 0, and that the model would be y = A sin(omega * t + pi) or alternatively y = - A sin (omega * t). There would have been no phase shift, and the function would not be the cosine, since the cosine would not place the oscillator at equilibrium when t = 0. You used the cosine function on the assumption that t = 0 occurred at the instant of release. In practical terms there is not necessarily a reason to trigger the clock at the instant of release, and there might well be a reason to trigger the clock at the instant the object passes downward through the equilibrium position. Your solution based on the cosine model was correct, except for this discrepancy.