A political pundit predicts that Candidate A will receive no more than half of the votes in an election against Candidate B. If we conduct a poll of 1000 voters and record the proportion of people who plan on voting for Candidate A, we can test this pundit's prediction. We will reject the pundit's prediction if our poll's sample average p is high enough compared to 1/2. If we assume the true proportion of voters for Candidate A will be p = 0.5, which sample proportion would be so "extreme" in this positive direction that it would be higher than 95% of the p 's we could see from such surveys? Please enter your responses rounded to 3 decimal places. (Hint: Use the normal approximation for p using these parameters.)