In theory, in 19 cases out of 20, the results based on such samples will differ by no more than two or three percentage points in either direction from what would have been obtained by seeking out all American adults, and by no more than four percentage points from what would have been obtained by seeking out all adult New Yorkers.
In theory, in 19 cases out of 20, the results based on such samples will differ by no more than 3 percentage points from what would have been obtained by seeking out all adult New Yorkers.
In theory, in 19 cases out of 20 the results based on such samples will differ by no more than 3 percentage points in either direction from what would have been obtained by seeking out all adult New York City residents.
In theory, in 19 cases out of 20 the results based on such samples will differ by no more than 3 percentage points in either direction from what would have been obtained by seeking out all adult New Yorkers.
In theory, in 19 cases out of 20, overall results based on such samples will differ by no more than three percentage points in either direction from what would have been obtained by seeking out all American adults.
In theory, in 19 cases out of 20, overall results based on such samples will differ by no more than 3 percentage points in either direction from what would have been obtained by seeking out all adult residents of New York City.
In theory, in 19 cases out of 20 the results based on such samples will differ by no more than four percentage points in either direction from what would have been obtained by seeking out all registered voters in Florida.
In theory, in 19 cases out of 20, overall results based on such samples will differ by no more than 4 percentage points in either direction from what would have been obtained by seeking out all American adults.
In theory, in 19 of 20 cases, the overall results based on such samples will differ by no more than three percentage points from what would have been obtained by seeking to interview all American adults.
