Scores on certain test are normally distributed with variance of 70. A researcher wishes to determine the mean score achieved by all adults on the test. Find out the sample size required to assure with 95 percent confidence that the sample mean will not differ from population mean by more than 2 units.