Scores upon certain test are normally distributed with variance of 44. A researcher wishes to determine the mean score achieved by all adults upon the test. Determine the sample size required to assure with 95 percent confidence that sample mean will not differ from the population mean by more than 3 units.