x +- 2 a). Therefore, the probability of a reading falling outside confidence limit (95%) is only 5%, i.e. 1 in 20 (P = 0.05). o So, For 1 SD, confidence limit 68% For 2 SD, confidence limit 95% For 3 SD, confidence limit --> 99.7% Note: Generally we consider that for 2SD, confidence limit is 95%. However, actual (exact) values are slightly different-o Actual parameters in a Normal (Gaussian) distribution. o Mean +- 1 SD ( la) limits include 68.27% values o Mean +- 2SD ( 26) limits include 95.45% values o Mean +- 1.96SD ( 1.966) limits include 95% values o Mean +- 3SD ( 3a) limits include 99.73% values o Mean +- 2.58SD ( +- 2.58 a) limits include 99% values Also know: o Values that differ from the mean by more than 2SD are rare, being only 4.55%. o Values higher or lower than the Mean +- 3SD ( 36) are very rare, being only 0.27%. o 6 SD (3 on either side of mean) cover almost the entire range of a variable character (SD divides the range into 6 equal sub-range) o Minimum sample size required for establishment of normal range for any health parameter: 300 healthy subjects. o Confidence limits of normal curve can never be 100%. therefore the limbs of curve never touch base line.">