SOC SCI 89 Lecture Notes - Lecture 5: Abraham De Moivre, Central Limit Theorem, Statistical Inference
Document Summary
If we are using sample data to make statements about a larger population (i. e. inferential statistics), we must account for the fact that sample data tends to underestimate population parameters . To correct for this we divide variance and sd by n-1. The variance, however, doesn"t tell us much about the average dispersion or spread of the scores (the variance is useful for other things that we will cover later) For the average spread of the scores, we must calculate the standard deviation, which is simply the square root of the variance of s^2. In summary, the sd helps us gauge how dispersed the values in our sample set are around the mean. If all of the values bunch close together the standard deviation is going to be closer to zero. If the values are spread out the sd is going to be much larger.