PSYC 202 1st Edition Lecture 4Measures of Dispersion/Validity - Common descriptive statistics o Number of participants: “N” (N = 171) o Central tendency (center, middle): Mean = average of all scores. Median = mid-point of all ‘ordered’ scores. Mode = most frequently occurring score.o Dispersion (variability, spread): Range = difference b/w maximum and minimum score. Variance = degree to which scores deviate from mean.- How much “spread” is there around the mean? Standard deviation = square root of the variance.- Rangeo Range = difference b/w the minimum and maximum scores.o Minimum score = 8 Maximum score = 20 20 – 8 = 12o Range of scores is from 8 to 20 and can be displayed as: 8, 20 12 (20 – 8)o Range can’t always determine variability though, you need variability to determine variability - Varianceo (sum of each person’s score minus the mean) / the total number of scores minus one- Standard Deviationo The square root of the varianceo Standard deviation just takes things one step farther o It’s the amount of “spread” from the average A low (small) stddev indicates that the data points are very close to the mean  A large stddev means that the data points are far spread out from the mean- Total Varianceo An example of error variance is like trying to measure stress, it’s SO hard to measure stress so we give room for erroro The more error, the harder it is to determine a relationship (like our psych study in our lab with sleep and tech and other factors that can mess with it)- Effect Sizeo Strength of association b/w variable and participants’ responses Effect Size - Strength of association between variable of interest and participants’ responses- It’s going to range from 0.00 – 1.00o The closer to 1 that you get the stronger If you have a .11 variance then you have 89% erroro Cohen’s Criteria 0.01 = small effect size 0.06 = medium effect size 0.15 = large effect sizeo Large effect size means strong association (?)o It may seem bad to have 85% error but that’s still good and still a large effect size Very rarely do you get .15, a large effect size- Meta-analysiso Analyze and integrate results from a large set of