Exam 2 Study Guide PSY 230Chapter 4o Arithmetic Mean (average): Mean=(sum of x)/No Weighted Mean: mean of a group of meanso Median: midpoint of a distribution of scores so that half fall above and half fall below. (50th percentile)o Use with skewed distributionso Good for skewed, truncated, and open ended distributions Truncated: use only part of the distribution Open-ended: top or bottom category has only 1 limito Median for arrays of scores:o If N is odd, put in ascending order, find middle #o If N is even, put in ascending order, take X of two middle #so If N is even but middle 2 #s are the same, use the formula for median.o Mode: most common score (this is a crude measure)o When to use each measurement:o Mode: quick and easy but crude, not unique.o Median: skewed, truncated, open-endedo Mean: most common, normal distributionso Mean, median and skew relationship:o Mean > median = positive skewo Mean < median = negative skewo Mean = median = no skewChapter 5o Dispersion: spread or variability of scores around a central tendency measureo Range: high score – low scoreo Semi-interquartile range (SIR) or Q2: use with median, median plus/minus SIR cuts off middle 50% of scoreso SIR=Q2=(Q3-Q1)/2o Variance or mean square and standard deviation or root mean squareo Use with meano Can use to compare distributionso Quite preciseo Used in statistical tests later ono Large values=high error, low precisiono Small values=low error, high precisiono Mean deviation method: shows how scores vary from the meano Raw score method: easier, less intuitive about meano Homogeneous sample: data values are similaro Heterogeneous sample: data values are dissimilaro Pearson’s coefficient of skew: + or – and how mucho Kurtosis: peaked or flatChapter 6o Z-score: represents distance between score and mean relative to standard deviation (s)o Can use to compare 2 different variables because z scores are abstract numbers without limitso If scores are normally distributed, they can relate directly to PR via the “standard normal distribution” (a theoretically ideal normal distribution)o When you transform data to z-scoreso Mean=0o Sum of squared z-scores=no