PSY 321 Lecture Notes - Lecture 13: Categorical Variable, Descriptive Statistics, Statistical Hypothesis Testing
Document Summary
Scores need to be placed in a data matrix. So we know how to interpret the numbers. Categorical (discrete) - scores fall into a finite set of groups when scoring the data. Continuous - has a larger range of possible scores on the measure. Interval - implies order, distance meaningful, but arbitrary zero. Ratio - implies order, distance meaningful, and true zero. Use univariate statistics (descriptive statistics, one variable statistics) Measures of central tendency - what typical score value is. Mean - best for normal distribution of continuous data. Median - best for skewed distribution of continuous data. Mode - most frequent score usually best for categorical data. Standard deviation and variance - average distance of each scores from the mean. Assessed using cronbach"s coefficient alpha (a) a > 0. 9 = excellent a < 0. 5 = unacceptable measures. Items (internal consistency) - consistently measuring some sort of concept with those items. Raters (inter-rater agreement) - different ratings sources are reliable.