Applied Statistics Test Notes

Validity- does the variable accurately reflect the phenomenon you are trying to measure. Reliability- does the indicator consistently assign the same number. Cross-sectional data- no time/multiple entries. Panel data set- multiple entries/over time. Time series-one entry/time. Population: the total set of items or people that a researcher is interested in studying. Sample: a subset of the population- random or nonrandom. Goal is to be representative of population.Criteria for evaluating causal relationships. Time order: which comes first (independent comes first). Theory: does it make sense. Covartiation: change in one and change in the other, will the two move together. Nonspurioussness: no alternative example. Mean- use when #’s are in interval and no outliers. Median-not sensitive to outliers, useful in ordinal- only captures small amount of information about the sample. Mode- the most frequent # will work with nominal, ordinal or interval. Positively skewed data- is positively skewed because most of the scores tend to cluster toward the lower end of the scale with increasingly fewer scores at the upper end of the scale. Negatively skewed data- is negatively skewed because most of the scores tend to occur toward the upper end of the scale while increasingly fewer scores occur toward the lower end. Standard deviation: number minus the mean, squared, added together and divided by number of values then take the square root. R=1, the numbers fall along a straight line, positive slope. / R=-1 fall along a straight line with a negative slope. \ R=0 no linear relationship between variables. --- Limits on correlation coefficients- can only use if both are interval, and sensitive to outliers, does not capture non-linear relationships, does not show the strength of relationships. Interquartile range- location measure, shows outliers. Regression- how close the value is to the line (r^2) y=a+bx a= intercept. B= slope. Intercept is y=? when x=0. Regression line explains the