Statistics Concepts

Question 1

level of measurement***

Answer 1

refers to way numbers assigned to categories of variable (nominal, ordinal, ratio/interval)

Question 2

discrete variable***

Answer 2

assumes distinct values, can be finite or infinite, always some minimum unit beyond which categories cannot be divided

Question 3

continuous variable***

Answer 3

can assume any value on continuum

Question 4

unit of analysis

Answer 4

Refers to specific group or entity on which statistical

analysis is performed

Question 5

data structure

Answer 5

types: cross-sectional, time series, panel

Question 6

array

Answer 6

univariate visualization involving list of observations in increasing or decreasing order

Question 7

stem-and-leaf plot

Answer 7

univariate visualization; Stem represents leading digit of number, Leaf represents trailing digits

Question 8

frequency distribution***

Answer 8

univariate visualization which counts number of times category repeated in data set

Question 9

percent distribution***

Answer 9

univariate visualization which shows what percent of cases fall into each category

Question 10

cumulative distribution***

Answer 10

univariate visualization which shows relative position of category in distribution (e.g. percentiles of GRE test takers)

Question 11

histogram***

Answer 11

univariate visualization which uses vertical bars to designate frequency/percent of category, no spaces between bars to convey that variable measured on continuum

Question 12

pie chart/bar graph

Answer 12

Pie charts/bar graphs are univariate visualizations used to graph nominal or ordinal variables, most useful when variable being graphed has few categories

Question 13

mean***

Answer 13

Most commonly used measure of central tendency; mean restricted to interval/ratio variable because
involves arithmetic, also used with dichotomous
variables

Question 14

trimmed mean

Answer 14

reduces influence of extreme values on calculation of mean by excluding portion of data in tails of distribution (e.g., 10 percent trimmed mean discards highest/lowest 10 percent of data)

Question 15

median***

Answer 15

Measure of central tendency used with ordinal or higher

Question 16

mode***

Answer 16

measure of central tendency used for all levels of measurement

Question 17

variance***

Answer 17

denoted s^2, most commonly used measure of variability; takes into account all observations of variable

Question 18

standard deviation***

Answer 18

measure of variability, denoted S. Disadvantage of variance uses square of deviation from mean measure not expressed in units of original variable, can return to original metric by taking square root of variance

Question 19

range

Answer 19

Measure of variability measuring span of data, or

maximum possible difference in categories

Question 20

interquartile range

Answer 20

Modified version of range, not as susceptible to outliers; measured as range of middle 50 percent of
observations and equal to difference b/w third quartile (75th
percentile) and first quartile (25th percentile)

Question 21

box plot

Answer 21

Graphical device used to display univariate summary measures, displays 5 key measures on axis: min, Q1, median, Q3,
max

Question 22

skewness

Answer 22

skewness measures describe whether distribution

symmetrical or skewed. Rightward skew = positive, leftward skew = negative

Question 23

kurtosis

Answer 23

Measure of skewness reflecting peakedness or flatness of distribution; Leptokurtic = tall peak, kurtosis value > 3, Platykurtic = flat peak, kurtosis value < 3, Mesokurtic = symmetric peak, kurtosis = 3