AP Stats unit 1 test review

standard deviation(o)

measures the typical distance from the mean

interquartile range(IQR)

mini medians: median of 1st half data and secede half data: median of median

Q3-Q1=IQR

3 medians in total

percentiles

the percent of data less than or equal to a certain value

sample/total

for example, data value 9 out of 10 values would be 9/10 so the 90th percentile

5 number summary

median, minimum, maximum, Q1, Q3

used to make a boxplot

determining outliers with IQR

lower bound: Q1 - 1.5(IQR). way to small < Q1-1.5(IQR)

upper bound: Q3 +1.5(IQR) way to big > Q3+1.5(IQR)

deterring outliers with standard deviation

lower bound: mean - 2(SD)

upper bound: mean -2(SD)

box plots

• each section represents 25% of plot = actual box represents 50%

z-scores

measures how many standard deviations a data value is above/below the mean

z = (x-u)/o

or z-score = (sample-mean)/SD

roughly/lightly symmetric

use because nothing is exactly symmetric

mode

most frequent number in a data set

SOCS

in resposes, explain every part of this:

Shape
Outliers
Center
Spread(min-max =range) but use term variability

median vs mean

when roughly symmetric use mean

when skewer use median

more resistant measures

medien, mean and SD change a lot when values are multiplied/divided/etc

how to interpret SD

rainfall for a city TYPICALLY VARIES BY 15.52 INCHES FROM THE MEAN OF 34.94 INCHES

perfect graph percents

0.15,

2.35,

13.5,

34,

34,

13.5,

2.35,

0.15

extra credit answers

• 60 girls

• x = 3, y=2

• a = 3, b = 8, c = 9

• 6 ,7,8,9

• 27 × 37 = 999(s=9)

standardizing(changing) values and how it affects measures of data

| shape |. center. | variability(spread |

---

(+) constant a. |. same. | + a |. same |

———————————————————————————————————————————————————————————————————

(-) constant a |. same. | - a |. same |

---

(x) constant b. |. same. | x b |. x b |

---

(div) constant b. |. same. | divided by b | divided by b. |

prospective study

following participants over time

retrospective study

using past records for data(past)

variables of interest

groups that are being compared in a study

conclusion of a study

answers the investigative question

population

all of everything in a place or concerned people

sample

subject of population that is smaller and represents population. goes wrong when people use biased samples

relative frequency

takes data and makes percentage of wholes(proportions)

graphs

need to start at zero and If necessary need to become a mosaic graph

marginal distribution

add up all columns vertically and horizontally

association

knowing the value of one variable helps us predict the other variable

classes/bins

intervals of equal width that cover the spread of distributive data of a quantitative

histogram

quantitatative data on x an y axis and no spaces between bins

bar graph

qualitative(categorial) on x axis and spaced between bins

categorical(qualitative) variable

takes on values that are category names or group labels. ex. eye color, zip code, fav food.

Quantitive variable

takes on numerical values(different numbers or measurements)

describing a distributions

CUSSC: center(center values of a data set),

unusual features(gap, outlier),

shape(uniform, unimodal(with peak), bimodal(2 peaks))

spread(vary between _ intervals)

context(use in all descriptions of graph, what type of sample with what groups of people)