Midterm

Question 1

Nominal

Answer 1

any order, many categories (gender, location, etc)

Question 2

Ordinal

Answer 2

ranked or in a specific order (5 star rating) but not same distance between ranks

Question 3

Interval

Answer 3

can be ordered and have no true zero (0 means attribute did not exist) and cannot be compared directly (10 degrees is not twice as warm as 5)

Question 4

Ratio

Answer 4

has an order with natural zero so can be compared (time, 10 min is twice as long as 5)

Question 5

Skewed

Answer 5

heavily weighted to one side
Negative skewness: weighted (taller) to the right
Positive skewness: weighted to the left

Question 6

Variance

Answer 6

the square of the deviation
High = data points are very spread out from each other and mean
Low = data points are very close to mean and each other

Question 7

Non-Probability/Non-Random Sampling

Answer 7

not selected at random therefore some units have no chance of selection.

Question 8

Judgement or purposive

Answer 8

chose based on porpose with limited people with expertise in the area being researched

Question 9

Quota

Answer 9

select until you reach predetermined # within a category (age, gender, etc.)

Question 10

Snowball

Answer 10

one person refers another

Question 11

Self-selection

Answer 11

based on if people agree to participate or not in the sample

Question 12

Convenience, haphazard, accidental

Answer 12

chosen based on convenience (research pool)

Question 13

Random sampling

Answer 13

each item has equal probability of being selected. Eliminates bias.

Question 14

Simple random

Answer 14

all equal chance of selection by chance. Table or generator.

Question 15

Stratified

Answer 15

divide N into subgroups (age, gender,…) and take sample from each

Question 16

Cluster

Answer 16

divide N into clusters with some from each subgroup and use entire cluster

Question 17

Systematic

Answer 17

units are selected from random starting point or fixed interval

Question 18

Relative frequency

Answer 18

Number of occurrences/number of experiments - %

Frequency = how many times an event occurs

Question 19

P(A)

Answer 19

probability of A occuring of all possible outcomes

Question 20

P(A’)

Answer 20

probability of everything but A occuring

Question 21

(A U B)

Answer 21

probability that either A or B or both occur P(A) + P(B) - P(A∩B) - probability A + probability B - intersection of A and B

Question 22

P (A∩B)

Answer 22

probability that both A and B occur at the same time

P(A l B) x P(B)

Question 23

P (A l B)

Answer 23

probability that A will occur given B has happened

Question 24

Empirical Rule

Answer 24

68, 95, 99.7

68% lie with +/- 1 standard deviation σ
95% within +/- 2 standard deviation σ
99.7 within +/- 3 standard deviation σ

Question 25

Kurtosis

Answer 25

peakness of distribution

Question 26

Lepto

Answer 26

higher peak (positive kurtosis)

Question 27

Platy

Answer 27

Lower peak (negative kurtosis)

Question 28

Chebyshev

Answer 28

At least 1 - 1/K^2 lie within k st. deviation | K = NUMBER of standard deviations

Question 29

Discrete distributions

Answer 29

outcomes are whole numbers (binomial and poisson)

Question 30

Binomial

Answer 30

each trial has either success of failure Binom.dist (x, n, p, 1/0) 1 = cumulative - prob you get x or less 0 = exact - prob you get exactly x

Question 31

Poisson

Answer 31

counting #’s over a specific period/interval Poisson.dist (x, mean, 0/1) Adjust your mean so that the intervals match Sq.rt (lambda) = standard deviation

Question 32

Poisson Approximation to the Binomial

Answer 32

Use if n > 20, and mean np <= 7 Large # of trials, small chance of success Poisson.dist (x, mean NP, 1/0)

Question 33

Continuous Probability Distribution

Answer 33

you can not count the outcome (uniform, normal) | Probability of getting EXACT number is ZERO - ALWAYS CUMULATIVE

Question 34

Uniform

Answer 34

constant probability between a max and minimum - Height x length - 1 / (b - a) * (d - c) - There is 0 probability outside of the max/min bounds

Question 35

Normal Distrubition

Answer 35

= norm.dist (x, μ, st. dev, 1) To find x value = norm.inv (p, μ, st. dev) ALWAYS 1 You can put exact values in and not use number line

Question 36

normal approximation

Answer 36

used to approximate binomial when Mean: np > 5 AND n(1-p) > 5 Must find mean and standard deviation Use CORRECTION FACTOR

Question 37

Standard Normal Distribution

Answer 37

curve with mean of 0, st. deviation of 1 Translate normal into standard normal using Z score Z score = point - mean / st. dev Norm.s.dist (z value, 1) To find z when we know probability = Norm.s.inv (probability)

Question 38

Z score

Answer 38

point - mean / st. dev

Question 39

Sampling Distributions

Answer 39

now model the average / random amount of x bar rather than single number

Question 40

Central limit theorem

Answer 40

if sample size is > 30 the sample mean will be normally dist.

Question 41

Sampling Distributions Boat 1

Answer 41

Know population st. deviation and no population size / n/N < 5% Find sample mean and sample standard deviation Sample average standard deviation = st. dev/√n

Question 42

Sampling Distributions Boat 2

Answer 42

Know population st. D and n/N > 5%, use finite correction factor Multiply sample average st. deviation by finite correction factor

Question 43

Sampling Distributions Boat 3

Answer 43

do NOT know know population standard deviation Only S (sample st. dev) → T-distribution T.dist (t value, df, 1) T.inv (p, df) x̄ is the number you are modeling

Question 44

Confidence interval

Answer 44

proportion of intervals that would contain the true mean x bar is sample mean

Question 45

Alpha

Answer 45

space outside of bounds of confidnce interval

Question 46

Elementary

Answer 46

simple/basic) event: cannot be broken down into smaller events.

Question 47

Compound event

Answer 47

composed of two or more elementary events

Question 48

Mutually exclusive events

Answer 48

events that cannot occur simultaneously

Question 49

Percentiles

Answer 49

percentile.inc (array,k) 1% = 1% of data is below, 99% is above Array = all data K = the percentile you are looking for (1% = 0.01)