Based On Video For Descriptive Stats

Front 1

Cross sectional Data

Back 1

Looking at a large number of subjects over one specific time period

Front 2

Time series data

Back 2

looking at one subject over various times will almost always use a sample since a full population is hard to see

Front 3

4 Ways to measure Data

Back 3

Nominal
Ordinal
Interval
Ratio

Front 4

Nominal Scale

Back 4

taken a qualative value and assigning it a number. The most non-descript way to measure

Front 5

Ordinal Scale

Back 5

Uses numbers to rank things. Tacitly assumes that spaces between numbers are relatively equal but NOT ALWAYS

Front 6

Interval Scale

Back 6

Data Measured using fixed intervals. i.e. a thermometer. DOES NOT HAVE A TRUE ZERO
ex. 0 degrees does not mean a total abscence of heat.

Front 7

Ratio Scale data

Back 7

Interval scale data with a true zero. ex. money, salary, etc.

Front 8

Frequency distribution

Back 8

A table that shows the data you are studying and divides it in to catergories

Front 9

Classes

Back 9

types of fields in a frequency distirubion (columns/categories). Classes should be mutually exclusive

Front 10

Relative Frequency

Back 10

Percentage of total values in a specific freq

Front 11

Cumulative frequency

Back 11

making each interval additive to the following interval

Front 12

Cumulative relative frequency

Back 12

making each interval additive to the following interval and then making that value a percentage

Front 13

Histogram

Back 13

Bar chart that presents the same info in the table

Front 14

Frequency Polygon

Back 14

Line chart that connects dots on an x/y plane. We use the midpoint from each category.

Front 15

Geometric mean

Back 15

nth root of the product of all n returns

Front 16

FInding the Geomtric mean

Back 16

Step 1: add 1 to each percentage value so that we can always have a positive value under the radical sign
Step 2: subtract 1 from final answer

Front 17

Median

Back 17

value that divides sample in half. Half of observations will be above the mediam, half will be below

Front 18

Median for even number of values

Back 18

add the two middle values and divide by 2

Front 19

Mode

Back 19

observation that occurs most frequently

Front 20

Symmetric Skewness

Back 20

Mean=Median=mode

Front 21

Positive Skewness

Back 21

Skewed tot he right. Have outliers that are high in value.
Mean>Median>mode

Front 22

Negative skewness

Back 22

Skewed to the left (long tail to the left) outliers are low in value
Mean<Median<mode

Front 23

Kurtosis

Back 23

Referst to the narowness of the distribution, and the tails of the distribution

Front 24

Normal Kurtosis

Back 24

Has normal looking tails and a fairly rounded point

Front 25

Excess Kurtosis

Back 25

High, thin peak, very fat long tails

Front 26

Range

Back 26

Largest observation - smallest observation

Front 27

Mean Absolute Deviation

Back 27

1/sample (sum of absolute values from it's sample mean
Not used much because of absolute values

Front 28

Population Variance

Back 28

1/n * (sum of all deviations-mean)^2
Not used so often because some units can't be squared so much i.e. squared dollars

Front 29

Population Standard Deviation

Back 29

Square root of population variance.

Front 30

Sample Variance

Back 30

Same as population variance but instead of multiplying by 1/n, we multiply by n-1

Front 31

sample standard deviation

Back 31

square root of sample variance

Front 32

Chebyshev's Theorum

Back 32

minimum proportion within k standard deviations of the mean is 1-(1/k^2)

Front 33

Example of Chebyshev's Theorum for k=2, 3, 4

Back 33

k=2=1-(1/4)=.75 Therefore 75% of all results are within 2 standard deviations.

Front 34

coefficient of variation

Back 34

sample standard deviation / sample standard mean
Allows us to compare

Front 35

Holding Period Return HPR

Back 35

Beginning Price-Ending Price + Dividends / beginning price

Front 36

Sharpe Ratio

Back 36

Sample mean return - risk free rate / Standard Deviation of portfolio

Front 37

What does Sharpe ratio tell us

Back 37

Tells us the expected return per unit of risk

Front 38

Example: Sharpe Ratio 1/3 vs. Sharpe Ratio 1/2. Which one to pick

Back 38

We would want to choose asset with 1/2 sharpe ratio as we get more bang for the buck

Front 39

Geometric Mean use

Back 39

used to find past compound rates of return

Front 40

Weighted or arithmetic mean

Back 40

Used to find expected future return

Front 41

Ex. Geometric Mean: stock
Period 1: 100
Period 2: 200
Period 3: 100

Back 41

Geometric Mean = Square root of (1+1) * (1-.5) = 1. Now need to take away 1 and you have a 0% return

Front 42

Ex. Arithmetic Mean: stock
Period 1: 100
Period 2: 200
Period 3: 100

Back 42

1+-.5 / 2 = .25. This is incorrect because it seems to imply that we made money when we didn't.