BUS 230 Quiz 1

Numerical variable

Quantitative, represents meaningful numbers

Categorical variables

Qualitative, represents categories

Discrete variable

Assumes a countable number of values

Continuous variable

Uncountable (Unlimited) values within an interval

Nominal scale

Uses names, labels or categories to sort items into distinct groups. Least sophisticated level of measurement

Ordinal scale

Uses numbers to categorize and rank items with respect to some characteristic or trait (Ex: Customer service ratings)

Interval scale

Uses numbers to categorize and rank items while finding meaningful differences between them (Ex: Temperature)

Ratio scale

Same as interval scales, but also has a true zero point (Ex: Sales, profit, inventory levels, …)

Omission strategy

Also called complete-case analysis, excludes observations with missing values from analysis

When to use omission strategy

When the amount of missing values is small and are expected to be distributed randomly across observations

Mean imputation strategy

Replacing missing values with the mean values across relevant observations

When not to use mean imputation strategy

When there is a large number of missing values, or if the missing data is not random (Ex: Sensitive data)

How to address missing values for a categorical variable

1. Use to most frequent category

2. Create an “Unknown” category

Subsetting

Extracting portions of a data set (variables) that are relevant to an analysis

Benefits of subsetting

1. Eliminate unwanted data

2. Can reveal insights in data (descriptive analytics)

Binning

Converts numerical variables into categorical variables by grouping numerical values into a small number of bins (groups)

Benefits of binning

1. Reduces noise in data caused by minor observation errors

2. Create equal intervals

3. Create bins of equal counts

Rescaling

Rescale numerical data so that variables in a data set are measured in a same scale

Category reduction

Collapse some categories to create fewer nonoverlapping categories

When do we use category reduction

1. When variables have too many categories

2. Variables have some categories that rarely occur

3. When a small sample doesn’t have any observations in certain categories

How to do category reduction

1. Create a “Other” category for categories with very few observations

2. Combine categories with similar impacts

Dummy variables

An indicator/binary variable, takes on values of 1 or 0 to describe 2 categories of a categorical variable

Amount of dummy variables to create

One less than the number of categories of the variable

Assigning bin number 1 to categories

The category you are testing or measuring

Assigning bin number 0 to categories

When the row does not match that category

Category scores

Use when the data is ordinal or have natural ordered categories

Appending data

Consolidating multiple data sources that share the same format and variables

Merging data

Using a common variable between 2 data sets to merge them together

=AVERAGE

Calculates the mean or average between a number of observations

=MEDIAN

Calculates the median as a measure of central location. Is the middle value of the ordered observations of a variable

=MODE

Calculates the value that occurs most frequently in a variable

Unimodal

If a variable has 1 mode

Bimodal

If a variable has 2 modes

Multimodal

If a variable has more than 2 modes

Range

Simplest measure of dispersion, is the difference between the maximum and minimum observations of a variable

Interquartile Range (IQR)

The difference between the third quartile and the first quartile. The range of the middle 50% of observations of the variable

Mean absolute deviation

An average of the absolute differences between the observations and the mean

2 most widely used measures of dispersion

Variance and standard deviation

Variance

An average of the squared differences between the observations and the mean

Standard deviation

The positive square root of the variance

Coefficient of variation (CV)

Standard deviation divided by the mean. Is a relative measure of dispersion and adjusts for differences in the magnitudes of the means

The Sharpe Ratio

Measures an investments excess return per unit of total risk or volatility

High Sharpe Ratio

The better the investment compensates its investors for risk

Skewness coefficient

The degree to which a distribution is not symmetric about its mean

Kurtosis Coefficient

Tells us whether the tails of the distribution are more or less extreme than the normal distribution

Covarience

The linear relationship between 2 variables

Correlation coefficient

The direction and the strength of the linear relationship between x and y (-1 to 1)

Boxplot

A summary that shows the min, Q1, Q2, Q3 and max value of the variable

Empirical rule 1

Approximately 68% of all observations fall in interval x + s

Empirical rule 2

Approximately 95% of all observations fall in the interval x + 2s

Empirical rule 3

Almost all (99.7%) of observations fall in the interval x + 3s

z score

Find the relative position of an observation by dividing the difference of the observation from the mean by the standard deviation

Treat observation as an outlier if z-score is

less than -3 or more than 3

In a boxplot, outliers are present when

They are farther than 1.5 x IQR from the IQR box

Lower fence of box plot

Q1 - (1.5 x IQR)

Upper fence

Q3 + (1.5 x IQR)

z-score formula

(xi - mean)/std dev

=COUNT

Counts the number of cells in range that contains values

=COUNTA

Counts the number of cells in a range that are not empty

=COUNTIF

Counts the number of cells in range that meet a certain criteria

=COUNTIFS

Counts the number of cells in range that meet multiple criterias

=COUNTBLANK

Counts the number of cells in a range that are blank

=LN

Returns the natural log of a certain cell

=SUM

Total value of all cells

=YEARFRAC

Returns the year fraction representing the number of whole days between 2 dates

=VLOOKUP

Searches for a specific value in the first column of a table and returns a value from another column in the same row

=MONTH

Returns the month from 1 (January) to 12 (December)

=MAX

Finds the maximum value of a range of cells

=MIN

Finds the minimum value of a range of cells

=STDEV.S

Finds the standard deviation based on a sample

=CORREL

Returns the correlation coefficient between 2 datasets

=VAR.S

Finds the variance based on the sample