03. Review of Basic Data Analytic Methods Using R

A factor is vector of values that are limited to a fixed set of values (categories). Factors always have levels (the things it allows to be put in - like it had a data validation drop down list). When you first create the factor it will assume the different options are the initial options, after that it will check against this set of levels.

A list is a vector of values which can contain different data types

An array can only contain the same type of data. Arrays can multi-dimensional. i.e. rows and columns and sheets and workbooks etc

A dataframe is a table of vectors or factors; all items of the same length. Individual columns are the same data type, but different columns can be different data types.

A single cell

It plots every variable against every other variable. Also known as a Splom or a ScatterPlot

Samples of the population (not the full population)

Samples of the population (not the full population) are tested to compare against a NULL hypothesis (i.e. checking if there is a statistical significance)

AVOVA are used in hypothesis testing when you have more than two sample populations

where you are picking between the null and the alternative hypothesis

Statistical Power is a measure of how well that test compares against the null

The data follows a normal distribution

The standard deviation is the square root of the deviance

Parametric. Can be one test and two test A “students t-test” is another name for the two sample t test.

Parametric. But can cope with different standard deviations (hence automatically two sample)

Non-Parametric. It tests is two populations of numbers are equally distributed.

We use ANOVA tests to perform multiple comparisons across more than two populations of data. Example: You have an online shop that gives two offers or none at all. We want to find out whether the offers are affecting the number of purchases. The point of the ANOVA is to determine whether the variance in the dataset is due to the spread of values within each group or because of the spread of values between groups.

The P-value is the likelihood that the NULL hypothesis is true. So a low p value means the NULL is very unlikely so happy days.

A vector can only consist of one class and represents a single column

A two dimensional array

Type 1 error

It’s like clustering and looks at the between groups mean sum of squares (between group variance) and the within group mean sum of squares (within group variance)