Statistics

2.

Random Variables

The values which vary

There will be some chance associated with all possible outcomes

There is no pattern for outcome

1.

Graphical Representation

barplot / dotchart

representing each datapoint with a bar/dot respectively

Histogram

frequency distribution

The entire range is split into classes / bins / intervals

count of the data points falling in each interval/bin is plotted

boxplot

100% of data is represented using 4 quarters

Min / lower extreme

Q1 / First quartile / lower quartile

lower 25% of data lies between min and Q1

Q2 / Median

Q3 / upper quartile / 3rd quartile

IQR = Q3 - Q1 = box region

IQR has middle 50% of data

Q4 / upper extreme

between Q4 and Q3 - upper 25% of data

Outliers

any data point which lies beyod = Q+/- 1.5(IQR)

click to edit

Data :

Qualitative

categorical data which cannot be quantified

Quantitative

Continuous datatypes

data that can be represented on a decimal scale

Discrete datatypes

data that cannot be represented on a decimal scale

Measurements

Nominal

Labeling the data

Ordinal

giving order/direction/preference

Interval

quantifying the magnitude of the order/direction

comparision is done on local scale

Ratio

quantifying the magnitude on a global scale

click to edit

click to edit

Probability distribution

Graphical representation of probabilities of events

X-axis = Events

Y-axis = Probabilities

Based on data type of X

Continuous Prob. Dist.

Discrete Prob. Dist.

click to edit