6. Statistics
▸ The discipline of collecting and
analysing numerical data
▸ Usually in large amounts of data
▸ Especially used for the purpose of
making future predictions
8. ▸ Analytical methods for risk management
▸ Statistical modelling of the stock market
▸ Using statistical techniques for making
recommender systems
Application of statistics
9. Modern Applications of
Statistics
▸ Statistical modelling is used to make stock market
predictions
▸ Analytical methods are used to sequence the gnome
▸ Statistics has also helped make significant
headway in the field of artificial intelligence
10. ▸ Statistical analysis depends heavily on the
data
▸ Flawed data can lead to faulty analysis
▸ Statistics cannot be used to understand a
problem in great depth or detail
Loopholes in Statistics
11. Two Key Insights
Results of statistical
analysis can only be
understood in the whole
context of things.
People usually think
they understand
everything if they are
confronted with easy
looking numbers
We should not confuse
“statistically significant” with
something that actually matters.
12. How to spot a bad statistic Diagram featured by
http://slidemodel.com
● Come to the conclusion in different ways. Find how robust a finding
is
● Be wary of scholars using high-powered statistical techniques as a
bludgeon to silence critics who are not specialists.
● Don’t mistake correlation for causation
● Always ask “so what”?
● Combine domain knowledge and statistical analysis to come to the
conclusion
14. FOCUSING ON THE BIG
PICTURE
Statistical analysis of a representative group of consumers
can provide a reasonably accurate, cost-effective snapshot
of the market with faster and cheaper statistics than
attempting a census of very single customer a company
may ever deal with. The statistics can also afford
leadership an unbiased outlook of the market, to avoid
building strategy on uncorroborated presuppositions.
16. ENSURING QUALITY
Statistics provide the means to measure and control
production processes to minimize variations, which lead to
error or waste, and ensure consistency throughout the
process. This saves money by reducing the materials used to
make or remake products, as well as materials lost to
overage and scrap, plus the cost of honoring warranties due
to shipping defective products.