Hypothesis Testing for Business Analytics

-

Overview

• The Question
• Definition of Hypothesis Testing
• Prerequisite Knowledge
• Problems with Traditional Business Decisions
  • Risks of Subjective Thinking
• Solutions for aforementioned Problems
  • Ensurances of the solution
  • Comparability of Hypothesis Testing to FIRST
• Major Considerations

-

The Question To Answer

• The Question: How do businesses make effective decisions that lead to the betterment of the business?
  • profitability
  • scalability
  • innovation

-

What is Hypothesis Testing

• a hypothesis seeks to explain why something has happened, or what may happen, contingent on given (presumed) conditions
• hypothesis testing is a means of drawing correlations between an assumption and a hypothesis
• We must assume something will happen as the result of an action; Also known as an expectation.

-

Prerequisite Knowledge

• To fully comprehend the upcoming information about hypothesis testing, it is critical that we first discuss
  • What is testing?
  • What are the components of testing?

-

Prerequisite Knowledge

Fundamental Clauses of Test

• There exists an inferrable
• given,
• when,
• and then
• clause for every test.

-

Prerequisite Knowledge

Fundamental Clauses: given

• The given clause declares, initializes, and instantiates all known variable values.
• The given clause of a test sets the state of the “environment” for the entity that needs to be tested.
• The expected output is always given prior to testing. (How else could one test?)
• An object must set its state for the respective test
  • An Inventory must add items before attempting to get an item
  • A Game must have a Player before attempting to play

-

Prerequisite Knowledge

Fundamental Clauses: when

• when clause invokes the method to be tested; it is the action to be performed
• when clause stores output of method invokation; this is the data that is collected as a result of the action

-

Prerequisite Knowledge

Fundamental Clauses: then

• then clause will assert something to be true regarding the output of the action being tested
  • If these assertions fail, then we will be pointed to the defect in the hypothesis.
  • If these assertions pass, then we know our hypothesis works as expected (non-nullary or correlated).

-

Problem

• subjective thinking often guides business strategy
• flawed insights can lead to wasted resources, missed opportunities, and disastrous outcomes.
  • A company’s ability to pivot from these disasters is known as disaster recovery
• “If a company launches a new marketing campaign, then sales will increase”

-

Risks of Subjective Thinking

• Risks can be caused by any of the following
  • flawed thinking
  • flawed observations
  • incomplete data
  • inaccurate data
  • presence of unknown variables

-

Solution

• Back decisions with measurable, repeatable, and verifiable data
• Enables hypotheses to be verified or rejected based objective data rather than our presumptions

-

Solution

Ensurances

• Hypothesis Testing ensures
  • decisions are maintaing innovation and thus profitability
  • decisions which have positive effects are reinforced
  • decisions which have negative effects are nullified quickly
  • theories and assumptions are tested before putting them into action

-

Solution

Hypothesis Testing is Comparable to the FIRST principle

• FIRST principle ensures that decisions are
  • Fast - if we fail, we discontinue
  • Isolated/Independent - we ensure we have measurable correlations, otherwise it is a null-hypothesis
  • Repeatable - if we succeed, we reiterate using our measured data
  • Self-validating - we ensure we have measurable correlations, otherwise it is a null-hypothesis
  • thorough - analyze our sample data to draw a conclusion

-

Major Considerations

1. Null Hypothesis
2. Alternative Hypothesis
3. Sampling

-

Major Considerations

Null Hypothesis

• An assertion that there exists no measurable relationship between variables being tested
• This is the presumed hypothesis until evidence suggests otherwise

-

Major Considerations

Alternative Hypothesis

• An assertion that there exists a measurable relationship between variables being tested
• This is which carries the burden of truth (validity)

-

Example and Discussion

• If we can build long-lasting relations with our clients, clients will feel comfortable paying a premium for our service

-

The Question To Answer

• The Question: How do businesses make effective decisions that lead to the betterment of the business?
• The Answer: Using hypothesis testing, we can make objective measurements against the goals of the busines.
  • Correlations between assumptions can be actualized or nullified