Two types of chi square tests:
goodness of fit
test of independence
Goodness of fit:
The goodness of fit of a statistical model describes how well it fits a set
of observations. Measures of goodness of fit typically summarize the
discrepancy between obs
Review of 4 data types (NOIR)
Nominal (identifier) student id, Joe, Address, credit card #s
naming data
Ordinal order matters (evaluating, identifying)
Star ratings of hotels, grades, Likert scales
-Interval numeric, measurement, has no absolute zero poi
Example: Suppose we want to look at all the different reasons why a
member quit our gym (categorical data), and then further analyze
whether or not the city they live in has something to do with it (also
categorical).
So, we still have categorical variabl
Chi square goodness of fit tests to see if the sample we have fits with the
hypothesized distribution close enough. There is some room for error,
but the test shows us if this is more than natural variation or error.
Continuing with the M&M example, for a
Data Gathering
participant observer: observer is gathering data but is also part
of the study
direct observation: observer is not part of the study and doesnt
influence data at all they are just a collector
interview
focus groups trained moderator, 6-
Feedback from Survey Writing Process
Hard to choose an effective way to administer surveys
Make sure you differentiate between sampling methods and
administration methods
o Its almost more important to have a good sampling method
than an administration me
Hypothesis Testing
Purpose: We are interested in whether or not applying a decision to a situation
yields a change in effect (profits, customers, etc.)
The application of a decision is called a treatment.
Were either interested in if applying the treatmen
Confidence Intervals for Proportions and Means
Before we begin, look at the requirements (assumptions and conditions):
Independence assumptions
o Randomized data condition
10% condition (no more than)
Proportions: success/failure condition
Means: near
Probability Distributions
Discrete versus continuous probability distributions
Events that occur or do not occur
Roulette, tossing a coin or a die, etc.
Binomial
Success/failure
Discrete possible outcomes
On or off, yes or no, 1 or 0
As we want more
Standard Normal
Were interested in finding probability of continuous values.
(examples: test grades, heights, weights, money, etc.)
i.e. What is the probability that someone will receive over an 89 on
the exam?
i.e. What is the probability that a car get