Statistics Summary
Importance of Statistics
Stats make you look more trustworthy, but you have to check them carefully so you aren't fooled.
People can use numbers to lie; the goal is to see what the numbers actually mean.
Population
The whole giant group you want to know about, like "everyone in the world" or "every student in a school."
A "Census" is when you actually talk to every single person in that big group.
Samples
A tiny chunk of the big group used to guess what the big group is thinking.
It's way easier than asking everyone on Earth.
Sample Bias
This happens when your "tiny chunk" doesn't represent the whole crowd fairly.
You want a "simple random sample" where everyone has a fair chance to be picked.
Types of Bias
Sample Bias: The group you picked is different from the big crowd.
Voluntary Response Bias: Only the loud or angry people who volunteer to talk get counted.
Self-interest Study: When a company does a study to make themselves look good.
Response Bias: When people lie or give the "right" answer instead of the truth.
Sample Statistic vs. Parameter
Sample Statistic: A number you got from your small group (like an average).
Parameters: The "true" real number for the whole population, which is usually impossible to find.
Variables
The specific stuff you are tracking, like "favorite soda."
Types of Data
Categorical (Qualitative): Sorting things into buckets (like "eye color").
Numerical (Quantitative): Anything you count with numbers (like degrees).
Graphical Data Representation
Pie Charts: Circle slices showing how big a group is compared to others.
Bar Graphs: Towers showing how many people are in each group.
Frequency Tables: A list showing how many times things popped up.
Histograms: Bar graphs for ranges of numbers (like "scores between and ").
Sampling Methods
Simple Random Sample: Picking names out of a hat so everyone is equal.
Stratified Sampling: Splitting the crowd into types (like "Job A" and "Job B") and then picking some from each type.
Cluster Sampling: Picking whole groups (like "3 whole classrooms") at once.
Systematic Sampling: Picking every th person (like every th person in line).
Convenience and Voluntary Response: Picking whoever is nearby or whoever wants to participate.
Levels of Measurement
Nominal Scale: Labels or names with no specific order.
Ordinal Scale: Categories you can rank (like "Happy, Neutral, Sad") but can't measure the exact distance between.
Interval Scale: Numbers indexed with equal gaps, but zero doesn't mean "none."
Ratio Scale: Real math numbers where zero means "nothing exists" (like weight).
Data Collection Methods
Observations: Just watching and taking notes on what happens.
Experiments: Doing something to a group (like giving medicine) to see what happens.
Frequency Distribution
A summary list that helps you see which things happen most often.
Class Width
How big the "bins" are for your numbers; picking the right size makes the data easier to read.