Continuous, Probability and Distributions Flashcards
A PDF is?
A probability distribution function, with a graph that has a definite integral of 1
(from a to b)
How do you find the CDF?
A cumulative distribution function, with a PDF graph that has a definite integral from a to x (where a is the first value)
Quantiles can be found by?
CDF = whatever value (in decimal form)
e.g.
3rd quartile, CDF = 0.75
6th decile, CDF = 0.6
37th percentile, CDF = 0.37
What is a normal distribution?
PDF that looks is symmetrical and looks like a bell, hence its other name the ‘bell curve’
How is a normal distribution arranged?
In the middle is the median, mean and mode (μ) and from there it goes out ±1σ, ±2σ, ±3σ
What does a STANDARD normal distribution have for its values of μ and σ?
μ = 0 σ = 1
Empirical Rule states that?
On a normal distribution:
The values between ±1σ, will represent 68% of the data
The values between ±2σ, will represent 95% of the data
The values between ±3σ, will represent 99.7% of the data
How do you find the z score, and what does a z score do?
z = (x - μ)/σ
Tells you how many standard deviations σ you are away from the mean μ
What is the method for finding probability of data on a normal distribution?
1. First step?
Find the z score, approximate to 2 decimals
What is the method for finding probability of data on a normal distribution?
2. Second step?
Check the table for the t-score:
If it’s less than → straight to the t-score
If it’s greater than → 1 minus the t-score
If it’s between → bigger t-score minus the smaller t-score
note table is in page 633-34
What do you want if x is above the mean?
Bigger z score so you can significantly outrank mediocrity
What do you want if x is below the mean?
Smaller z score so you can significantly outrank complete failure
How do you draw on data onto a normal distribution?
1. First step?
Change to decimals
e.g. If question says value between bottom 10% and top 40% on standard dis table
bottom 10% = 0.1
top 40% = 60% = 0.6
How do you draw on data onto a normal distribution?
2. Second step?
Reverse engineer, look for at t-score FOR z score
e.g. If question says value between bottom 10% and top 40% on standard dis table
t1 = 0.1 → z1 ≈ 1.285
t2 = 0.6 → z2 ≈ 0.265
How do you draw on data onto a normal distribution?
3. Third step?
Place onto PDF for x-values
e.g. If question says value between bottom 10% and top 40% on standard dis table
z1 ≈ 1.285 → x1 ≈ 1.285
z2 ≈ 0.265 → x2 ≈ 0.265