STA 100 Midterm 2
STA 100 Midterm 2
Lecture 6
2
σ𝑌 is the binomial variance = variance of the number of success in a trial
STA 100 Midterm 2 Consolidated Notes
𝑃𝑟 ( 𝐴 𝑎𝑛𝑑 𝐵)
𝑃𝑟 (𝐴|𝐵) = 𝑃𝑟 (𝐵)
Lecture 7
Normal Random Variables
Facts about Normal Random Variables
● The median equals the mean
● 𝑃𝑟 ( 𝑥 > 𝑢 + 𝑎) = 𝑃𝑟 (𝑌 < 𝑢 − 𝑎)
● If Y is distributed normal with the mean µ𝑌, the standard deviation σ𝑌, we
denote this as 𝑌 ∼ 𝑁 ( µ𝑌, σ𝑌)
STA 100 Midterm 2 Consolidated Notes
Z Score Table
● Columns = second decimal place
● Rows = first decimal place
● Middle = probabilities
● The Z table provided only gives the probabilities to the left side of the Z.
Z score
If the Z is less than you can compute the probability directly from Z table:
𝑃𝑟 (𝑍 < 1) = 0. 8413
STA 100 Midterm 2 Consolidated Notes
If the Z is greater than, you cannot compute the probability directly from Z table,
you must take the compliment:
If Z if it is between A and B:
To solve for the Percentile for Y, first find the percentile for Z, then solve for Y:
𝑡ℎ 𝐾
● The 𝐾 Percentile of a Z score is determined by looking up 100
or the
closest value to it in the middle of the Z table, and by seeing what Z score
that is by looking at which row and column it is in.
𝑡ℎ
● The value that is closest to it will become 𝐾
● Use equation provided to solve for the percentile
𝑎 −µ𝑌
I. 𝑃𝑟 ( 𝑌 < 𝑎 ) = 𝑃𝑟 ( 𝑍 < σ𝑌
)
𝑎 −µ𝑌
II. Pr ( Y > a) = 1 - Pr ( Z < σ𝑌
)
𝑏 −µ𝑌 𝑎 −µ𝑌
III. Pr ( a < Y < b ) = Pr ( Z < σ𝑌
) − 𝑃𝑟 ( 𝑍 < σ𝑌
)
IV. To solve for the Percentile for Y, first find the percentile for Z, then solve for
(𝐾) (𝐾)
Y using: 𝑌 = µ𝑌 + 𝑍 σ𝑌
Lecture 8
STA 100 Midterm 2 Consolidated Notes
1. The Empirical Rule: data that is approximately normal follows the 68%,
95%, and 99.7% rule
● 68% of data should lie between one standard deviation of the mean
● 95% of data should lie between two standard deviations of the mean
STA 100 Midterm 2 Consolidated Notes
● 99.7% of data should lie between three standard deviations of the mean.
2. QQ plot: data that is approximately normal will have its percentiles closely
match up with the percentiles of the normal
● If data is normal it will be a straight diagonal line where y=x