Continuous Distributions
PubH 6450 Biostatistics I
Random Variable
Definition:
The outcome of a random process
• X, Y, Z: notation for the random variable
- Y = cigar smoker status variable
• x, y, z: notation for a value of the random variable
- y1 = 0, indicates person 1 doesn’t smoke cigars
- y3 = 1, indicates person 3 is a cigar smoker
PubH 6450 Biostatistics I slide 2
Continuous Random Variables
A (theoretically) infinite number of possible values
Z = systolic blood
pressure
z1 = 133 mm Hg
z2 = 115 mm Hg
z3 = 140 mm Hg
z4 = 132 mm Hg
z5 = 133 mm Hg
z6 = 138 mm Hg
and so on
PubH 6450 Biostatistics I slide 3
Distributions
Describe how often different values arise from a random
process.
Distribution Types
• Sample distribution: data in a specific sample
- Bar plot, histogram
• Population distribution: data in the population of interest
- Density plots
• Sampling distribution: sample statistics in all possible
samples
PubH 6450 Biostatistics I slide 4
Continuous Distributions
Sample Distribution Population Distribution
X = Systolic Blood Pressure (mmHg)
PubH 6450 Biostatistics I slide 5
Continuous Population Distribution
• Trends
• Proportion below (or
above) a certain value
• Cannot find probability
for a single number
PubH 6450 Biostatistics I slide 6
The Normal Distribution
X = Continuous variable
Notation:
X ~ Norm(µ, σ)
Where
µ = population mean
and σ = population
standard deviation
PubH 6450 Biostatistics I slide 7
Normal Distribution Example
X = Cholesterol in
women aged 20-34
µ = 185 mg/dL and
σ = 39 mg/dL
X ~ Norm(185, 39)
P(X > 240) = 0.079
P(X < 240) = 0.921
P(200 < X < 240) =
0.271
PubH 6450 Biostatistics I slide 8
Normal Probabilities By Formula
How can we find the Normal probability of being below
a specific value, x?
1 𝑥𝑥−μ 2
𝑥𝑥 1 −
P(X ≤ x) = P(X < x) =∫−∞ 𝑒𝑒 2 𝜎𝜎
σ 2𝜋𝜋
PubH 6450 Biostatistics I slide 9
Normal Probabilities By Formula
How can we find the Normal probability of being below
a specific value, x?
1 𝑥𝑥−μ 2
𝑥𝑥 1 −
P(X ≤ x) = P(X < x) =∫−∞ 𝑒𝑒 2 𝜎𝜎
σ 2𝜋𝜋
Hooray for software!
PubH 6450 Biostatistics I slide 10
Normal Probabilities By Software
It will require you to provide:
• The distribution name [normal or gaussian]
• Value of interest [x]
• Population mean [µ]
• Population standard deviation [σ]
Software returns P(getting a value less than x)
PubH 6450 Biostatistics I slide 11
Normal Distribution Example #1
Software P(X > 240) =
gives this
proportion
Interested 1 - P(X < 240) =
in this
proportion 1– 0.921 = 0.079
([Link]=F in R)
PubH 6450 Biostatistics I slide 12
Normal Distribution Example #2
P(200 < X < 240) =
PubH 6450 Biostatistics I slide 13
Normal Distribution Example #2
Software
gives this
proportion
Interested P(X < 240) –
in this P(X < 200)
proportion
= = 0.921 –
Software 0.650
gives this
proportion = 0.271
P(200 < X < 240)
PubH 6450 Biostatistics I slide 14
Normal Distribution Example #3
Middle 95% of
values between
185 – 2*39 = 107
mg/dL, and
95% of values
185 + 2*39 = 263
mg/dL
PubH 6450 Biostatistics I slide 15
The Standard Normal Distribution
Can standardize X
x−μ
Z=
σ
Notation:
Z ~ Norm(0, 1)
Where
µ = 0 and σ = 1
PubH 6450 Biostatistics I slide 16
Standard Normal Distribution Example
x = 240 → z = ?
240 −μ 240 −185
z= = = 1.41
σ 39
Standardize
P(Z > 1.41) =
= P(Z < -1.41) =
0.0793
PubH 6450 Biostatistics I slide 17
68 – 95 – 99.7 rule
68% of values 95% of values 99.7% of values
are within 1 sd of are within 2 sd of are within 3 sd of
the mean the mean the mean
P(-1<z<1) = 0.68 P(-2<z<2) = 0.95 P(-3<z<3) = 0.997
PubH 6450 Biostatistics I slide 18
Other Continuous Distributions
• Exponential
• Lognormal
• Beta
• Gamma
PubH 6450 Biostatistics I slide 19
Key concepts
Density curves
Continuous distributions
• Can’t find P(X = x)
• P(X≤x) = P(X<x)
Normal distribution
• Standard Normal distribution
• 68 – 95 – 99.7 rule
PubH 6450 Biostatistics I
© 2016 Regents of the University of Minnesota. All rights reserved. The University of Minnesota is an equal opportunity
educator and employer. This material is available in alternative formats upon request. Direct requests to 612-624-6669.
