Normal distribution Flashcards
(14 cards)
Features of Normal distribution (how the graph looks, type of data)
- Normal distribution graphs are symmetrical
- They show continuous data
What adds to 1 in Normal distribution
In normal distribution, area under the graph = 1
Normal distribution: p(x=n) ?
For continuous values, p(x=n)=0, because the value would be infinitesimally small
Normal distribution: Where are the points of inflection on the graph?
Points of inflection = μ ± σ (mean ± one standard deviation)
Normal distribution: How much data is within 1, 2, and 3 standard deviations of the mean? (percentage rule)
1σ = 68%
2σ = 95%
3σ = 99.7%
The normal distribution model format
X ~ N ( μ, σ² )
where μ = mean
and σ = standard deviation
What is the standard normal distribution model?
Z ~ N (0 , 1²)
- Used when μ (mean) or σ (standard deviation) are unknown
- Followed by the coding formula
What is the coding formula for standard normal distribution?
Z = (X - μ) / σ
where Z is number of standard deviations from mean (the value gained from the standard normal model [Z ~ N (0 , 1²)]
μ is the mean
σ is th standard deviation
X is the variable that we want to find the Z value of
What is a Z value in standard normal distribution?
- Z is the number of standard deviations away from the mean
- It tells you how extreme a value is
Using phi (Φ) to represent standard normal probabilities
Φ (a) = p (Z < a)
Normal distribution: How to find μ and σ when given probabilities are given
- Use standard normal distribution [ Z ~ N (0 , 1²) ] to find Z values that correspond to given probabilities
- Insert into coding equation [ Z = (X - μ) / σ ] (in terms of mu and sigma)
- solve simulataneous equations
What are the conditions that allow binomial distributions to be approximated as normal distributions?
- P value is close to 0.5 (means the bd is symmetrical)
- n is large (means the bd is smoother curve, closer to continuous nd)
Approximating binomial distributions as normal distributions: how to find mean (μ) and standard deviation (σ)
X~B (n, p) —-> Y~N (μ, σ²)
μ = n x p n=trials p=probability
σ² = np(1-p)
Approximating binomial distributions as normal distributions: continuity corrections
- if > or < change to ≥ or ≤
- enlarge range by 0.5
ex. p(x>5) = p(x≥6) = p(x≥5.5)