Chapter 3
BME120
NAUDINE BAKKES
1
Experiment
• Process by which measurement is taken or observations are made or as a procedure that
generates outcomes.
o Deterministic
▪ The same outcome is observed each time experiment is done
o Random
▪ One of several outcomes is observed each time experiment is done
Sample space (S)
• Enumeration of all possible outcomes of an experiment
Events
• Collections of outcomes
o Denoted with capital letters
Simple events
• Individual outcomes
Probability (P)
• Likelihood that the outcome or event occurs
• Each outcome has an equal chance of occurring
1
o 𝑃(𝑜𝑢𝑡𝑐𝑜𝑚𝑒) = 𝑁
# 𝑜𝑓 𝑜𝑢𝑡𝑐𝑜𝑚𝑒𝑠 𝑖𝑛 𝑒𝑣𝑒𝑛𝑡
o 𝑃(𝑒𝑣𝑒𝑛𝑡) = 𝑁
Complement
• Consists of all outcomes in the sample space that are not in the event
o Denoted with a ’
• Complement rule
o 𝑃(𝐴′ ) = 1 − 𝑃(𝐴)
2
Union of events
• Consists of outcomes in either or both events
o Denoted by 𝐴 ∪ 𝐵 -OR- 𝐴 𝑜𝑟 𝐵
Intersection of events
• Consists of outcomes that are in both events
o Denoted by: 𝐴 ∩ 𝐵 -OR- 𝐴 𝑎𝑛𝑑 𝐵
Addition rule
• 𝑃(𝐴 ∪ 𝐵) = 𝑃(𝐴) + 𝑃(𝐵) − 𝑃(𝐴∩𝐵)
Mutually exclusive
• Two events that have no outcomes in common
o 𝑃(𝐴∩𝐵) = 0
Conditional probability rule
𝐴 𝑃(𝐴∩𝐵)
• 𝑃 (𝐵 ) = 𝑃(𝐵)
Independence
• Two events where one is not “influenced” by the occurrence or non-occurrence of the
other
𝐴 𝐵
o 𝑃(𝐴) = 𝑃 (𝐵) -OR- 𝑃(𝐵) = 𝑃 (𝐴 )
o 𝑃(𝐴 ∩ 𝐵) = 𝑃(𝐴) × 𝑃(𝐵)
Permutations
• Number of distinct arrangements of n individuals from a population of size N
3
Sampling without replacement
• Done in most applications
o Ensure that sample is comprised of different individuals rather than the same
individual selected more than once
𝑁!
▪ 𝑃𝑁(𝑛) = (𝑁−𝑛)!
Sampling with replacement
• Order of the number of samples is not important
𝑁!
o 𝐶𝑁(𝑛) = 𝑛!(𝑁−𝑛)!
Random variables (X)
• Numerical description of the outcome of a statistical experiment.
o Discrete
▪ May assume only a finite number or an infinite sequence of values
o Continuous
▪ May assume any value in some interval on the real number line
• Generates a probability model or probability distribution
Binomial distribution model
• Three specific attributes:
1. Each trial of the experiment results in only one of two possible outcomes
o Either success or failure
2. The probability of a success on each trial is constant
o 0≤𝑃≤1
3. The trials are independent
Binomial distribution formula
• Two methods used to generate probabilities:
𝑛!
o 𝑃(𝑋 = 𝑥) = 𝑥!(𝑛−𝑥)! 𝑝 𝑥 (1 − 𝑃)𝑛−𝑥
o 𝑃(𝑋 = 𝑥) = 𝐶𝑛(𝑥) 𝑃 𝑥 (1 − 𝑝)𝑛−𝑥
▪ 𝑥 = number of successes of interest
4
▪ 𝑃 = P(success) of any trial
▪ 𝑛 =number of trails
▪ 𝐶𝑛(𝑥) = number of combinations of 𝑥 successes in n trials
Mean and variance of the binomial distribution
• 𝜇 = 𝑛𝑝
• 𝜎 2 = 𝑛𝑝(1 − 𝑃)
Normal distribution
• Most widely used for continuous random variables
• Characteristic of a normal distribution is if the distribution is bell-shaped:
o Horizontal axis
▪ Displays the values of the continuous normal random variable X
o Vertical axis
▪ Each value of the random variable reflects the probability or relative frequency
o Total area under the normal curve is equal to 1
▪ Values in the centre of the distribution are more likely than values at the
extremes
▪ Values in the centre have higher probabilities
Normal probability distribution
𝑥−𝜇 2
1
𝑓(𝑥) = 𝜎√2𝜋 ⅇ −0.5[ ]
• 𝜎
o 𝜇 = Mean of the random variable X
o 𝜎 =Standard deviation of the random variable X
Attributes of normal distribution:
1. Normal distribution is symmetric about the mean
o 𝑝𝑋>𝜇=𝑃𝑋<𝜇=0.5
2. The mean = the median = the mode
3. The mean (𝜇) and variance (𝜎 2 ), completely characterize the normal distribution
o The mean and variance are the only parameters required to compute probabilities in
normal distribution.
5
o X is distributed as 𝑁(𝑛, 𝜎 2 )
4. Empirical rule:
o 𝑃(𝜇 − 𝜎 < 𝑋 < 𝜇 + 𝜎) = 0.68
o 𝑃(𝑛 − 2𝜎 < 𝑥 < 𝑛 + 2𝜎) = 0.95
o 𝑃(𝑛 − 3𝜎 < 𝑋 < 𝑛 + 3𝜎) = 0.99
5. 𝑝(𝑎 < 𝑋 < 𝑏) = the area under the normal curve from a to b
Standard normal distribution
• Normal distribution with 𝜇 = 0 and 𝜎 2 = 1
• Random variable Z is used to refer to the standard normal distribution
𝑥−𝜇
o 𝑍= 𝜎
Percentiles of the normal distribution
• Kth percentile holds k% of scores below it
o 𝑋 = 𝜇 + 𝑍𝜎
Normal approximation to the binomial
• Standardizes binomial random variable X
• Normal approximation is appropriate if:
o 𝑛𝑝 ≥ 5 and 𝑛(1 − 𝑝) ≥ 5
𝑋−𝑛𝑝
o 𝑧=
√𝑛𝑝(1−𝑝)
