Lecture #4 (Processes) Flashcards
What is the difference between deterministic and probabistic?
Deterministic implies the liklihood of an outcome is 100% certain while probabilistic means the outcome is not with 100% certainty
What is the difference between random and stochastic?
Random is when the outcome is completely random and stochastic is when not all outcomes yield the same probability (chance of seeing a bear on campus)
What is formula for probability?
P (A) = F (A) / F (E) where F (A) = the frequency of something occuring and F (E) = the frequency of all outcomes
What is a compliment?
The frequency of something not happening
What is statistical independance?
Statistical independence means the outcome of the first event has no impact on the next outcome
What is conditional dependence?
The probability changes after the first outcome (you take a king from the deck of cards. Now the chances of getting a queen went from 4/52 to 4/51)
Discrete probability is when
Probability exists with finite value (dice)
Continuous probability is
when the probability is a continuous value (weight of baby)
Dichotomous means
There are two outcomes (yes/no)
This can be expressed using a binary distrubution
Integer or discrete means
There are more than two outcomes
This can be expressed in a poisson distribution
Continuous means
There exists fractional or decimal possibilites
This can be expressed using a normal distribution
What do you call data that doesn’t conform to these probabilty distributions?
Non-parametric statistics
What are some characteristics of a normal distribution probability function?
Bell shaped, mean = 0, standard deviation = 1, total area underneath = 100%,
What are Z-scores?
Z scores calculate the area (probabilty) to the left or right of the mean
What is the percentage of area per each standard deviation?
+/- 1 SD = 68%
+/- 2 SD = 95%
+/- 3 SD = 99.7%
How do you calculate Z-Scores?
Z = (value for data - mean of the original data) / SD of original data
Z = (x - o) / u
What does the central limit theroum argue?
The mean of a large number of independent random samples will be normally distributed, and will be the same as the population mean
What affects the standard error of the mean?
a) the variability of the population which can be measured by the variance and SD
b) the sample size
Is a smaller SEM (standard error of the mean) good or bad?
A smaller SEM is good (more accurate)
How do we calculate the SEM?
SEM = SD / √n
What is a binomial distribution?
There is either an event or there is not, the likelihood is statistically independent
A geometric distribution is
The question of the likelihood of observing an event for the first time after n years
A type of binomial distribution
Poisson distributions can be classified as
Probability distribution for integer or discrete data (catagorized); there are more than two possible outcomes but each can occur completely
Poisson distribution is perfect for
discrete or integer
Normal distribution is perfect for
continuous data
