Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapter 20: Non-Normal Probability Distributions - Coggle Diagram
Chapter 20: Non-Normal Probability Distributions
This chapter covers data sets that do not follow the standard bell curve, categorized by continuous and discrete types
Non-Normal Continuous Distributions
Exponential
Used for arrival times or the Mean Time Between Failures (MTBF); it decreases exponentially across the x-axis
Lognormal
Constrained by zero and always features a positive skew; common in wealth distribution or time duration data
Weibull
Extremely flexible; often used in reliability applications where failure probabilities change over time.
Uniform
Data points divided evenly among bins; often points to an error in the measurement system or a non-random sample.
Other Types
Includes Cauchy, Logistic, Laplace, Beta, Gamma, and Triangular distributions.
Non-Normal Discrete Distributions
Binomial
Used for attribute data with only two outcomes (pass/fail, yes/no) where trials are independent.
Poisson
Applied to data distributed randomly within a unit of measurement (e.g., calls per hour); often identified by the word "per" in a metric.
Geometric
Measures the number of trials or "waiting time" before the first occurrence of a success.
Negative Binomial
Determines the probability of a certain number of outcomes before reaching the "" pass or fail.