CONCEPTS Flashcards
Used to infer effects from a set of numbers without knowing why the numbers are like they are
Statistics
Yields the same output when stimulated several times with the same input
Deterministic
Yields different output from the same input
Random
Partly deterministic and random
Stochastic
Deterministic is ________ because it has no uncertainty however it is _________
Superior; Rare
Numerical value assigned to an observation which reflects the magnitude or amount of some characterisitc
Measurement
Determines the type of analysis that can be made from the data
Scale of Measurement
Classification of observation into mutually exclusive categories of equal rank
Nominal
Ranked in hierarchy of states
Ordinal
Length of successive intervals in constant
Interval
Have a true zero point
Ratio
Type of probability that the outcomes are countable or finite
Discrete Probability
What are the four properties of binomial distribution?
-There are only TWO POSSIBLE OUTCOMES (“success” or “failure”) for each trial.
-Each trial is INDEPENDENT of all the others.
-The probability of success DOES NOT CHANGE from trial to trial.
-The trials are performed on a FIXED number of times.
Type of probability distribution where the number of trials is not fixed
Negative binomial
Type of probability distribution where the probability of success is very small
Poisson Distribution
A special case of the negative binomial when the interest is the number of trials prior to the initial success
Geometric Distribution
An extension of the binomial where more than two mutually exclusive outcomes are possible
Multinomial Distribution
Governs an experiment called sampling without replacement
Hypergeometric Distribution
Variation in the measurements due to instrumental error becomes apparent with repeated measurements
Experimental error
Variation due to experimental error and between members of a set are sometimes mixed with each other
Confounded error
Statistical tests are based on the assumption that random variables are _________
Normally distributed
Well-defined set (finite or infinite) of elements
Population
Subset of elements taken from a population
Sample
Deliberate or accidental exclusion of certain characteristics from a sample
Bias
Characteristics that describe a population
Paremeters
Characteristics that refer to samples
Statistics
Continuous variable is divided into discrete categories and the number or proportions of observations within each category is represented by the heights of each bars
Histogram
The vertical scale is the relative frequency
Relative frequency histogram
Successive categories show the cumulative numbers or proportions of observations
Cumulative plot
Successive divisions of a distribution (percentile, decile, quartile)
Quantiles
A five-number summary of the median, first and third quartiles, and the minimum and maximum
Box and whisker plot
Average squared deviations
Variance
Square root of variance; same as the units of the measurements
Standard deviation
Dimensionless measure of variability expressed as a fraction of the mean
Coefficient of variation
Joint variation of two variables about their common mean
Covariance
Ratio of the covariance of two variables to the product of their standard deviation (Between +1 and -1)
Correlation Coefficient
Some correlations between variables do not reflect the relationship between them, but are induced by an operation or transformation that has been performed on the variables
Induced Correlation
One in which all variables measured on an individual add to a fixed total such as 1.00 or 100%, which means the individual variables are proportions of a whole
Closed data
Measurements are not expressed as proportions
Open data
The process of computing variances and covariances of ratios can be simplified using logarithms
Logratio Transformation
States that if sets of random samples are taken from any population, and the means calculated for these samples, the sample means will tend to be normally distributed. This becomes more pronounced for larger size.
Central Limits Theorem
The probability of committing a type I error is called the
Level of significance
Null hypothesis and its alternative are mutually exclusive and not all inclusive (T or F)
False; all inclusive