Stats Flashcards
Poisson distribution
Describes the number of successes for count data and used to test whether events are randomly distributed in time/space
Binomial test
Tests whether a population proportion matches the null expectation
๐2 goodness-of-fit
Use to test how the distribution of one categorical variable (if continuous use data binning) fits to a null distribution
Agresti-Coull
Used to estimate 95% confidence interval for proportions
Relative risk
The probability of an outcome in the treatment group divided by the probability of the same outcome in a control group
Odds ratio
Used to quantify the magnitude of association between two categorical variables which each have two categories
Fisherโs exact test
Use with a 2x2 contingency table when ๐2 assumptions are not met
Standard normal distribution
Standard normal deviate (Z) indicates how many standard errors the sample mean is, away from the true mean
One (student) sample t-test
Used to compare the mean from a sample of individuals with a value for the proposed population mean
Two (independent) sample t-test
Simplest method to compare the means of a numerical variable between two independent groups
Paired (dependent) sample t-test
Used to compare scores on two different variables but for the same group of cases. Both treatments applied to every sample unit
F test
Used to show variance are the same so that an independent samples t-test can be used
Mann-Whitney U-test
Compares the distribution of two groups. Non-parametric equivalent of the two-sample t-test when non-normal data cannot be transformed
Sign test
Non-parametric alternative to one-sample or paired-sample t-tests when normality assumption is violated
Welchโs approximate t-test
Used to compare means (student t-test) when two samples have unequal variances and possible unequal sample sizes
Shapiro-Wilk test
Evaluates the goodness-of-fit of a normal distribution to a set of data randomly sampled from the population
Permutation test (bootstrapping)
Used to produce a null distribution for the association between two variables with a large sample size
Analysis of variance (ANOVA)
Used for testing the difference among means of
๐ (3 or more) groups simultaneously
One-way ANOVA
Used to compare difference in group means when there is one independent variable (not changed by other variables and organises data into categories)
Two-way ANOVA
Used to compare differences in group means when there is two independent variables (measures the effect of two factors on the independent variable and whether they effect each other)
๐ 2
Use to distinguish between statistically significance and biological importance as shows how much variation in Y is explained by the differences among groups
Analysis of covariance (ANCOVA)
Used to investigate the interaction between continuous explanatory variable (covariate X) and categorical explanatory variable (A) on a response variable . It combines one-way ANOVA and linear regression, therefore, reduced error variance and eliminates confounding variables
Linear regression
Used to predict the value of a response (numerical) variable from values of an explanatory (numerical) variable
Non-linear regression
Used to predict the value of a response (numerical) variable from values of an explanatory (numerical) variable using piecewise linear model
Kruskal-Wallis test
Used to test for the differences among ๐ populations in the means or medians of their distribution (non-parametric equivalent to one-way ANOVA)
Planned comparisons
Used to determine specifically which means are significantly different from each other in and ANOVA test
Tukey-Kramer test
Use to test the differences of all pairs of ๐ means (unplanned)
