Exam three I think

ANOVA

Categorical independent, numerical dependent

Binomial

One categorical variable, which is a dichotomous success/failure

Linear regression

Numerical independent, numerical dependent (keyword: predictive)

Chi-square contingency test

Categorical independent, categorical dependent

T-tests (one-sample, two-sample, paired)

Categorical independent, numerical dependent

Correlation

Numerical independent, numerical dependent

Chi-square goodness of fit test

Frequencies associated with one categorical variable

Hypotheses of ANOVA

Null: there is no difference between group means; alternate: at least one group has a mean that is significantly different from the rest

Hypotheses of binomial

Null: probability of success for a single trial is equal to the hypothesized population value; alternate: probability of success for a single trial does not equal the hypothesized population value

Hypotheses of linear regression

Null: there is no predictive relationship between the independent and dependent variables; alternate: there is a predictive relationship between the independent and dependent variables

Hypotheses of the chi-square contingency test

Null: there is no dependent relationship between the categories; alternate: there is a dependent relationship between the two categories

Hypotheses of t-tests

Null: there is no significant difference between population means; alternate: there is a significant difference between population means

Hypotheses of correlation

Null: there is no predictive relationship between the independent and dependent variables; alternate: there is a predictive relationship between the independent and dependent variables

Hypotheses of chi-square goodness of fit

Null: the probability of an event in evenly distributed across all categories of observation; alternate: the probability of an event is not evenly distributed across all categories of observation

Binomial assumptions

Random sampling

Chi-square assumptions

Random sampling, no category with an expected frequency less than one, no more than 20% of categories with an expected frequency less than five

T-test assumptions

Random sampling, normality, homogeneity of variance; alternative: Mann-Whitney U test or Wilcoxon signed-rank test

Correlation assumptions

Random sampling, measurements have a bivariate normal distribution (x/y is linear, frequency distributions for x and y individually are normal); alternative: Spearman’s rank correlation

Regression assumptions

At each x-value there is a population of possible y-values whose mean lies on the regression line, the possible y’s at each x-value are normally distributed, the variance of y-values is the same at all x-values, at each x the y’s represent a random sample of all possible y’s; alternative: logistic regression

ANOVA assumptions

The measurements in each group represent a random sample from the corresponding population, the variable is normally distributed in each of the populations, the variance is the same in all the populations; alternative: Kruskal-Wallis

Tukey-Kramer test

Compares means among groups using t-tests; part of an ANOVA