ANCOVACheckedAssumptionsAndResults: Checks ANCOVA assumptions and runs ANCOVA

In yhhc2/ancovall: Package that automates the process of doing ANCOVA on a large number of dependent variables

Description

The homogeneity of slopes and variance assumptions are checked Normality of residual distribution is not checked. ANCOVA is then performed with all covariates and repeated with only covariates that are significant.

Usage

ANCOVACheckedAssumptionsAndResults(
  inputted.data,
  dependent.variable,
  independent.variable,
  covariates
)

Arguments

inputted.data	A dataframe
dependent.variable	A string that specifies the column name of the column to use as the dependent variable. Column must be numeric.
independent.variable	A string that specifies the column name of the column to use as the independent variable. Column can be numeric or factor. If it's a factor, then it can only have two levels.
covariates	A vector of strings that specifies the columns names of the columns to be used as covariates. Columns can be numeric or factor. If it's a factor, then it can only have two levels.

Details

Homogeneity of slopes and homogeneity of variance are both checked. If the p-value is significant for any of the interaction terms or Levene's test, then this means the assumptions are not met. This is done using the CheckAllAssumptionsANCOVA() function.

ANCOVA computation is performed 3 ways by the ANCOVAWithFormattedOutput() function:

1. Using independent variable and all covariates.

2. Using independent variable and only covariates with p-value <0.05 as determined by method 1.

3. Using independent variable and/or covariates only if they are selected by AIC.

Value

A matrix with two rows. The first row specifies what the values are in the second row. The second row:

The first element is the formula used to evaluate p-value of interaction terms. The next elements are the p-values for each interaction term. Following the p-value for interaction terms is the formula used to evaluate Levene test. The next element is the p-value from the Levene test.

The next element is the formula used to evaluate ANCOVA with all covariates. The elements that are between this formula and the next formula are the p-values for each variable. The element that comes next is the formula that only includes significant covariates along with the independent variable. The elements that are between this formula and the next formula are the p-values for each variable after doing ANCOVA with only variables with coefficient that have p-value <0.05. The element that comes next is the formula that only includes significant covariates (determined by AIC), and independent variable is only included if it's determined to be significant by AIC. The following elements are the p-values for the variables after doing ANCOVA with only the variables determined to be significant by AIC.

Examples

dependent.col <- c(10.1, 11.3, 12.1, 13.7, 14.2, 1.6, 2.3, 3.2, 4.1, 5.3)
independent.col <- as.factor(c(1, 1, 1, 1, 1, 0, 0, 0, 0, 0))
covariate.one.col <- c(1, 2, 3, 4, 5, 1, 2, 3, 4, 5)
covariate.two.col <- as.factor(c(1, 0, 1, 0, 1, 0, 1, 0, 1, 0))

inputted.data <- data.frame(dependent.col, independent.col, covariate.one.col,
                            covariate.two.col)

results <- ANCOVACheckedAssumptionsAndResults(inputted.data, "dependent.col",
                                         "independent.col",
                                         c("covariate.one.col", "covariate.two.col"))

yhhc2/ancovall documentation built on Dec. 23, 2021, 7:19 p.m.