repMean, repTable, repQuantile, and repGlmreport.RdSummarizes the output of the four main functions repMean, repTable,
repQuantile, and repGlm, and provides a single data.frame with all
results.
report(repFunOut, trendDiffs = deprecated(), add=list(),
exclude = c("NcasesValid", "var"),
printGlm = FALSE, round = TRUE, digits = 3, printDeviance = FALSE,
printSE_correction = FALSE)
report2(repFunOut, add=list(), exclude = c("NcasesValid", "var"), printGlm = FALSE,
round = TRUE, digits = 3, printDeviance = FALSE, printSE_correction = FALSE)output of one of the four eatRep-functions.
deprecated. In earlier versions of the package, this argument was used to determine differences in trends. As differences in trends are equivalent to the trend of differences (no matter whether group or cross-level differences), the argument was deprecated. If the user specifies group or cross-level difference along with trends, trends of differences are computed as well.
Optional: additional columns for output. See examples of the jk2-functions
Which parameters should be excluded from reporting?
Only relevant for repGlm: print summary on console?
Logical: should the results be rounded to a limited number of digits?
How many digits should be used for rounding?
Only relevant for repGlm when other than the
identical function is used as link function, and if printGlm is TRUE.
Should the deviance information printed additionally? Note: To print deviance information,
the argument poolMethod of the repGlm function must be set
to "scalar".
Logical: Print the differences of original SEs of cross
differences (method "old") and SEs obtained by the "wec" or "rep" method.
report and report2 differ in the output which is returned. The output of report2
is optimized for further processing, i.e. drawing plots by means of the eatPlot package. For
report, the output is a data frame with at least nine columns.
Denotes the group an analysis belongs to. If no groups were specified and/or analysis for the whole sample were requested, the value of ‘group’ is ‘wholeGroup’.
Denotes the name of the dependent variable in the analysis.
Denotes the mode of the analysis. For example, if a JK2 regression analysis was conducted, ‘modus’ takes the value ‘JK2.glm’. If a mean analysis without any replicates was conducted, ‘modus’ takes the value ‘CONV.mean’.
Denotes whether group mean comparisons or cross-level comparisons were conducted. Without any comparisons, ‘comparison’ takes the value ‘NA’
Denotes the parameter of the corresponding analysis. If regression analysis was applied,
the regression parameter is given. Amongst others, the ‘parameter’ column takes the values
‘(Intercept)’ and ‘gendermale’ if ‘gender’ was the independent variable, for instance.
If mean analysis was applied, the ‘parameter’ column takes the values ‘mean’, ‘sd’,
‘var’, or ‘Nvalid’. See the examples of repMean,repTable,
repQuantile, or repGlm for further details.
Denotes the name of the dependent variable (only if repGlm was called before)
Denotes the estimate of the corresponding analysis.
Denotes the standard error of the corresponding estimate.
Denotes the p value of the estimate.
For report2, the output is a list with four data.frames. The first data.frame plain summarizes the
results in human-readable form. The data.frames 2 to 4 (comparisons, group, estimate) are redundant
to plain and contain the results in a technical presentation suitable for further processing in eatPlot.
The complete results in human-readable form.
An allocation table that indicates which comparison (group comparison or cross-level comparison) relates to which groups.
A table that assigns an ID to each analysis unit. This makes it easier later on to read from the output which comparison relates to which groups. This simplifies the assignment, especially when comparing comparisons (i.e., cross-level differences of group differences).
The results of the analyses, assigned to their IDs.
### see examples of the eatRep main functions.