Improved the printout of the summary()
of std_selected()
and std_selected_boot()
outputs. It now prints the R-squared increase of the highest order term, as well as the F test for the increase, if the model has one and only one highest order term (e.g., an interaction term). (0.2.9.1)
Added the argument w_values
to cond_effect()
and plolmod()
. Users can specify the values of the moderator (w
) to be used to compute the conditional effects. (0.2.9.2)
update.std_selected()
. Though still not recommended, it should now work more reliably if it needs to be called. (0.2.9.1)stdmod-package
. (0.2.8.9001)summary()
of std_selected()
and std_selected_boot()
outputs. Small numbers are rounded to prevent the use of scientific notation, and small p-values can be printed in formats like p<.001. Users can also control the number of digits in the printout. See the help page of print.summary.std_selected()
to learn more about new arguments (0.2.8.9002).dplyr
from the tests and Suggests. (0.2.7.2)visreg
will be skipped if visreg
is not installed. (0.2.7.3)cond_effect
-class object and the summary of a std_selected
-class object. If one or more variables are standardized but bootstrapping is not requested, users will be recommended to use std_selected_boot()
. (0.2.7.4)stdmod_lavaan()
switched to the bootstrapping algorithm used by lavaan()
. It also updated to allow for partial standardization. To use the older algorithm, set use_old_version()
to TRUE
. (0.2.7.5)print()
method of the summary()
output of std_selected()
. (0.2.6.2)to_standardize
to std_selected()
and std_selected_boot()
. (0.2.6.3)confint.std_selected()
when type = "lm"
and bootstrapping is requested. Should not be an issue because t-based CIs should not be used when bootstrapping is requested. This option is just for testing. (0.2.6.4)to_standardize
or mention it as a shortcut. (0.2.6.5)to_standardize
. (0.2.6.6)summary()
of std_selected()
and std_selected_boot()
outputs. (0.2.4.9001).ggplot2
. (0.2.4.9002)summary()
of std_selected()
. (0.2.4.9003)bibentry()
in CITATION. (0.2.6)std_selected()
: It now works correctly when a variable in the data frame is a factor. (0.2.0.1)confint()
and coef()
methods for cond_effect
-class objects. confint()
can return confidence intervals based on t statistics, which are appropriate in some situations. (0.2.2)print()
method for cond_effect
-class objects can print confidence intervals based on t statistics. (0.2.2)do_boot
to std_selected_boot()
. If set to FALSE
, it will not do bootstrapping. (0.2.3)cond_effect_boot()
will disable bootstrapping in the original call if the output is generated by std_selected_boot()
, to avoid redundant bootstrapping inside bootstrapping. (0.2.3)do_boot
to cond_effect_boot()
. If set to FALSE
, it will not do bootstrapping. (0.2.4)confint()
and vcov()
for std_selected
-class object. If bootstrap CIs are requested, then bootstrap CIs and VCOV based on bootstrapping should be returned. (0.2.0.0)(All major changes after 0.1.7.1)
plotmod()
. It now correctly handles more than two levels when w_method
is set to"percentile"
. (0.1.7.2, 0.1.7.3)(All major changes after 0.1.5)
plotmod()
for plotting moderation effects. This function will check whether a variable is standardized. If yes, will note this in the plot.plotmod()
can also plot a Tumble graph (Bodner, 2016) if graph_type
is set to "tumble"
.plotmod()
instead of visreg::visreg()
.cond_effect()
for computing conditional effects. This function will check which variable(s) is/are standardized. If yes, will note this in the printout.cond_effect_boot()
, a wrapper of cond_effect()
that can form nonparametric bootstrap confidence intervals for the conditional effects, which may be partially or completely standardized.std_selected()
and std_selected_boot()
.stdmod_lavaan()
now returns an object of the class stdmod_lavaan
, with methods print, confint, and coef added.std_selected_boot()
output. Bootstrap confidence intervals are placed next to parameter estimates.vcov()
method for std_selected()
output. If bootstrapping is used, it can return the variance-covariance matrix of the bootstrap estimates.confint()
method for std_selected()
output. If bootstrapping is used, it can return the bootstrap percentile confidence intervals if requested.std_selected()
.