Title:	Mediation Analysis Using Joint Significance
Version:	0.2.3
Description:	A set of helper functions to conduct joint-significance tests for mediation analysis, as recommended by Yzerbyt, Muller, Batailler, & Judd. (2018) <doi:10.1037/pspa0000132>.
License:	MIT + file LICENSE
URL:	https://jsmediation.cedricbatailler.me/, https://github.com/cedricbatailler/JSmediation
BugReports:	https://github.com/cedricbatailler/JSmediation/issues
Depends:	R (≥ 2.10)
Imports:	broom, data.table, dplyr, glue, knitr, magrittr, MASS, purrr, rlang (≥ 0.4.9), stats, tibble, tidyselect
Suggests:	covr, lavaan, markdown, mediation, patchwork, performance, rmarkdown, roxygen2, see (≥ 0.6.5), stringr, testthat (≥ 3.0.2), vdiffr, withr
VignetteBuilder:	knitr
ByteCompile:	true
Config/testthat/edition:	3
Config/testthat/parallel:	true
Encoding:	UTF-8
LazyData:	true
RoxygenNote:	7.3.1
NeedsCompilation:	no
Packaged:	2025-01-26 18:54:21 UTC; cedri
Author:	Cédric Batailler [aut, cre], Dominique Muller [aut], Vincent Yzerbyt [aut], Charles Judd [aut], Yoann Julliard [ctb], Arnold Ho [dtc], Nour Kteily [dtc], Jacqueline Chen [dtc], Simone Dohle [dtc], Michael Siegrist [dtc]
Maintainer:	Cédric Batailler <cedric.batailler@gmail.com>
Repository:	CRAN
Date/Publication:	2025-01-26 19:10:02 UTC

JSmediation

Description

A set of functions to conduct joint-significance tests for mediation analysis.

Details

The main functions of the JSmediation package follow an ⁠mdt_*⁠ pattern. ⁠mdt_*⁠ family functions allow you to conduct joint-significance tests for various mediation models (see Judd, Yzerbyt, & Muller, 2014; Muller, Judd, & Yzerbyt, 2005; Yzerbyt, Muller, Batailler, & Judd, 2018).

The syntax for ⁠mdt_*⁠ family functions is usually the same. The first argument is always a data frame (data) which is followed by the variable names involved in the model (e.g., DV, IV). Because ⁠mdt_*⁠ family functions use non-standard evaluation, these variable names must generally be specified unquoted.

⁠mdt_*⁠ family functions allow you to create an object of class "mediation_model" for which various methods are implemented. The add_index method computes the (moderated) indirect effect index using Monte Carlo for the different mediation models JSmediation offers.

See vignette("jsmediation") for a general introduction and overview of JSmediation.

Author(s)

Maintainer: Cédric Batailler cedric.batailler@gmail.com (ORCID)

Authors:

• Dominique Muller (ORCID)

• Vincent Yzerbyt (ORCID)

• Charles Judd

Other contributors:

• Yoann Julliard (ORCID) [contributor]

• Arnold Ho [data contributor]

• Nour Kteily [data contributor]

• Jacqueline Chen [data contributor]

• Simone Dohle [data contributor]

• Michael Siegrist [data contributor]

References

Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is mediated and mediation is moderated. Journal of Personality and Social Psychology, 89(6), 852–863. doi: 10.1037/0022-3514.89.6.852

Judd, C. M., Yzerbyt, V. Y., & Muller, D. (2014). Mediation and Moderation. In H. T. Reis & C. M. Judd (Eds.), Handbook of Research Methods in Social and Personality Psychology (2nd ed., pp. 653–676). New York: Cambridge University Press.

Yzerbyt, V., Muller, D., Batailler, C., & Judd, C. M. (2018). New recommendations for testing indirect effects in mediational models: The need to report and test component paths. Journal of Personality and Social Psychology, 115(6), 929–943. doi: 10.1037/pspa0000132

See Also

Useful links:

• https://jsmediation.cedricbatailler.me/

• https://github.com/cedricbatailler/JSmediation

• Report bugs at https://github.com/cedricbatailler/JSmediation/issues

Pipe operator

Description

See ⁠magrittr::[\%>\%][magrittr::\%>\%]⁠ for details.

Usage

lhs %>% rhs

Adds an indirect effect index to a fitted mediation model

Description

add_index is a generic function that adds a (moderated) indirect effect index to an object created with an ⁠mdt_*⁠ family function. This index is computed using Monte Carlo methods. This function invokes particular methods depending of the class of the mediation model. For example, with a model fitted with mdt_simple, add_index will invoke add_index.simple_mediation.

Usage

add_index(mediation_model, times = 5000, level = 0.05, ...)

Arguments

Value

An object of the same class as mediation_model, but with index added for later use.

add_index method for moderated mediation

Description

Adds the confidence interval for the index of moderated mediation to a model fitted with mdt_moderated.

Usage

## S3 method for class 'moderated_mediation'
add_index(mediation_model, times = 5000, level = 0.05, stage = NULL, ...)

Arguments

Details

Indirect effect moderation index for moderated mediation uses a, a \times Mod, b, and b \times Mod estimates and their standard errors to compute the appropriate index product distribution using Monte Carlo methods (see Muller, Judd, & Yzerbyt, 2005).

JSmediation supports different types of mediated indirect effect index:

• Stage 1: computes the product between a \times Mod and b.

• Stage 2: computes the product between a and b \times Mod.

• Total: computes the sum of Stage 1 and Stage 2 distribution.

References

Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is mediated and mediation is moderated. Journal of Personality and Social Psychology, 89(6), 852-863. doi: 10.1037/0022-3514.89.6.852

Examples

## getting a stage 1 moderated indirect effect index
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
ho_et_al <- standardize_variable(ho_et_al, c(linkedfate, sdo), suffix ="c")
moderated_model <- mdt_moderated(data = ho_et_al,
                                 IV = condition_c,
                                 DV = hypodescent,
                                 M = linkedfate_c,
                                 Mod = sdo_c)
add_index(moderated_model, stage = 1)

add_index method for simple mediation

Description

Adds confidence interval for the index of mediation to a model fitted with mdt_simple.

Usage

## S3 method for class 'simple_mediation'
add_index(mediation_model, times = 5000, level = 0.05, ...)

Arguments

Details

Indirect effect index for simple mediation uses a and b estimates and their standard errors to compute the ab product distribution using Monte Carlo methods (see MacKinnon, Lockwood, & Williams, 2004).

References

MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence Limits for the Indirect Effect: Distribution of the Product and Resampling Methods. Multivariate Behavioral Research, 39(1), 99-128. doi: 10.1207/s15327906mbr3901_4

Examples

## getting an indirect effect index
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
simple_model <- mdt_simple(data = ho_et_al,
                           IV = condition_c,
                           DV = hypodescent,
                           M = linkedfate)
add_index(simple_model)

add_index method for within-participant mediation

Description

Adds the confidence interval for the index of within-participant mediation to a model fitted with mdt_within or mdt_within_wide.

Usage

## S3 method for class 'within_participant_mediation'
add_index(mediation_model, times = 5000, level = 0.05, ...)

Arguments

Details

Indirect effect index for within-participant mediation uses a and b estimates and their standard error to compute the ab product distribution using Monte Carlo methods (see MacKinnon, Lockwood, & Williams, 2004).

References

MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence Limits for the Indirect Effect: Distribution of the Product and Resampling Methods. Multivariate Behavioral Research, 39(1), 99-128. doi: 10.1207/s15327906mbr3901_4

Examples

## getting an indirect effect index
within_model <- mdt_within(data = dohle_siegrist,
                           IV = name,
                           DV = willingness,
                           M = hazardousness,
                           grouping = participant)
add_index(within_model)

Creates an APA formatted report from a significance test

Description

Create an APA formatted report from the test of a specific term in a linear model.

Usage

apastylr(model, term)

Arguments

Value

An APA formatted character string.

Examples

data(ho_et_al)
test <- lm(hypodescent ~ linkedfate, ho_et_al)
apastylr(test, "linkedfate")

Builds a contrast code from character vector

Description

This function constructs a contrast code from a character variable. It is useful when one needs to recode a two-category character variable to a numeric one.

Usage

build_contrast(vector, cond_a, cond_b)

Arguments

Details

The lm method supports factor and character variables by dummy coding them. Dummy coding can make the interpretation of regression coefficient difficult or at least more difficult than contrast coding. Contrast-coded-variable coefficients interpretation is particularly useful when conducting a joint-significance test.

Value

A numeric vector.

See Also

scale for centering continuous numeric variable.

Examples

  data(ho_et_al)

  ho_et_al$condition_contrast <- build_contrast(ho_et_al$condition,
                                                "Low discrimination",
                                                "High discrimination")

 head(ho_et_al)

Test assumptions for models underlying the mediation

Description

When conducting a joint-significant test, different models are fitted to the data. This function tests assumptions regarding these models using the performance package.

The assumptions test are performed using check_normality, check_heteroscedasticity, and check_outliers.

Note that check_assumptions returns a mediation_model object.

Usage

check_assumptions(
  mediation_model,
  tests = c("normality", "heteroscedasticity")
)

Arguments

Value

Invisibly returns an object of class mediation_model.

See Also

Other assumption checks: plot_assumptions()

Examples

data(ho_et_al)

ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")

my_model <-
  mdt_simple(data = ho_et_al,
             IV = condition_c,
             DV = hypodescent,
             M = linkedfate)

check_assumptions(my_model)

Compute the indirect effect index for a specific value of the moderator

Description

When computing a moderated mediation, one assesses whether an indirect effect changes according a moderator value (Muller et al., 2005). mdt_moderated makes it easy to assess moderated mediation, but it does not allow accessing the indirect effect for a specific moderator values. compute_indirect_effect_for fills this gap.

Usage

compute_indirect_effect_for(
  mediation_model,
  Mod = 0,
  times = 5000,
  level = 0.05
)

Arguments

Details

The approach used by compute_indirect_effect_for is similar to the approach used for simple slope analyses. Specifically, it will fit a new moderated mediation model, but with a data set with a different variable coding. Behind the scenes, compute_indirect_effect_for adjusts the moderator variable coding, so that the value we want to compute the indirect effect for is now 0.

Once done, a new moderated mediation model is applied using the new data set. Because of the new coding, and because of how one interprets coefficients in a linear regression, a \times b is now the indirect effect we wanted to compute (see the Models section).

Thanks to the returned values of a and b (b_51 and b_64, see the Models section), it is now easy to compute a \times b. compute_indirect_effect_for uses the same approach than the add_index function. A Monte Carlo simulation is used to compute the indirect effect index (MacKinnon et al., 2004).

Models

In a moderated mediation model, three models are used. compute_indirect_effect_for uses the same model specification as mdt_moderated:

• Y_i = b_{40} + \mathbf{b_{41}} X_i + b_{42} Mo_i + \mathbf{b_{43}} XMo_i

• M_i = b_{50} + \mathbf{b_{51}} X_i + b_{52} Mo_i + \mathbf{b_{53} XMo_i}

• Y_i = b_{60} + \mathbf{c'_{61}} X_i + b_{62} Mo_i + \mathbf{b_{63} Xmo_i} + \mathbf{b_{64} Me_i} + \mathbf{b_{65} MeMo_i}

with Y_i, the outcome value for the ith observation, X_i, the predictor value for the ith observation, Mo_i, the moderator value for the ith observation, and M_i, the mediator value for the ith observation.

Coefficients associated with a, a \times Mod, b, b \times Mod, c, c \times Mod, c', and c' \times Mod, paths are respectively b_{51}, b_{53}, b_{64}, b_{65}, b_{41}, b_{43}, b_{61}, and b_{63} (see Muller et al., 2005).

References

MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence Limits for the Indirect Effect: Distribution of the Product and Resampling Methods. Multivariate Behavioral Research, 39(1), 99-128. doi: 10.1207/s15327906mbr3901_4

Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is mediated and mediation is moderated. Journal of Personality and Social Psychology, 89(6), 852-863. doi: 10.1037/0022-3514.89.6.852

Examples

# compute an indirect effect index for a specific value in a moderated
# mediation.
data(ho_et_al)
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
ho_et_al <- standardize_variable(ho_et_al, c(linkedfate, sdo))
moderated_mediation_model <- mdt_moderated(data = ho_et_al,
                                           DV = hypodescent,
                                           IV = condition_c,
                                           M = linkedfate,
                                           Mod = sdo)
compute_indirect_effect_for(moderated_mediation_model, Mod = 0)

Displays models from a mediation object

Description

When conducting a joint-significance test, different models are fitted to the data. This function helps you see a summary of the models that have been used in an object of class mediation_model.

Usage

display_models(mediation_model)

Arguments

Value

A list of summary.lm objects.

Examples

data(ho_et_al)
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
my_model <-
  mdt_simple(data = ho_et_al,
             IV = condition_c,
             DV = hypodescent,
             M = linkedfate)

display_models(my_model)

Dohle and Siegrist (2014, Exp 1) illustrating within-subject analysis (long-format)

Description

A data set containing data from Dohle and Siegrist (2014)'s Experiment 1 that can be used to conduct within-subject joint-significance test. In this experiment, researchers are interested in the effect of name complexity on willingness to buy a drug. The specific hypothesis is that complex drug names are perceived as more hazardous, which makes someone less likely to buy the drug. Researchers used a within-subject design.

This data set is in a long-format, see mdt_within to conduct a within-participant mediation analysis with this data set.

Usage

data("dohle_siegrist")

Format

A data frame with 44 rows and 4 variables:

participant

Participant number.

name

Name of the drugs ("simple" vs. "complex").

hazardousness

Mean estimated hazardousness.

willingness

Mean willingness to buy.

References

Dohle, S., & Siegrist, M. (2014). Fluency of pharmaceutical drug names predicts perceived hazardousness, assumed side effects and willingness to buy. Journal of Health Psychology, 19(10), 1241-1249. doi: 10.1177/1359105313488974

Dohle and Siegrist (2014, Exp 1) illustrating within-subject analysis (wide-format)

Description

A data set containing data from Dohle and Siegrist (2014)'s Experiment 1 that can be used to conduct within-subject joint-significance test. In this experiment, researchers are interested in the effect of name complexity on willingness to buy a drug. The specific hypothesis is that complex drug names are perceived as more hazardous, which makes someone less likely to buy the drug. Researchers used a within-subject design.

This data set is in a wide format, see mdt_within_wide to conduct a within-participant mediation analysis with this dataset.

Usage

data("dohle_siegrist_wide")

Format

A data frame with 22 rows and 5 variables:

participant

Participant number.

hazardousness_c

Hazardousness for complex drug name.

hazardousness_s

Hazardousness for simple drug name.

willingness_c

Willingness to buy for complex drug name.

willingness_s

Willingness to buy for simple drug name.

References

Dohle, S., & Siegrist, M. (2014). Fluency of pharmaceutical drug names predicts perceived hazardousness, assumed side effects and willingness to buy. Journal of Health Psychology, 19(10), 1241-1249. doi: 10.1177/1359105313488974

Extracts a single model from a mediation_model object

Description

When conducting a joint-significant test, different models are fitted to the data. This function helps you access the models used in an object of class mediation_model.

Usage

extract_model(mediation_model, step = NULL)

Arguments

Value

An lm object.

See Also

extract_models to access a list of every model relevant to joint-significance testing.

Other extract functions: extract_models(), extract_tidy_models()

Examples

data(ho_et_al)
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
my_model <-
  mdt_simple(data = ho_et_al,
             IV = condition_c,
             DV = hypodescent,
             M = linkedfate)

extract_model(my_model, step = "X -> M")

Extracts models from a mediation_model object

Description

When conducting a joint-significant test, different models are fitted to the data. This function helps accessing the models used in an object of class mediation_model.

Usage

extract_models(mediation_model)

Arguments

Value

A list of lm objects.

See Also

Other extract functions: extract_model(), extract_tidy_models()

Examples

data(ho_et_al)
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
my_model <-
  mdt_simple(data = ho_et_al,
             IV = condition_c,
             DV = hypodescent,
             M = linkedfate)

extract_models(my_model)

Extracts models from a mediation object as a data frame

Description

When conducting a joint significant test, different models are fitted to the data. This function helps you access the models used in an object of class mediation_model.

Usage

extract_tidy_models(mediation_model)

Arguments

Value

A data frame.

See Also

Other extract functions: extract_model(), extract_models()

Examples

data(ho_et_al)
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
my_model <-
  mdt_simple(data = ho_et_al,
             IV = condition_c,
             DV = hypodescent,
             M = linkedfate)

extract_tidy_models(my_model)

Data set showing simple and moderated mediation analysis

Description

A data set containing data from Experiment 3 from Ho, Kteiley, and Chen (2017). In this experiment, the authors hypothesized that presenting a text stating that Black-White biracials were discriminated against would lead Black participants to associate Black-White biracials more with their lower status parent group than their higher status parent group, according to the rule of hypodescent. In this experiment, the authors tested if this effect was mediated by the sense of linked fate between discriminated Black-White biracials and Black participants.

Note that this data set does not include the participants who were in the discrimination control condition in the study conducted by Ho, Kteiley and Chen (2017).

See mdt_simple and mdt_moderated to conduct a simple mediation or a moderated mediation analysis with this dataset.

Usage

data("ho_et_al")

Format

A data frame with 824 rows and 5 variables:

id

An incremental index.

condition

Experimental condition (High discrimination vs. Low discrimination).

sdo

Score at an SDO scale.

linkedfate

Score at an 8-item linked fate measure.

hypodescent

Score at a 3-item measure of hypodescent.

References

Ho, A. K., Kteily, N. S., & Chen, J. M. (2017). “You’re one of us”: Black Americans’ use of hypodescent and its association with egalitarianism. Journal of Personality and Social Psychology, 113(5), 753-768. doi: 10.1037/pspi0000107

Fits a moderated mediation model

Description

Given a data frame, a predictor (IV), an outcome (DV), a mediator (M), and a moderator (Mod) conducts a joint-significant test for moderated mediation (see Yzerbyt, Muller, Batailler, & Judd, 2018).You can learn about moderated mediation in vignette("moderated-mediation")

add_index.moderated_mediation computes the moderated mediation index. compute_indirect_effect_for is used to compute the indirect effect index for a specific value of the moderator.

Usage

mdt_moderated(data, IV, DV, M, Mod)

Arguments

Details

With moderated mediation analysis, one tests whether the indirect effect of X on Y through M is moderated by Mod. The hypothesis behind this test is that X has an effect on M (a) which has an effect on Y (b), meaning that X has an indirect effect on Y through M.

Total moderation of the indirect effect of X on Y can be described as follows:

c * Mod = c' * Mod + (a * Mod) * b + a * (b * Mod)

with c * Mod the total moderation of the indirect effect, c' * Mod the moderation of the direct effect, (a * Mod) * b, the moderation of the indirect effect passing by the moderation of a, and a * (b * Mod), the moderation of the indirect effect passing by the moderation of b (see Models section; Muller et al., 2005).

Either both a * Mod and b or both a and b * Mod need to be simultaneously significant for a moderation of the indirect effect to be claimed (Muller et al., 2005).

Value

Returns an object of class "mediation_model".

An object of class "mediation_model" is a list containing at least the components:

Models

In a moderated mediation model, three models will be used:

• Y_i = b_{40} + \mathbf{b_{41}} X_i + b_{42} Mo_i + \mathbf{b_{43}} XMo_i

• M_i = b_{50} + \mathbf{b_{51}} X_i + b_{52} Mo_i + \mathbf{b_{53} XMo_i}

• Y_i = b_{60} + \mathbf{c'_{61}} X_i + b_{62} Mo_i + \mathbf{b_{63} Xmo_i} + \mathbf{b_{64} Me_i} + \mathbf{b_{65} MeMo_i}

with Y_i, the outcome value for the ith observation, X_i, the predictor value for the ith observation, Mo_i, the moderator value for the ith observation, and M_i, the mediator value for the ith observation.

Coefficients associated with a, a \times Mod, b, b \times Mod, c, c \times Mod, c', and c' \times Mod, paths are respectively b_{51}, b_{53}, b_{64}, b_{65}, b_{41}, b_{43}, b_{61}, and b_{63} (see Muller et al., 2005).

Variable coding

Because joint-significance tests use linear models behind the scenes, variables involved in the model have to be numeric. mdt_simple will give an error if non-numeric variables are specified in the model.

If you need to convert a dichotomous categorical variable to a numeric one, please refer to the build_contrast function.

Note that variable coding is especially important in models with multiple predictors as is the case in the model used to conduct a joint-significance test of moderated mediation. Muller et al. (2005) recommend using variables that are either contrast-coded or centered. Using mdt_moderated with a DV, a mediator, or a moderator that is neither contrast-coded nor centered will give a warning message.

References

Muller, D., Judd, C. M., & Yzerbyt, V. Y. (2005). When moderation is mediated and mediation is moderated. Journal of Personality and Social Psychology, 89(6), 852-863. doi: 10.1037/0022-3514.89.6.852

Yzerbyt, V., Muller, D., Batailler, C., & Judd, C. M. (2018). New recommendations for testing indirect effects in mediational models: The need to report and test component paths. Journal of Personality and Social Psychology, 115(6), 929–943. doi: 10.1037/pspa0000132

See Also

Other mediation models: mdt_simple(), mdt_within()

Joint-significance test for simple mediation

Description

Given a data frame, a predictor (IV), an outcome (DV), and a mediator (M), conducts a joint-significant test for simple mediation (see Yzerbyt, Muller, Batailler, & Judd, 2018).

Usage

mdt_simple(data, IV, DV, M)

Arguments

Details

With simple mediation analysis, one is interested in finding if the effect of X on Y goes through a third variable M. The hypothesis behind this test is that X has an effect on M (a) that has an effect on Y (b), meaning that X has an indirect effect on Y through M.

The total effect of X on Y can be described as follows:

c = c' + ab

with c the total effect of X on Y, c' the direct of X on Y, and ab the indirect effect of X on Y through M (see Models section).

To assess whether the indirect effect is different from the null, one has to assess the significance against the null for both a (the effect of X on M) and b (effect of M on Y controlling for the effect of X). Both a and b need to be simultaneously significant for an indirect effect to be claimed (Cohen & Cohen, 1983; Yzerbyt, Muller, Batailler, & Judd, 2018).

Value

Returns an object of class "mediation_model".

An object of class "mediation_model" is a list containing at least the components:

Models

In a simple mediation model, three models will be fitted:

• Y_i = b_{10} + \mathbf{c_{11}} X_i

• M_i = b_{20} + \mathbf{a_{21}} X_i

• Y_i = b_{30} + \mathbf{c'_{31}} X_i + \mathbf{b_{32}} M_i

with Y_i, the outcome value for the ith observation, X_i, the predictor value for the ith observation, and M_i, the mediator value for the ith observation (Cohen & Cohen, 1983; Yzerbyt, Muller, Batailler, & Judd, 2018).

Coefficients associated with a, b, c, and c' paths are respectively a_{21}, b_{32}, c_{11}, and c'_{31}.

Variable coding

Because joint-significance tests uses linear models behind the scenes, variables involved in the model have to be numeric. mdt_simple will give an error if non-numeric variables are specified in the model.

To convert a dichotomous categorical variable to a numeric one, please refer to the build_contrast function.

References

Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed). Hillsdale, N.J: L. Erlbaum Associates.

Yzerbyt, V., Muller, D., Batailler, C., & Judd, C. M. (2018). New recommendations for testing indirect effects in mediational models: The need to report and test component paths. Journal of Personality and Social Psychology, 115(6), 929–943. doi: 10.1037/pspa0000132

See Also

Other mediation models: mdt_moderated(), mdt_within()

Examples

## fit a simple mediation model
data(ho_et_al)
ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")
mdt_simple(data = ho_et_al,
           IV = condition_c,
           DV = hypodescent,
           M = linkedfate)

Joint-significance test for within-participant mediation

Description

Given a data frame, a predictor (IV), an outcome (DV), a mediator (M), and a grouping variable (group) conducts a joint-significant test for within-participant mediation (see Yzerbyt, Muller, Batailler, & Judd, 2018).

Usage

mdt_within(data, IV, DV, M, grouping, default_coding = TRUE)

Arguments

Details

With within-participant mediation analysis, one tests whether the effect of X on Y goes through a third variable M. The specificity of within-participant mediation analysis lies in the repeated measures design it relies on. With such a design, each sampled unit (e.g., participant) is measured on the dependent variable Y and the mediator M in the two conditions of X. The hypothesis behind this test is that X has an effect on M (a) which has an effect on Y (b), meaning that X has an indirect effect on Y through M.

As with simple mediation, the total effect of X on Y can be conceptually described as follows:

c = c' + ab

with c the total effect of X on Y, c' the direct of X on Y, and ab the indirect effect of X on Y through M (see Models section).

To assess whether the indirect effect is different from the null, one has to assess the significance against the null for both a (the effect of X on M) and b (effect of M on Y controlling for the effect of X). Both a and b need to be simultaneously significant for an indirect effect to be claimed (Judd, Kenny, & McClelland, 2001; Montoya & Hayes, 2011).

Value

Returns an object of class "mediation_model".

An object of class "mediation_model" is a list containing at least the components:

Models

For within-participant mediation, three models will be fitted:

• Y_{2i} - Y_{1i} = c_{11}

• M_{2i} - M_{1i} = a_{21}

• Y_{2i} - Y_{1i} = c'_{31} + b_{32}(M_{2i} - M_{1i}) + d_{33}[0.5(M_{1i} + M_{2i}) - 0.5(\overline{M_{1} + M_{2}})]

with Y_{2i} - Y_{1i} the difference score between DV conditions for the outcome variable for the ith observation, M_{2i} - M_{1i} the difference score between DV conditions for the mediator variable for the ith observation, M_{1i} + M_{2i} the sum of mediator variables values for DV conditions for the ith observation, and \overline{M_{1} + M_{2}} the mean sum of mediator variables values for DV conditions across observations (see Montoya & Hayes, 2011).

Coefficients associated with a, b, c, and c' paths are respectively a_{21}, b_{32}, c_{11}, and c'_{31}.

Data formatting

To be consistent with other ⁠mdt_*⁠ family functions, mdt_within takes a long-format data frame as data argument. With this kind of format, each sampled unit has two rows, one for the first within-participant condition and one for the second within-participant condition. In addition, each row has one observation for the outcome and one observation for the mediator (see dohle_siegrist for an example.

Because such formatting is not the most common among social scientists interested in within-participant mediation, JSmediation contains the mdt_within_wide function which handles wide-formatted data input (but is syntax-inconsistent with other ⁠mdt_*⁠ family functions).

Variable coding

Models underlying within-participant mediation use difference scores as DV (see Models section). Because the function input does not allow the user to specify how the difference scores should be computed, mdt_within has a default coding.

mdt_within's default behavior is to compute the difference score so the total effect (the effect of X on Y) will be positive and compute the other difference scores accordingly. That is, if mdt_within has to use Y_{2i} - Y_{1i} (instead of Y_{1i} - Y_{2i}) so that c_{11} is positive, it will use M_{2i} - M_{1i} (instead of M_{1i} - M_{2i} in the other models.

User can choose to have a negative total effect by using the default_coding argument.

Note that DV and M have to be numeric.

References

Judd, C. M., Kenny, D. A., & McClelland, G. H. (2001). Estimating and testing mediation and moderation in within-subject designs. Psychological Methods, 6(2), 115-134. doi: 10.1037//1082-989X.6.2.115

Montoya, A. K., & Hayes, A. F. (2017). Two-condition within-participant statistical mediation analysis: A path-analytic framework. Psychological Methods, 22(1), 6-27. doi: 10.1037/met0000086

Yzerbyt, V., Muller, D., Batailler, C., & Judd, C. M. (2018). New recommendations for testing indirect effects in mediational models: The need to report and test component paths. Journal of Personality and Social Psychology, 115(6), 929–943. doi: 10.1037/pspa0000132

See Also

Other mediation models: mdt_moderated(), mdt_simple()

Joint-significance test for simple mediation (wide-format input)

Description

Given a data frame, a predictor (IV), an outcome (DV), a mediator (M), and a grouping variable (group) conducts a joint-significant test for within-participant mediation (see Yzerbyt, Muller, Batailler, & Judd, 2018).

Usage

mdt_within_wide(data, DV_A, DV_B, M_A, M_B)

Arguments

Details

With within-participant mediation analysis, one tests whether the effect of X on Y goes through a third variable M. The specificity of within-participant mediation analysis lies in the repeated measures design it relies on. With such a design, each sampled unit (e.g., participant) is measured on the dependent variable Y and the mediator M in the two conditions of X. The hypothesis behind this test is that X has an effect on M (a) which has an effect on Y (b), meaning that X has an indirect effect on Y through M.

As with simple mediation, the total effect of X on Y can be conceptually described as follows:

c = c' + ab

with c the total effect of X on Y, c' the direct of X on Y, and ab the indirect effect of X on Y through M (see Models section).

To assess whether the indirect effect is different from the null, one has to assess the significance against the null for both a (the effect of X on M) and b (effect of M on Y controlling for the effect of X). Both a and b need to be simultaneously significant for an indirect effect to be claimed (Judd, Kenny, & McClelland, 2001; Montoya & Hayes, 2011).

Value

Returns an object of class "mediation_model".

An object of class "mediation_model" is a list containing at least the components:

Data formatting

To be consistent with other ⁠mdt_*⁠ family functions, mdt_within takes a long-format data frame as data argument. With this kind of format, each sampled unit has two rows, one for the first within-participant condition and one for the second within-participant condition. In addition, each row has one observation for the outcome and one observation for the mediator (see dohle_siegrist for an example.

Because such formatting is not the most common among social scientists interested in within-participant mediation, JSmediation contains the mdt_within_wide function which handles wide-formatted data input (but is syntax-inconsistent with other ⁠mdt_*⁠ family functions).

Variable coding

Models underlying within-participant mediation use difference scores as DV (see Models section). mdt_within_wide uses M_A - M_B and DV_A - DV_B in these models.

Models

For within-participant mediation, three models will be fitted:

• Y_{2i} - Y_{1i} = c_{11}

• M_{2i} - M_{1i} = a_{21}

• Y_{2i} - Y_{1i} = c'_{31} + b_{32}(M_{2i} - M_{1i}) + d_{33}[0.5(M_{1i} + M_{2i}) - 0.5(\overline{M_{1} + M_{2}})]

with Y_{2i} - Y_{1i} the difference score between DV conditions for the outcome variable for the ith observation, M_{2i} - M_{1i} the difference score between DV conditions for the mediator variable for the ith observation, M_{1i} + M_{2i} the sum of mediator variables values for DV conditions for the ith observation, and \overline{M_{1} + M_{2}} the mean sum of mediator variables values for DV conditions across observations (see Montoya & Hayes, 2011).

Coefficients associated with a, b, c, and c' paths are respectively a_{21}, b_{32}, c_{11}, and c'_{31}.

References

Judd, C. M., Kenny, D. A., & McClelland, G. H. (2001). Estimating and testing mediation and moderation in within-subject designs. Psychological Methods, 6(2), 115-134. doi: 10.1037//1082-989X.6.2.115

Montoya, A. K., & Hayes, A. F. (2017). Two-condition within-participant statistical mediation analysis: A path-analytic framework. Psychological Methods, 22(1), 6-27. doi: 10.1037/met0000086

Yzerbyt, V., Muller, D., Batailler, C., & Judd, C. M. (2018). New recommendations for testing indirect effects in mediational models: The need to report and test component paths. Journal of Personality and Social Psychology, 115(6), 929–943. doi: 10.1037/pspa0000132

Returns diagnostic plots for the linear model used in a mediation

Description

When conducting a joint-significant test, different models are fitted to the data. This function returns diagnostic plots for each of the model used in the mediation model. check_assumptions_plot uses the performance and see packages behind the scenes to provide the different plots.

This function is best used in an interactive context.

Usage

plot_assumptions(
  mediation_model,
  tests = c("normality", "heteroscedasticity", "outliers")
)

Arguments

Value

Invisibly returns an object of class mediation_model.

See Also

Other assumption checks: check_assumptions()

Examples

data(ho_et_al)

ho_et_al$condition_c <- build_contrast(ho_et_al$condition,
                                       "Low discrimination",
                                       "High discrimination")

my_model <-
  mdt_simple(data = ho_et_al,
             IV = condition_c,
             DV = hypodescent,
             M = linkedfate)

plot_assumptions(my_model)

Print method for object of class indirect_index

Description

Print a summary for an indirect effect index created with add_index() method.

Usage

## S3 method for class 'indirect_index'
print(x, digits = 3, ...)

Arguments

Print method for object of class mediation_model

Description

Print a summary for a mediation model represented by a mediation_model object.

Usage

## S3 method for class 'mediation_model'
print(x, digits = 3, ...)

Arguments

Standardize variables in a data set.

Description

standardize_variable() standardizes the selected columns in a data.frame using base::scale(). By default, this function overwrites the column to be scaled. Use the suffix argument to avoid this behavior.

standardize_variable() and standardise_variable() are synonyms.

Usage

standardize_variable(data, cols = dplyr::everything(), suffix = NULL)

standardise_variable(data, cols = dplyr::everything(), suffix = NULL)

Arguments

Value

A data frame with the standardized columns.

standardize_variable and grouped_df

Note that standardize_variable ignores grouping. Meaning that if you call this function on a grouped data frame (see dplyr::grouped_df), the overall variables' mean and standard deviation will be used for the standardization.

Examples

ho_et_al %>%
  standardize_variable(sdo)

ho_et_al %>%
  standardize_variable(c(sdo, linkedfate), suffix = "scaled")

Tidy eval helpers

Description

• rlang::sym() creates a symbol from a string and rlang::syms() creates a list of symbols from a character vector.

• rlang::expr() and rlang::quo() quote one expression. quo() wraps the quoted expression in a quosure.

  The plural variants rlang::exprs() and rlang::quos() return a list of quoted expressions or quosures.

• rlang::enexpr() and rlang::enquo() capture the expression supplied as argument by the user of the current function (enquo() wraps this expression in a quosure).

  rlang::enexprs() and rlang::enquos() capture multiple expressions supplied as arguments, including ....

exprs() is not exported to avoid conflicts with Biobase::exprs(), therefore one should always use rlang::exprs().

To learn more about tidy eval and how to use these tools, visit https://rlang.r-lib.org/ and the Metaprogramming section of Advanced R.