Help for package semnova

Type:

Package

Title:

Latent Repeated Measures ANOVA

Version:

0.1-6

Author:

Benedikt Langenberg [aut, cre], Axel Mayer [ctb]

Imports:

lavaan, Matrix, parallel, MASS, stats, methods

Suggests:

testthat, knitr, rmarkdown

Depends:

R (≥ 3.4.0)

Description:

Latent repeated measures ANOVA (L-RM-ANOVA) is a structural equation modeling based alternative to traditional repeated measures ANOVA. L-RM-ANOVA extends the latent growth components approach by Mayer et al. (2012) <doi:10.1080/10705511.2012.713242> and introduces latent variables to repeated measures analysis.

Maintainer:

Benedikt Langenberg <benedikt.langenberg@gmail.com>

License:

GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]

Encoding:

UTF-8

LazyData:

true

RoxygenNote:

7.1.0

VignetteBuilder:

knitr

NeedsCompilation:

Packaged:

2020-06-18 13:31:54 UTC; benedicens

Repository:

CRAN

Date/Publication:

2020-06-19 05:50:07 UTC

Comparing the fit of LGC objects.

Description

Comparing the fit of LGC objects.

Usage

## S4 method for signature 'lgc'
anova(object, ...)

Arguments

object

lgc object. An lgc object to be compared against other lgc objects.

...

lgc object. More lgc objects to be compared.

Specifying a measurement model.

Description

Specifying a measurement model.

Usage

create_mmodel(..., list = NULL, lv_scaling = "effect", invariance = NULL)

Arguments

...

Named arguments each representing a latent variable. The arguments are character vectors containing the variable names the latent variables are measured by.

list

List. Each list element represents a latent variable. List elements are character vectors containing the variable names the latent variables are measured by.

lv_scaling

Character vector. Defines the strategy for latent variable scaling. Default is lv_scaling = "effect". Possible strategies are: c("effect", "referent").

invariance

Not yet implemented.

Value

Object of classe mmodel.

Examples


mmodel <- create_mmodel(
    A1B1 = "var1",
    A2B1 = "var2",
    A3B1 = "var3",
    A1B2 = "var4",
    A2B2 = "var5",
    A3B2 = "var6",
    lv_scaling = "referent"
)

General function to specify a general latent growth components model.

Description

General function to specify a general latent growth components model.

Usage

lgc(
  data,
  mmodel,
  C_matrix,
  hypotheses = NULL,
  covariates = NULL,
  groups = NULL,
  append = NULL,
  verbose = FALSE,
  compound_symmetry = FALSE,
  sphericity = FALSE,
  multiv_tests = c("wilks", "wald"),
  univ_tests = NULL,
  randomization = list(ncores = 1, nsamples = 1000),
  ...
)

Arguments

data

Dataframe. Data object to be passed to lavaan.

mmodel

Object of class mmodel. If not provided, manifest variables from the formula object will be used. Otherwise, use create_mmodel() to specify measurement model.

C_matrix

Contrast matrix. Must be invertible.

hypotheses

List of numeric vectors. Each list element represents a hypothesis. For each hypothesis, the contrasts indicated by the elements of the vectors are tested against zero.

covariates

Not implemented yet.

groups

Not implemented yet.

append

Character. Syntax that is to be appended to lavaan syntax.

verbose

Boolean. Print details during procedure.

compound_symmetry

Boolean. When set to TRUE, compound symmetry is assumed.

sphericity

Boolean or formula. When set to TRUE, sphericity is assumed for all effects.

multiv_tests

Character vector. Multivariate test statistics that are to be computed. Possible statistics are: c("wilks", "wald"). Default is multiv_tests = c("wilks", "wald").

univ_tests

Character vector. Univariate test statistics that are to be computed. Possible statistics are: c("F"). Default is univ_tests = NULL.

randomization

Not yet supported.

...

Additional arguments to be passed to lavaan.

Value

Function returns an lgc object. Use summary(object) to print hypotheses. Otherwise use object@sem_obj to get access to the underlying lavaan object.

Examples


set.seed(323412431)

data("semnova_test_data", package = "semnova")

mmodel <- create_mmodel(
    A1B1 = "var1",
    A2B1 = "var2",
    A3B1 = "var3",
    A1B2 = "var4",
    A2B2 = "var5",
    A3B2 = "var6",
    lv_scaling = "referent"
)

hypotheses <- list(
    Intercept = c(1),
    A        = c(2, 3),
    B        = c(4),
    AB       = c(5, 6)
)

C_matrix <- matrix(
    c(1, 1, 0, 1, 1, 0,
      1, 0, 1, 1, 0, 1,
      1,-1,-1, 1,-1,-1,
      1, 1, 0,-1,-1, 0,
      1, 0, 1,-1, 0,-1,
      1,-1,-1,-1, 1, 1),
    nrow=6
)

fit_lgc <- lgc(data = semnova_test_data, mmodel, C_matrix, hypotheses)
summary(fit_lgc)

LGC Class.

Description

LGC Class.

Latent repeated-measures ANOVA using the LGC approach

Description

Function specifies an LGC model. The idata object is used to create the contrast matrix that is passed to the lgc() function. Typical hypotheses are specified as well.

Usage

semnova(
  formula,
  idesign,
  idata,
  data,
  mmodel = NULL,
  covariates = NULL,
  groups = NULL,
  append = NULL,
  icontrasts = c("contr.poly", "contr.sum"),
  verbose = FALSE,
  compound_symmetry = FALSE,
  sphericity = FALSE,
  multiv_tests = c("wilks", "wald"),
  univ_tests = c("F"),
  randomization = list(ncores = 1, nsamples = 1000),
  ...
)

Arguments

formula

Formula.

idesign

Formula. Within-subjects design formula.

idata

Dataframe. The dataframe contains the factorial design.

data

Dataframe. Data object to be passed to lavaan.

mmodel

Object of class mmodel. If not provided, manifest variables from the formula object will be used. Otherwise, use create_mmodel() to specify measurement model.

covariates

Not implemented yet.

groups

Not implemented yet.

append

Character vector. Syntax that is to be appended to lavaan syntax.

icontrasts

Character vector. Use this argument to select the type of contrasts to be used. Default is c("contr.sum", "contr.poly") (not ordered, ordered).

verbose

Boolean. Print details during procedure.

compound_symmetry

Boolean. When set to TRUE, compound symmetry is assumed among dependent variables.

sphericity

Boolean or formula. When set to TRUE, sphericity is assumed for all effects.

multiv_tests

Character vector. Multivariate test statistics that are to be computed. Possible statistics are: c("wilks", "wald"). Default is multiv_tests = c("wilks", "wald").

univ_tests

Character vector. Univariate test statistics that are to be computed. Possible statistics are: c("F"). Default is univ_tests = NULL.

randomization

Not yet supported.

...

Additional arguments to be passed to lavaan.

Value

Function returns an lgc object. Use summary(object) to print hypotheses. Otherwise use object@sem_obj to get access to the underlying lavaan object.

Examples


set.seed(323412431)

data("semnova_test_data", package = "semnova")

idata  <- expand.grid(A = c("A1", "A2", "A3"), B = c("B1", "B2"))

mmodel <- create_mmodel(
    A1B1 = "var1",
    A2B1 = "var2",
    A3B1 = "var3",
    A1B2 = "var4",
    A2B2 = "var5",
    A3B2 = "var6",
    lv_scaling = "referent"
)

fit_semnova <-
    semnova(
        formula = cbind(A1B1, A2B1, A3B1, A1B2, A2B2, A3B2) ~ 1,
        data = semnova_test_data,
        idata = idata,
        idesign = ~ A * B,
        mmodel = mmodel
    )

summary(fit_semnova)

This data set serves for examples and tests.

Description

This is a simulated data set that 100 observation of six normally distributed variables with mean = 0, variance = 1 and covariance 0.5.

Usage

semnova_test_data

Format

A data frame with 100 rows and 6 variables:

Printing the summary for an LGC object.

Description

Printing the summary for an LGC object.

Usage

## S4 method for signature 'lgc'
summary(object, ...)

Arguments

object

lgc object. The object to get a summary about.

...

Additional arguments. Currently none supported.