Help for package ccpsyc

Type:

Package

Title:

Methods for Cross-Cultural Psychology

Version:

0.2.6

Description:

With the development of new cross-cultural methods this package is intended to combine multiple functions automating and simplifying functions providing a unified analysis approach for commonly employed methods.

License:

GPL-3

Depends:

R (≥ 4.0.0)

Imports:

magrittr, dplyr, lavaan, readr, MCMCpack, psych, ufs, xlsx, tibble, rlang, RcppAlgos, tidyr

Suggests:

knitr, rmarkdown, testthat (≥ 3.0.0)

Config/testthat/edition:

Encoding:

UTF-8

LazyData:

true

RoxygenNote:

7.1.1

NeedsCompilation:

Packaged:

2022-05-10 03:45:45 UTC; karljo

Author:

Johannes Karl [aut, cre]

Maintainer:

Johannes Karl <johannes.a.karl@gmail.com>

Repository:

CRAN

Date/Publication:

2022-05-11 05:50:06 UTC

Pipe operator

Description

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

Usage

lhs %>% rhs

Arguments

lhs

A value or the magrittr placeholder.

rhs

A function call using the magrittr semantics.

Value

The result of calling rhs(lhs).

Non-Centrality Index

Description

Non-Centrality Index

Usage

MNCI(object)

Arguments

object

A lavaan object that was fitted with a MLM estimator/

Bootstrapped pairwise differences in psychometric function of groups.

Description

Bootstrapped pairwise differences in psychometric function of groups.

Usage

boot_inv_eff(
  n,
  n_sample,
  df,
  items,
  group,
  eff_sizes = c("SDI2", "UDI2", "WSDI", "WUDI", "dmacs"),
  seed = 2711
)

Arguments

n

Number of bootstraps

n_sample

Number of participants to sample

df

Data to resample

items

Items to resample for the model as vector of strings

group

String variable indicating grouping variable

eff_sizes

Effect sizes to be returned

seed

Seed for replicability

Value

Returns a dataframe with the bootstrapped effect sizes based on the invariance_eff function in this package for two country comparisons.

Examples


two_country <- dplyr::filter(example, country %in% c("NZ" , "BRA"))
boot_inv_eff(n = 10,
             n_sample = 200, df = two_country, group = "country",
              items = paste0("voice",1:3, "m"))

Function to quickly organize and clear psych factor loadings

Description

Function to quickly organize and clear psych factor loadings

Usage

clearing_fa(
  psych_fa,
  cutoff = 0.4,
  dbl_dist = 0.2,
  key_file = NULL,
  cleaned = TRUE
)

Arguments

psych_fa

Output from the psych package, can be either fa or principal with at least two dimensions

cutoff

Desired cutoff below which loadings are omitted defaults to .40

dbl_dist

Desired distance between highest and second highest loading for an item to remove double loadings, defaults to .20

key_file

Optional: Either a .csv or .xlsx file with at least two columns: 1 labeled item containing the item labels as in the data frame, 2 a column labeled wording containing the item wording.

cleaned

If true (default), only the cleaned solution with a message for descriptive stats are returned. If false the function returns a list of data frames one cleaned and one showing all in-between steps

Value

clean This column contains the assignment after removing NAs and double loadings

dir This column contains the direction (positive or negative) of the highest loading.

Examples

library(psych)
fa_solution <- fa(example[c(paste0("help", 1:6, "m"), c(paste0("voice", 1:5, "m")))], nfactors = 2)
clearing_fa(fa_solution)

Computes dMACS

Description

Computes dMACS

Usage

dMACS(fit.cfa, group1, group2)

Arguments

fit.cfa

Lavaan output object with two groups and a single factor.

group1

String for first group in the grouping factor

group2

String for second group in the grouping factor

Value

Returns dMACS for each item.

Examples

dMACS

One-step equivalence testing The function allows for a simple one step test of configural, metric, and scalar equivalence between multiple groups.

Description

One-step equivalence testing The function allows for a simple one step test of configural, metric, and scalar equivalence between multiple groups.

Usage

equival(x, dat, group, standart_lv = TRUE, orthog = TRUE, estim = "MLM")

Arguments

x

CFA model identical to models provided to lavaan.

dat

A data frame or tibble containing the raw data for the specified model.

group

A character string that indicates the column of dat that contains the grouping variable. e.g "country"

standart_lv

A boolean that indicates whether the latent variables should be standardised.

orthog

A boolean that indicates whether the latent variables should be orthogonal.

estim

A string indicating the estimator to be used MLM for complete data and MLR for incomplete data. Defaults to MLM

Value

Returns a data frame with the fit indices for each model and delta values comparing the different levels of equivalence. For a step by step interpretation see.

Examples


model <- "voice =~ voice1m + voice2m + voice3m
          help =~ help1m + help2m + help3m"
equival(x = model, dat = example, group = "country")

Help and Voice Behavior in different countries

Description

Help and Voice Behavior in different countries

Usage

example

Format

A data frame with 5201 rows and 13 variables:

country: Country of sample
help1m: First Help Item
help2m: Second Help Item
help3m: Third Help Item
help4m: Fourth Help Item
help5m: Fifth Help Item
help6m: Sixth Help Item
help7m: Seventh Help Item
voice1m: First Voice Item
voice2m: Second Voice Item
voice3m: Third Voice Item
voice4m: Fourth Voice Item
voice5m: Fifth Voice Item

...

Source

https://www.frontiersin.org/articles/10.3389/fpsyg.2019.01507/full

Improving boot effectsize output

Description

Improving boot effectsize output

Usage

format_boot_inv_eff(x)

Arguments

x

The output of a bootstrapped invariance effect call

Value

A formatted tibble with all effect sizes reported by boot_inv_eff from this package and significant determined by 95% CIs either crossing 0 or .30

Examples


two_country <- dplyr::filter(example, country %in% c("NZ" , "BRA"))
boot_inv_eff_result <- boot_inv_eff(n = 10,
                        n_sample = 200, df = two_country, group = "country",
                        items = paste0("voice",1:3, "m"))
format_boot_inv_eff(boot_inv_eff_result)

Gamma Hat from MLM fitted lavaan object

Description

Gamma Hat from MLM fitted lavaan object

Usage

gamma_hat_scaled(object)

Arguments

object

A lavaan output object that was fitted with a MLM estimator

Invariance Effect Sizes

Description

Invariance Effect Sizes

Usage

invariance_eff(
  df,
  items,
  group,
  nodewidth = 0.01,
  intercept_fix = 1,
  lowerLV = -10,
  upperLV = 10,
  ...
)

Arguments

df

Multi-group dataframe

items

vector of items for the target construct

group

string defining grouping variable

nodewidth

Steps tested

intercept_fix

Which item should have a fixed intercept defaults to the first item

lowerLV

Lower range of latent variable tested

upperLV

Upper range of latent variable tested

...

Passes on to lavaan CFA functions

Value

Returns a dataframe with a row for each item comprising the uni-factorial solution and one column for each invariance effect size. A detailed interpretation of each effect size is provided in Gunn et al. (2019).

Get more comprehensible output from lavTestScore

Description

Get more comprehensible output from lavTestScore

Usage

lavTestScore.clean(lavaan.fit, ndigit = 3, ...)

Arguments

lavaan.fit

Model fitted with lavaan

ndigit

Defines the rounding

...

Arguments passed to lavTestScore

Value

Returns a dataframe which contains one row for each constrained parameter in the model together with a chi-square test indicating whether the parameter significantly differs between groups. This is a cleaned version identical to lavTestScore.

Author(s)

Maksim Rudnev

Multi-group reliability table

Description

Multi-group reliability table

Usage

mg_rel_table(df_s, measure_list, group, digitn = 3, seed = 2711)

Arguments

df_s

The full dataframe with all groups and items.

measure_list

A named list of vectors containing the item names. The format should be list(measure_name1 = c('Item1', 'Item2', 'Item3'), measure_name2 = c('Item1', 'Item2', 'Item3'))

group

Grouping variable in the dataset as string for example "country"

digitn

Controls the amount of digits shown in the output

seed

Seed for the bootstrapped confidence intervals

Value

Returns a formatted dataframe with the reliability of all constructs by group

Pairwise Effect sizes of similarities and difference in the psychometric structure between multiple groups

Description

Pairwise Effect sizes of similarities and difference in the psychometric structure between multiple groups

Usage

multi_group_eff(
  df,
  group,
  items,
  eff_sizes = c("SDI2", "UDI2", "WSDI", "WUDI", "dmacs")
)

Arguments

df

Multi-Group data frame

group

String variable indicating the grouping variable

items

Vector of strings indicating items for the uni-factorial construct

eff_sizes

Effect sizes to be returned

Value

The function returns a list of dataframes with the first reporting the averaged results per item and the second reporting the pairwise comparisons.

Examples


example_s <- dplyr::filter(example, country %in% c("NZ", "BRA"))
multi_group_eff(df = example, group = "country", items = paste0("voice",1:3, "m"))

Creating a Pan-Cultural Loading Matrix

Description

Creating a Pan-Cultural Loading Matrix

Usage

pancultural(df, group, nfactors)

Arguments

df

A data frame contains the variables for the exploratory factor analysis and the grouping variable.

group

The name of the column tht cointains the grouping supplied as a string.

nfactors

The number of factors expected.

Value

returns a Pan-Cultural loading matrix.

Examples

pancultural(example, "country", 5)

Procrustes rotation function, returning Tucker's Phi

Description

Procrustes rotation function, returning Tucker's Phi

Usage

prost(loading, norm, rotated = FALSE, digits = 2)

Arguments

loading

A correlation matrix to be rotated towards a target

norm

A correlation matrix that is the goal of the rotation

rotated

A TRUE/FALSE operator indicating if the rotated matrix should be returned in addition to Tucker's Phi

digits

The number of digits to be displayed in the output, defaults to 2

Value

Returns Tuckers Phi evaluating the congruence of the loading matrix to the normative matrix

Examening chisquare improvement if paths are unconstrained. The function returns the paths to be unconstrained based on chisquare change. Adjusted P-values are calculated based on iterative Bonferroni corrections.

Description

Examening chisquare improvement if paths are unconstrained. The function returns the paths to be unconstrained based on chisquare change. Adjusted P-values are calculated based on iterative Bonferroni corrections.

Usage

release_bonferroni(lavaan.fit, ndigit = 3, exp_p = 0.05, ...)

Arguments

lavaan.fit

Model fitted with lavaan

ndigit

Number of digits to round chi and p to

exp_p

Expected p-value

...

Arguments passed to lavTestScore

Value

Returns a dataframe representing a Bonferroni corrected version of lavTestScore.clean.

Author(s)

Maksim Rudnev

Split by groups

Description

Split by groups

Usage

splitgroup(df, group, named = FALSE, name.list = NA)

Arguments

df

Dataframe

group

Variable from the dataset that defines the groups

named

TRUE/FALSE argument wheter the resulting list should be named

name.list

Supply a list of names same length as number of groups

Value

Returns a list of dataframes with only the selected groups