umxCLPM: Runs cross-lagged panel models

In tbates/umx: Structural Equation Modeling and Twin Modeling in R

Runs cross-lagged panel models

Description

One way of assessing causal relationships is by introducing time into the analyses. umxCLPM implements three cross-lagged panel models (CLPM) from the literature. The first is the classic CLPM from Heise (1969), the second is the CLPM from Hamaker et al. (2015), and the third is the CLPM from STARTS (1995). You simply pass the number of waves and the data set along with the model you wish to run.

Sketch mode is available; if you pass column names to data, a model object is returned for manipulation later.

Usage

umxCLPM(
  waves,
  name = NULL,
  model = c("Hamaker2015", "Heise1969", "STARTS1995", "IV_RI_CLPM"),
  data = NULL,
  counts = NULL,
  summary = !umx_set_silent(silent = TRUE),
  autoRun = getOption("umx_auto_run"),
  tryHard = c("no", "yes", "ordinal", "search"),
  verbose = FALSE,
  batteries = c("scale", "ordinaloptim"),
  std = FALSE,
  ivs = NULL,
  defn = NULL,
  defto = NULL,
  type = c("Auto", "FIML", "cov", "cor", "WLS", "DWLS", "ULS"),
  allContinuousMethod = c("cumulants", "marginals")
)

Arguments

waves	Number of waves of data.
name	The name of the model (defaults to "Heise1969", "Hamaker2015", "STARTS1995" or "IV_RI_CLPM").
model	Model type ("Hamaker2015", "Heise1969", "STARTS1995", or "IV_RI_CLPM").
data	Data frame for the analysis.
counts	Optional vector of count data columns.
summary	Logical indicating whether to show a summary (default: TRUE if silent is not set).
autoRun	Logical indicating whether to run the model (default to getOption("umx_auto_run")).
tryHard	Method for fitting the model ("no", "yes", "ordinal", "search").
verbose	Logical to control verbose output (default: FALSE).
batteries	A character vector of pre-processing options ("scale", "ordinaloptim", "thresholds").
std	Logical indicating whether to standardize the output (default: FALSE).
ivs	Optional vector of instrumental variable column names.
defn	Optional definition variable.
defto	Optional variable to which to define.
type	The method for handling missing data ("Auto", "FIML", "cov", "cor", "WLS", "DWLS", "ULS").
allContinuousMethod	Method for handling continuous data ("cumulants", "marginals").

Value

An OpenMx model object.

References

Kenny, D.A., & Zautra, A. (1995). The trait-state-error model for multiwave data. Journal of Consulting and Clinical Psychology, 63, 52–59. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1037/0022-006X.63.1.52")} Hamaker E.L., Kuiper R.M., & Grasman R. (2015). A critique of the cross-lagged panel model. Psychological Methods, 20, 102–116. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1037/a0038889")} Heise D. R. (1970). Causal inference from panel data. Sociological Methodology, 2, 3–27. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.2307/270780")}

Examples

## Not run: 

# ================
# = 1. Load Data =
# ================
data(docData)
dt <- docData[2:9]

# ============================
# = 2. Make a CLPM model     =
# ============================
hamaker <- umxCLPM(waves = 4, name = "mymodel", model = "Hamaker2015", data = dt)

## End(Not run)

tbates/umx documentation built on Dec. 14, 2024, 11:28 a.m.