simfit_practice_all: Simulate Practice-Effect Data and Fit All Models
In dalejbarr/autocorr: Autocorrelation toolkit for multilevel data

Description Usage Arguments Details Value See Also

Simulate data using sim_practice and then fit three models: a non-linear mixed-effects model (NLME), a linear mixed-effects model (LMEM) and a generalized additive mixed-effects model (GAMM).

simfit_practice_all(
  n_subj = 48,
  n_trials = 48,
  within_eff = 10,
  between_eff = 24,
  asymptote = 400,
  sweep = 400,
  learning_rate = -2,
  rslope_sd = 20,
  rasym_sd = 40,
  rsweep_sd = 40,
  rlrate_sd = 0,
  err_sd = 20,
  verbose = FALSE
)

`n_subj`	Number of subjects.
`n_trials`	Number of trials per subject.
`within_eff`	Mean effect of within-subject factor, coded by `xi`.
`between_eff`	Mean effect of between-subject factor, coded by `xj`.
`asymptote`	Asymptotic value.
`sweep`	The 'sweep' of the decline; i.e., the distance from the asymptote to the starting value.
`learning_rate`	Speed of the decay.
`rslope_sd`	By-subject standard deviation for the random slope.
`rasym_sd`	By-subject standard deviation for the random asymptote.
`rsweep_sd`	By-subject standard deviation for the random sweep.
`rlrate_sd`	By-subject standard deviation for the learning rate.
`err_sd`	Error standard deviation.
`verbose`	Whether to return random effects in the data frame.

Used in Monte Carlo simulation. See sim_practice for details on data generation and fit_practice_models for details on model fitting.

A 44-element numeric vector, containing statistical results for all three models. Paramters are prefixed by G for GAMM results, by L for LMEM results, and by N for NLME results. Following the prefix, is the type of statistic, with est for parameter estimates, se for standard errors, t for t-values, and p for p-values. Finally, the suffix identifies the parameter in question, with int for the intercept, wij for the within-subject effect, and bi for the between-subject effect. There are also specific parameters for the nonlinear model, including asym for the asymptote, sweep for the sweep, and lrc for the learning rate. In cases where the non-linear model estimation fails, the error is trapped and the corresponding N.xxx.xxx values will be set to NA.