fit_2x2: Fit models to 2x2 data with autocorrelated errors
In dalejbarr/autocorr: Autocorrelation toolkit for multilevel data

Description Usage Arguments Details Value

Fit models to 2x2 data with autocorrelated errors

fit_2x2(
  dat,
  cs = FALSE,
  by_subj_fs = TRUE,
  dontfit = FALSE,
  m = NA,
  k = -1,
  bam_args = NULL,
  fit_blocked = TRUE,
  fit_lmem = TRUE
)

`dat`	Data generated using `sim_2x2()`.
`cs`	Fit smooths as main effect of trial (in addition to by-subject factor smooths) for the GAMM models?
`by_subj_fs`	Include by-subject factor smooths in GAMM?
`dontfit`	If `TRUE`, don't fit the models, just return the model formulas. Used for debugging.
`m`	The `m` parameter to be passed on to any factor smooths (specified in the `mgcv::s()` function).
`k`	The `k` parameter to be passed on to any factor smooths (specified in the `mgcv::s()` function).
`bam_args`	Any other arguments to be passed on to `mgcv::bam()` (for the fitting of the GAMM models only.)
`fit_blocked`	Whether to fit the blocked version in addition to the randomized version.
`fit_lmem`	Whether to fit the LMEM in addition to the GAMM.

Fits four models using bam:

1: A Generalized Additive Mixed Model (GAMM) for the blocked DV (Y_b);
2: A Linear Mixed-Effects Model (LMEM) for the blocked DV (Y_b);
3: A GAMM for the randomized DV (Y_r); and
4: A LMEM for the randomized DV (Y_r).

15x2x2 array with model statistics (or just the model formulas if dontfit is TRUE). Second dimension is GAMM or LMEM, third is randomized or blocked, and first dimension is as follows.