Process-state HMM with an IRT response layer — fit_process_hmm

Fits a diagonal-Gaussian HMM to standardized process features within each sequence, then uses state occupancy as explicit process evidence in a response model. This two-stage reference engine is deliberately interpretable and should be distinguished from a fully joint HMM-IRT likelihood.

Usage

fit_process_hmm_irt(
  data,
  sequence_id = "trial_id",
  order = "timestamp",
  process_features = c("x", "y"),
  response = "response",
  person = "participant_id",
  item = "item_id",
  n_states = 3L,
  max_iter = 100L,
  tol = 1e-05,
  seed = 1
)

Arguments

data: Input data frame or compatible tabular object.
sequence_id: Sequence identifier.
order: Within-sequence ordering variable.
process_features: Names of process-derived features.
response: Response variable or response-column name.
person: Person or participant identifier column.
item: Item identifier, name, or item column.
n_states: Number of latent process states.
max_iter: Maximum number of iterations.
tol: Numerical convergence tolerance.
seed: Random-number seed.