Skip to content

Troubleshooting and quality checks

Purpose

Use this page to isolate an unexpected result without changing several variables at once.

Before you begin

Record the active experiment, dataset, session, profile, and selected method before changing settings. Those facts explain many apparent discrepancies.

Empty plot or report

  1. Confirm that data are loaded and not excluded.
  2. Check the selected plot level and active child data.
  3. Inspect latency-window names and timing; an out-of-bounds window can produce NaN or no usable result.
  4. Review Diagnostic notices and Help → Open Log Folder.

Windows changed more data than expected

Dataset and Experiment scope copy the draft to every affected active child session. Check Values from and Apply target before Apply. A heterogeneity warning means differing child window sets will be replaced. See Latency windows.

Zero, NaN, or inconsistent amplitudes

  • Confirm that the window contains enough samples at the session sampling rate.
  • Inspect raw and filtered traces separately; filtered values depend on cutoffs/order and unrectified methods depend on polarity.
  • For extrema_ptt, a valid window with no complete eligible pair is 0.0; invalid/out-of-bounds windows are NaN. See Extrema peak-to-trough methods.
  • For aggregate plots, verify bin size and included/excluded sessions.

Implausible relative-to-M-max values

Confirm the M-wave window and amplitude method, then inspect the M-max plot and its detected plateau. If MonStim reports M-max unavailable, the protocol did not provide a qualifying stable high-stimulus response; do not use automatic relative-to-M-max normalization. Review M-max algorithm and Configuration reference.

Import or interface problems

Use Experiment import for missing/misgrouped files. Use the UI scaling guide for a clipped interface or stale saved position. For support, include program version, OS, profile, data level, reproducible steps, and an error report; do not share identifiable data without permission.