Help: Polygraph Suite v9.7

1. Empirical Scoring System — Multinomial (ESS-M) Methodology

Empirical Scoring System — Multinomial (ESS-M) is an update of the original ESS, published in 2017 by Raymond Nelson. ESS-M applies a Bayesian classifier with a multinomial distribution to analyze Comparison Question Test (CQT) data.

1.1. Theoretical Basis

• Empirical foundation: ESS is a modification of the federal 3-position scoring system, developed based on research at the University of Utah and Johns Hopkins University.
• Bayesian analysis: ESS-M uses a Bayesian decision-making model that combines test data with a prior probability and a likelihood function to compute posterior odds.
• Multinomial distribution: The likelihood function is based on the analytic theory of polygraph testing and describes the probabilities for all possible combinations of sensor scores.

1.2. Scoring Principles

• 3-position scale: +1 (larger reaction to the relevant question), -1 (larger reaction to the comparison question), 0 (no difference). Rule — "bigger-is-better".
• EDA: Extended scale: +2, -2, or 0.
• Primary features: EDA, Cardio, PPG, Pneumograph (Upper + Lower).
• Time constraints: Only timely reactions are scored (no later than 5 sec. after the stimulus).

1.3. Decision Rules

Diagnostic Tests

DI / NDI / INC

Screening Tests

SR / NSR / INC

1.4. Step-by-Step Instructions: ESS-M

1. Enter the examinee's full name and date in the top panel.
2. Go to the ESS-M tab → click "+ Add ESS-M Calculation".
3. Select the Test Format (DLST, DLDT, LEPET, AFMGQT, BOST, Federal, Utah).
4. If needed, change the type: Screening / Diagnostic.
5. Click Questions and enter the texts of relevant questions R1–R4.
6. Enter scores: Breathing Thoracic/Abdominal (–1, 0, +1), EDA (–2, 0, +2), Cardio/PPG (–1, 0, +1). Symbol A or ∅ marks an artifact. Press – for quick negative value entry.
7. The system automatically calculates subtotals, Grand Total, status, and the probability estimate (Pr).
8. Review the "State Dynamics" block to assess the trend between presentations.

2. Concealed Information Test (CIT) Methodology

CIT (Guilty Knowledge Test) was proposed by David Lykken in 1959. It is the only psychophysiological method properly substantiated by scientific research and the orienting response (OR) theory.

2.1. Theoretical Basis

• Stimulus structure: A series of multiple-choice questions, each containing one relevant alternative (Key) and several neutral alternatives (Foils).
• Orienting response: Significant stimuli evoke enhanced OR. Systematically stronger reaction to the Key → indicates awareness.
• Probabilistic protection: The probability of a random positive result is mathematically controlled.

2.2. Lykken Scoring System

• 2 points — largest reaction to the Key.
• 1 point — second-largest reaction to the Key.
• 0 points — Key is not among the two largest reactions.

3. "Notes" Tab NEW v8.1

The " Notes" tab is the examiner's workspace for recording unstructured information: observations during testing, behavioral markers, conclusions, and examination context. Notes are tied to a specific examination (examinee), not to an individual test.

3.1. Markdown Editor

The editor supports two modes, switched via buttons at the top of the tab:

Supported Markdown Syntax

3.2. Attaching Images

Below the editor is the "Attached Images" section. Images are stored in the browser's IndexedDB — a separate storage that is not limited by the localStorage quota.

• Adding: "Add Image" button or drag files into the Drag & Drop zone.
• Formats: JPG, PNG, WEBP, GIF, BMP — any image.
• Limit: Up to 20 images of 5 MB each.
• Viewing: Click on a thumbnail → full-screen lightbox.
• Deleting: "×" button on the thumbnail on hover.

3.3. Where Notes Are Stored

Note text is stored as a string in localStorage in the notes field along with all other session data. Images are stored in IndexedDB, while localStorage only holds their metadata (name, size, date).

4. "Supervision" Tab NEW v8.5

The "Supervision" tab is designed for a second specialist (a supervisor) who reviews ESS-M tests already completed by the examiner. The tab is positioned separately on the right and works with the same session .json file — the examiner fills in tests as usual and saves the file, and the supervisor opens it on their own device to perform the review.

4.1. How it works

1. The examiner completes one or more ESS-M tests and saves the session (.json or .zip).
2. The supervisor opens the same file on their device via the "Open" button.
3. The "Supervision" tab's dropdown shows only the tests the examiner has actually filled in.
4. For the selected test, a table appears with the same structure as ESS-M (charts, R1–R4, channels), but each cell contains three elements: the supervisor's input field, the examiner's non-editable value, and a discrepancy marker.
5. The supervisor enters their own score using the same scoring format as ESS-M (–1/0/+1 for most channels, –2/0/+2 for EDA, symbols A/∅ for artifacts).

4.2. Discrepancy markers

4.3. Summary panel

Above the table, three aggregated metrics are shown for the selected test: how many cells the supervisor has filled in, the percentage match with the examiner's scores, and the total discrepancy. Metrics update immediately as you type, with no need to reload the page.

4.4. Where supervision data is stored

Supervisor scores are stored separately for each test in the session file's sv field, alongside ess, cit, and notes. Switching between tests in the dropdown does not lose entered data.

5. "R&I" Tab NEW v8.12

The Relevant/Irrelevant (RI) technique is the oldest polygraph method, alternating relevant (case-related) and irrelevant/neutral questions. The primary method of decision-making in RI is global qualitative evaluation by the examiner across the entire chart — no officially published formula exists for automatically determining SR/NSR.

5.1. The FROSS Method

This tab implements FROSS (Field Rank Order Scoring System), also known as the "High Three" — a documented supplementary rank-analysis method used to objectively support the conclusion and identify which relevant questions warrant further breakdown testing.

5.2. Ranking Procedure

The number of relevant questions is fixed at 5, and the number of presentations is fixed at 4. For each of the three physiological channels (P — respiration, E — electrodermal activity, C — cardiovascular activity), up to three most pronounced reactions are identified separately within each presentation:

• The most pronounced reaction is assigned rank 3, the second most pronounced — 2, the third — 1.
• If two reactions show equal intensity, both receive the same rank.
• If only two reactivities are observed within a channel, they receive ranks 3 and 2 (no 1).
• If no noticeable difference in reactivity is observed, the channel is not scored (the cell remains blank).

Ranks across channels are summed for each relevant question (SUBTOTALS), and across multiple presentations they are added into a "Grand Rank Total." On the significance chart, exactly three questions with the highest grand total are always highlighted in color (red — highest, orange — second, yellow — third), the rest in gray — even when several questions share the same score. Questions with a zero rank always stay gray.

The "Questions" button opens a form to record the wording of relevant questions R1–R5 (for reference only, it does not affect scoring); the saved text appears as a tooltip on the column header and on the chart bar. The "Clear Data" button clears only the ranks entered in the table — the test title and question texts are preserved. Presentation 4 can be temporarily hidden with a checkbox in its header (same pattern as the chart toggles in ESS-M) — hidden data is not deleted and is counted again once shown; while a presentation is hidden it is excluded from the "Grand Rank Total."

5.3. Sources

• Krapohl D., Shaw P. Fundamentals of Polygraph Practice. 2015. P. 122.
• Baur D. Federal Psychophysiological Detection of Deception Examiner Handbook, Counterintelligence Field Activity Technical Manual. 2006. P. 58-59.
• Krapohl D., Senter S., Stern B. An exploration of methods for analysis of multiple-issue relevant/irrelevant screening data. Polygraph. 2005. 34(1). P. 47-61.
• Shapovalov V.O., Aleksieieva-Protsiuk D.O., Zubovskyi D.S., Aleksieiev O.O. Technology of Applying RI, POT, and CIT Methods in Polygraph Examinations. Ukrainian Polygraph Association, 2020. P. 34-40.

7. Interface Overview

7.1. Global Toolbar

• Save — forces a save to localStorage.
• Open — imports .json or .zip from disk or Drag & Drop.
• Save — exports .json (without images) or .zip (with images).
• Export — generates a Markdown report, including notes.
• Print / PDF — optimized printing (hidden controls, preserved colors).
• Clear All — deletes localStorage, IndexedDB (images), and reloads the page.
• Help — this page.

7.2. Tabs

8. Methodological Sources and References

• Nelson, R. (2017). Multinomial reference distributions for the Empirical Scoring System. Polygraph and Forensic Credibility Assessment, 46(2), 81–115.
• Nelson, R. (2016). Scientific (analytic) theory of polygraph testing. APA Magazine, 49(5), 69–82.
• Krapohl, D. J. (2010). Reference Guide for the Empirical Scoring System — Multinomial (ESS-M). Lafayette Instrument Company.
• Lykken, D. T. (1959). The GSR in the detection of guilt. Journal of Applied Psychology, 43(6), 385–388.
• Lykken, D. T. (1960). The validity of the guilty knowledge technique. Journal of Applied Psychology, 44(4), 258–262.
• Verschuere, B., Ben-Shakhar, G., & Meijer, E. (2011). Memory Detection: Theory and Application of the Concealed Information Test. Cambridge University Press.
• Meijer, E. H., et al. (2016). Deception detection with behavioral, autonomic, and neural measures. Psychophysiology, 53(5), 592–599.