**Emotional Stop-Signal Task (ESST) User Manual v2024** Date of most recent update: 02.13.2024 **1. INTRODUCTION** The Emotional Stop-Signal Task (ESST; Allen & Hooley, 2015, 2019; Allen, Johnson, Burke et al., 2021) is a behavioral assessment tool designed to measure *emotional response inhibition*, a core component of *affective control* or (low-level) "hot" executive functioning. It is particularly useful in exploring the relationship between response inhibition, emotional regulation, and other transdiagnostic constructs associated with psychopathology. Specifically, evidence suggests that emotional response inhibition, as measured by the ESST, may comprise a neurocognitive mechanism underlying emotion-related impulsivity (or *urgency*) in diverse psychiatric syndromes. The ESST has strong psychometric properties comparable to gold-standard neuropsychological tests used in clinical settings (Allen et al., 2021). **2. EXPERIMENTAL DESIGN** The ESST is a computerized go/no-go task that follows parallel procedures to the standard stop-signal task (Logan & Cowan, 1994; Verbruggen & Logan, 2008; Verbruggen et al., 2019), an assessment of late-stage prepotent motor response inhibition or action cancellation. In typical stop-signal tasks, participants are asked to perform a primary response task, such as classifying stimuli, and to inhibit cued responses when a stop-signal (often an auditory cue) is presented. In the most widely used variant, participants must quickly and accurately indicate (via keypress) the direction of serially-presented arrow stimuli, except on an unpredictable subset of trials that include a stop-signal, which requires *post-hoc* termination of the initiated or ongoing motor response. The stop-signal delay (SSD) from stimulus onset varies, testing the participant's ability to inhibit responses under different levels of difficulty and permitting calculation of stop-signal reaction time (SSRT), a continuous index of motor response inhibition speed. In the ESST, the primary task involves responding to the valence of emotional stimuli, usually images with positive, negative, or neutral emotional content adapted from the International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert, 2008). The stop-signal in ESST adds an emotional regulation component to the standard response inhibition task. On most ESST (go/no-signal) trials, participants are prompted to quickly and accurately categorize these image stimuli as either positive (i.e., "pleasant"; default key: 'a') or negative (i.e., "unpleasant"; default key: 'l') by keypress, based on their automatic, reflexive, or "gut" reaction the the affective content of each image. However, at random intervals, an auditory "stop-signal" (buzzer or tone) is presented with the target stimulus either concurrently or after a brief delay, which cues participants to inhibit their emotional reaction and accompanying motor response. These no-go/stop trials only occur together with (or closely following) Negative or Positive target stimuli; Ambiguous/neutral images are shown exclusively during go/no-signal trials. On no-go/stop-signal trials, the latency or SSD between the target image and stop-signal presentation is adjusted stepwise depending on participant performance, using separate staircase algorithms for Negative and Positive stimuli, in order to achieve a ~50% "false alarm" rate for commission errors in both emotional valence categories. This fixed error rate enables estimation of independent SSRT values – the ESST’s primary metric of emotional response inhibition – associated with negative and positive affective reactions. In sum, the ESST is designed to evaluate emotional response inhibition, providing insight into how emotions influence impulse control and decision-making, especially under stress or emotional arousal. **2.1 Structure & timing parameters:** Following a brief practice block of 48 trials (*n* = 12 no-go/stop-signal trials), the ESST includes four experimental blocks of 120 trials each (*N* = 480 total trials), with 75% containing no stop-signal (go/no-signal trials; *n* = 360; 90 per block) and the remaining 25% comprising no-go/stop-signal trials that require emotional response inhibition (*n* = 120 ; 30 per block). The total duration of the ESST is approximately 15-20 minutes. Each ESST trial spans 2000 ms (i.e., stimulus onset asynchrony or SOA), which includes a fixed "pre-trial pause" period of 500 ms, followed by a central fixation cross that is displayed for ~250 ms, depending on a jittered latency value assigned pseudo-randomly on a trial-by-trial basis (i.e., 150-350 ms) to facilitate integration with event-based psychophysiology data collection. The target image stimulus then replaces the fixation cross in the center of the screen, where it is displayed for 1250 ms, i.e., 1400-1600 ms post-fixation onset. The initial (50 ms) stop-signal is presented 250 ms after stimulus onset (i.e, 400-600 ms post-fixation onset) -- separately for positive and negative images -- and subsequently adjusted upwards (if inhibition is successful) or downwards (if a commission error occurred) in 50 ms increments, to capture a broad range of SSD values (i.e., 0-1200 ms) that may include no-go/stop-signal trials requiring early-stage emotion-triggered action suppression or withholding rather than late-stage cancellation or termination, i.e., SSD = 0. **2.2. Conditions & trial types:** ESST trial types are pseudo-randomized and counterbalanced across blocks to ensure the same number of trials in each condition are presented in each block. (a) Ambiguous go/no-signal trials (*n* = 120; *n* = 30 trials per block) (b) Negative go/no-signal trials (*n* = 120; *n* = 30 trials per block) (c) Positive go/no-signal trials (*n* = 120; *n* = 30 trials per block) (d) Negative no-go/stop-signal trials (*n* = 60; *n* = 15 trials per block) (e) Positive no-go/stop-signal trials (*n* = 60; *n* = 15 trials per block) **2.3. ESST-Adaptive (ESST-A):** The ESST-A is an adaptive variant of the task designed as a key component of "Neurobehavioral Affective Control Training (N-ACT)", a novel cognitive remediation intervention targeting emotion dysregulation, which is currently being evaluated in a randomized waitlist-controlled trial. It is largely identical to the original version, with some additional design features intended to facilitate enhancement of emotional response inhibition over repeated administrations. Specifically, only emotional (negative or positive) images are shown on no-go/stop-signal trials, and conversely, only ambiguous images are shown on go/no-signal trials without an inhibitory cue. The ESST-A accordingly requires participants to inhibit emotional reactions to only the most arousing/evocative stimuli, in order to promote implicit learning that salient affective content (and/or strong emotional reactions to this information) serves as a reliable “cue” that predicts upcoming inhibitory demand (i.e., stop-signals). Repeated training with the ESST-A is therefore meant to eventually facilitate more automatic engagement of inhibitory control at early stages of emotional information processing. The other adaptive component of the ESST-A is equivalent to the assessment version of the ESST (and similar to other stop-signal tasks). Specifically, the ESST-A includes two independent staircase algorithms that adjust valence-specific SSDs (i.e., distinct tracking values for negative and positive images) in 50 ms increments depending on participant performance, such that successful inhibition results in a longer SSD on the next no-go/stop trial with the same type of image, thereby increasing inhibitory demand. In contrast, commission errors (i.e., false alarms) shorten SSD by 50 ms on subsequent no-go/stop trials with images from the same emotional valence category. These two stepwise algorithms attempt to maintain commission error rates at approximately 50% of no-go/stop-signal trials, separately for negative and positive images, which ultimately enables calculation of valence-specific stimates of stop-signal reaction time. **3. PARTICIPANT DIRECTIONS** Participants are instructed to react to the images as quickly and accurately as possible. They must also heed the auditory stop-signals, which indicate when they should not respond. The emphasis is on balancing speed with accuracy in response (as in traditional stop-signal tasks). Please see below for specific instructions shown in the current (PsychoPy) version of the ESST: "This is a timed test. On each trial, you will see an image. Your task is to decide whether each image is POSITIVE or NEGATIVE: (1) Press the 'A' key (marked with a smiling sticker) if the image is pleasant or POSITIVE. (2) Press the 'L' key (marked with a frowning sticker) if the image is unpleasant or NEGATIVE. If you are unsure how a picture makes you feel, follow your instinct or 'gut reaction'. Please respond as quickly and accurately as possible! However, if you hear a sound (called the 'STOP SIGNAL'), you should STOP your response on that trial. DO NOT respond to the image when you hear this sound. You should stop your response regardless of whether the image has just appeared or if the STOP SIGNAL plays afterwards. Nevertheless, DO NOT WAIT for the STOP SIGNAL. If you wait, the test will get harder on later trials. Press <SPACEBAR> to practice." **3.1 Performance feedback:** During the practice block, the ESST provides trial-level feedback (for 1000 ms per trial) to familiarize participants with its parameters. Specifically, participants will see the word "Correct!" following accurate classification responses to (positive and negative) emotional images on go/no-signal trials as well as after a response of either valence to ambiguous stimuli (default keys: 'a' or 'l'). Inaccurate responses to emotional stimuli on go/no-signal trials are followed by text indicating either a commission ("Oops! That was the wrong response.") or omission error occurred ("Miss! Please try to respond more quickly."). Finally, participants receive separate notifications for successful inhibition ("Correct! No response needed on stop trials.") and commission errors on no-go/stop-signal trials ("False alarm! Please do not respond on stop trials.") At the end of each block (including Practice), participants are shown a feedback summary of their performance on the trials within the preceding block. Specifically, the ESST will display the following feedback information alongside a 10-second timer countdown before proceeding to the next block: (1) Mean accuracy (%) on go/no-signal trials; (2) Mean reaction time on go/no-signal trials; and (3) Mean accuracy (%) on no-go/stop-signal trials, which should approximate 50%. Participants should also achieve a "miss" rate under 25% as well as a "hit" rate of at least 75% accuracy on go/no-signal trials during the Practice block to proceed with the experimental blocks; participants who do not reach these thresholds will receive a reminder prompt ("Remember to respond as quickly *and* accurately as possible to each image, and **DO NOT WAIT for the stop-signal**. If you wait, the test will get harder on later trials.") and should be asked to re-do the practice trials. Once the countdown timer reaches "0", the feedback screen disappears and the next block begins (with pre-block mood ratings; see below). 3.2. **Stimuli & mood ratings:** The IAPS and related emotional stimuli have been used in at least 70 published studies, including many focused on mental health (e.g., see Branco, Gonçalves, & Badia, 2023). It is fair to state that use of this stimulus battery is a standard research approach in the study of emotion. Although the photos are designed to provoke responses, it is important to consider that most people are exposed to disturbing (and pleasing) imagery across a variety of digital and traditional media. It is perhaps not surprising, then, that IAPS images are not found to provoke adverse events, even for people who experience mental health symptoms. Nonetheless, it is important to respect participants' autonomy and ability to make informed decisions. Thus, it is recommended for researchers to highlight the nature of these photos during the informed consent process, and consider reminding participants again prior to launching the ESST. Participants are presented with a Visual Analogue Scale (VAS) to rate their current mood (from 'extremely negative' to 'extremely positive') using a slider (default values: 0-100) before each ESST experimental block (i.e., starting after the Practice block). Each pre-block VAS is shown with the following text instructions (replacing the block feedback screen after the 10-second waiting period): "Before you continue with the test, please rate your current mood below by using the slider. Use your mouse to click and drag the red bar to how positive or negative you feel RIGHT NOW. You can then press <SPACEBAR> to proceed after you rate your mood." Once participants navigate away from the VAS screen, they are shown a final pre-block text message (or the exit screen): "There are four blocks of trials in this test. Press <SPACEBAR> when you are ready to begin "+ block_title+" *(*Practice, Block 1, etc.)*."* The ESST exit screen displays the following text after participants complete a fifth and final VAS rating following completion of the fourth block of experimental trials: "Well done! You've finished this test. Thank you for participating in our study. Please press the <SPACEBAR> to end the experiment and inform a research team member." **4. DATA COLLECTION & ANALYSIS** In setting up the ESST, researchers need to ensure a controlled environment to minimize external distractions, ideally providing noise-canceling earbuds or headphones. The software used should be capable of randomly presenting emotional stimuli and auditory stop-signals, while accurately recording response times and inhibition successes or failures. The current (v2024) ESST version is programmed in PsychoPy (.psyexp) freeware, but legacy variants in Millisecond software's Inquisit (.iqx) are available upon request (contact: jd.allen@berkeley.edu or jd.allen@brainhealthconsulting.org). The ESST begins with a calibration phase or practice block, which includes trial-level feedback absent from subsequent experimental or test blocks. During the task, the software records each response and the adherence to stop-signals. Researchers or clinicians should monitor the participant for consistent engagement and compliance with instructions. Key data include valence-specific accuracy and reaction time to images on go/no-signal trials and the accuracy of stopping responses when signaled. This data is then analyzed to calculate the stop-signal reaction time (SSRT), which provides insights into the participant's ability to inhibit responses under emotional contexts. **4.1 ESST summary indices:** - Emotional response accuracy (%): Proportion of all go/no-signal trials with correctly-identified negative and positive stimuli - Negativity bias (%): Proportion of all go/no-signal trials classified as 'negative' (regardless of stimulus valence) - Miss rate (%): Proportion of all go/no-signal trials with omission errors - Go/no-signal reaction time (ms): Mean response time across all go/no-signal trials with valid responses - Negative stop-signal reaction time (SSRT) (ms): Estimated speed of inhibitory processing of negative emotional content (i.e., negative emotional response inhibition), calculated as mean reaction time on go/no-signal trials with negative images minus the median stop-signal delay (SSD) on no-go/stop-signal trials with negative images - Positive stop-signal reaction time (SSRT) (ms): Estimated speed of inhibitory processing of negative emotional content (i.e., positive emotional response inhibition), calculated as mean reaction time on go/no-signal trials with positive images minus the median stop-signal delay (SSD) on no-go/stop-signal trials with positive images - *P*(Negative false alarm) (%): Proportion of no-go/stop-signal trials with negative images and responses (i.e., commission errors), reflecting the likelihood or probability of making a "false alarm" representing a negative emotional reaction - *P*(Positive false alarm) (%): Proportion of no-go/stop-signal trials with positive images and responses (i.e., commission errors), reflecting the likelihood or probability of making a "false alarm" representing a positive emotional reaction **5. TROUBLESHOOTING & FAQ** **6. REFERENCES**