Skip to main content
v2026.11,986 entries · CC-BY 4.0

Attentional Blink: Measurement, Formula and Guide

Attentional blink is a temporal attention deficit where identifying a first target impairs reporting a second target appearing within 200 to 500 milliseconds.

Written and maintained by CASRAI Editorial Board

Published

On this page

The attentional blink is a temporal visual phenomenon in which identifying a first target temporarily impairs reporting a second target appearing within 200 to 500 milliseconds, reflecting capacity constraints in working memory consolidation.

Definition in One Paragraph

The attentional blink (AB) is an experimental paradigm and quantitative measure of temporal visual attention defined as the transient impairment in reporting a second target stimulus (T2) when presented between 180 and 500 milliseconds following a successfully identified initial target (T1) within a single-location Rapid Serial Visual Presentation stream. Occurring under high perceptual load at display rates of 6 to 20 items per second (typically 10 items per second, or 100 milliseconds per item), the attentional blink stems from an active central bottleneck: while preconscious semantic processing of T2 continues uninterrupted, working memory encoding of T2 is suspended while capacity-constrained Stage 2 neural consolidation mechanisms process T1.

Formula with Every Symbol Explained

Quantitative analysis of the attentional blink relies on three foundational equations defining target timing, conditional reporting accuracy, and deficit magnitude.

First, temporal spacing between targets is determined by Stimulus-Onset Asynchrony (SOA):

$$text{SOA} = text{Lag} times Delta t$$

Where:

  • Lag: Position of T2 relative to T1 (Lag 1 = adjacent, Lag 2 = one distractor intervening).
  • Δt: Display duration per item in milliseconds (typically 80 to 100 ms).
  • SOA: Total elapsed time in milliseconds from T1 physical onset to T2 physical onset.

Second, performance is measured as Conditional Target 2 Accuracy ($P(T2|T1)$) to isolate central resource allocation from primary perceptual identification failures on T1:

$$P(T2|T1) = frac{N(T1_{text{correct}} cap T2_{text{correct}})}{N(T1_{text{correct}})}$$

Where:

  • N(T1correct): Count of trials where T1 was correctly reported.
  • N(T1correct ∩ T2correct): Count of trials where both T1 and T2 were correctly reported.
  • P(T2|T1): Conditional accuracy proportion, serving as the primary dependent variable in RSVP attention experiments.

Third, researchers quantify deficit depth via Attentional Blink Magnitude (ABM):

$$text{ABM} = frac{P(T2|T1)_{text{baseline}} – P(T2|T1)_{text{trough}}}{P(T2|T1)_{text{baseline}}}$$

Where:

  • P(T2|T1)baseline: Asymptotic conditional accuracy at late recovery lags (typically Lag 7 or 8, SOA ≈ 700–800 ms).
  • P(T2|T1)trough: Minimum conditional accuracy at the blink trough (typically Lag 2 or 3, SOA ≈ 200–300 ms).
  • ABM: Proportional index of attentional suppression from 0.0 (no deficit) to 1.0 (complete suppression).

Worked Example with Real Numbers and Arithmetic Shown

Consider an adult participant completing a dual-target RSVP task at 10 items per second ($Delta t = 100text{ ms}$). Target 1 is a white letter among black distractors; Target 2 is a black letter ‘X’. The task contains 500 total trials evenly divided across 5 test lags (100 trials per lag).

Lag Index SOA (ms) Total Trials T1 Correct Both Correct Conditional Accuracy P(T2|T1)
Lag 1 100 ms 100 90 74 74 / 90 = 0.8222 (82.2%)
Lag 2 200 ms 100 88 37 37 / 88 = 0.4205 (42.0%)
Lag 3 300 ms 100 89 31 31 / 89 = 0.3483 (34.8%)
Lag 4 400 ms 100 91 56 56 / 91 = 0.6154 (61.5%)
Lag 7 700 ms 100 92 81 81 / 92 = 0.8804 (88.0%)

Step 1: Calculate Conditional Accuracy for Each Lag

Conditional accuracy at Lag 1 remains high at 82.2% (74/90, classic Lag-1 sparing). Accuracy drops to 42.0% (37/88) at Lag 2 and reaches its minimum trough of 34.8% (31/89) at Lag 3 ($P(T2|T1)_{text{trough}}$). Performance recovers at Lag 4 (56/91 = 61.5%) and reaches baseline asymptote at Lag 7 (81/92 = 88.0%, $P(T2|T1)_{text{baseline}}$).

Step 2: Calculate Attentional Blink Magnitude (ABM)

Using baseline accuracy from Lag 7 and trough accuracy from Lag 3:

$$text{ABM} = frac{P(T2|T1)_{text{Lag 7}} – P(T2|T1)_{text{Lag 3}}}{P(T2|T1)_{text{Lag 7}}} = frac{0.8804 – 0.3483}{0.8804} = frac{0.5321}{0.8804} approx 0.6044$$

The participant exhibits an Attentional Blink Magnitude of 0.6044 (60.4%), demonstrating that T1 consolidation suppresses Target 2 detection by over 60% relative to baseline recovery.

Interpretation: What the Result Means

Interpreting attentional blink trajectories involves evaluating performance against established neurocognitive benchmarks:

Lag (SOA) Observed Pattern Typical P(T2|T1) Cognitive State and Mechanism
Lag 1 (100 ms) Lag-1 Sparing 75% to 90% Both targets fall within a single attentional window, entering Stage 2 consolidation together.
Lag 2 (200 ms) Blink Onset 35% to 55% Attentional gate closes to protect T1; T2 features register but P3b waves are suppressed.
Lag 3 (300 ms) Maximum Trough 25% to 45% Peak working memory bottleneck; vulnerable T2 representation decays before conscious consolidation.
Lag 4–5 (400–500 ms) Recovery Phase 55% to 75% T1 consolidation finishes; central resources reallocate, allowing P3b amplitudes to rebound.
Lag 7+ (700+ ms) Baseline Asymptote 80% to 95% Full recovery; Target 2 identification matches single-target control baselines.

Standard Attentional Blink Magnitude benchmarks:

  • ABM < 0.25 (Attenuated): Minimal interference (action gamers, meditators, or unmasked tasks).
  • ABM 0.25–0.50 (Moderate): Lower perceptual load or highly distinct targets.
  • ABM 0.50–0.75 (Robust): Normative healthy adult baseline in dual-target RSVP tasks.
  • ABM > 0.75 (Severe): Pronounced suppression from elevated T1 difficulty or clinical deficits.

Assumptions and Common Mistakes

Valid experimental execution requires controlling for four common methodological pitfalls:

  1. Using Unconditional Accuracy: Reporting raw Target 2 accuracy ($N(T2_{text{correct}}) / N_{text{total}}$) without conditioning on T1 accuracy conflates sensory misses on T1 with authentic dual-target attentional competition. Only trials with correct T1 identification isolate central bottlenecks.
  2. Omitting Trailing Masks: Raymond et al. (1992) established that backward masking is essential. Omitting distractor masks preserves items in sensory iconic memory, eliminating the blink effect.
  3. Misclassifying Order Inversions: During Lag-1 sparing, participants often perceive both targets but swap their reported order. In strict position scoring, valid detections are erroneously scored as errors. Scoring criteria must specify whether order-free scoring applies.
  4. Overlooking Floor and Ceiling Effects: Excessive T1 difficulty induces floor effects that truncate conditional trials, while salient T2 stimuli induce ceiling effects. For diagnostic strategies, see our guide on floor and ceiling effects, and review construct boundaries in convergent and discriminant validity.

How to Report It (APA Example)

When reporting attentional blink findings under APA 7th edition conventions, report RSVP timing, conditional scoring criteria, repeated-measures ANOVA model parameters, sphericity tests, and effect sizes. Review verification steps in our guide on repeated-measures ANOVA assumptions and effect size in research.

“A one-way repeated-measures ANOVA evaluated the effect of stimulus-onset asynchrony across five lags (Lag 1: 100 ms, Lag 2: 200 ms, Lag 3: 300 ms, Lag 4: 400 ms, Lag 7: 700 ms) on conditional Target 2 accuracy (T2|T1). Mauchly’s test indicated sphericity was violated, χ²(9) = 24.18, p = .004; degrees of freedom were adjusted using Greenhouse-Geisser estimates (ε = .74). The main effect of Lag was significant, F(2.96, 71.04) = 43.18, p < .001, ηp² = .64. Pairwise comparisons with Bonferroni correction confirmed Lag-1 sparing: accuracy at Lag 1 (M = 82.5%, SD = 7.4%) significantly exceeded the trough at Lag 3 (M = 34.8%, SD = 8.1%), p < .001. Performance recovered at Lag 7 (M = 88.2%, SD = 6.5%), which did not differ significantly from single-target baseline control trials (M = 89.4%, SD = 5.9%), p = .38. Mean Attentional Blink Magnitude across participants was 60.5% (SD = 9.2%).”

How to Calculate It in Excel, R, or Python

Below are code snippets to calculate conditional accuracy per lag and compute attentional blink magnitude.

1. Excel / Google Sheets

With trial data in columns: Column B = Lag (1, 2, 3, 4, 7), Column C = T1_Correct (1/0), Column D = T2_Correct (1/0) across rows 2 to 501:

  • T1 Correct count at Lag 2: =COUNTIFS($B$2:$B$501, 2, $C$2:$C$501, 1)
  • Both Correct count at Lag 2: =COUNTIFS($B$2:$B$501, 2, $C$2:$C$501, 1, $D$2:$D$501, 1)
  • Conditional Accuracy P(T2|T1) at Lag 2: =COUNTIFS($B$2:$B$501, 2, $C$2:$C$501, 1, $D$2:$D$501, 1) / COUNTIFS($B$2:$B$501, 2, $C$2:$C$501, 1)
  • Attentional Blink Magnitude: With Lag 3 in H3 and Lag 7 in H5: =(H5 - H3) / H5

2. R (tidyverse and rstatix)

library(dplyr)
library(tidyr)
library(rstatix)

# Conditional accuracy per participant and lag
ab_summary <- trial_data %>%
  filter(t1_correct == 1) %>%
  group_by(participant_id, lag) %>%
  summarize(p_t2_given_t1 = sum(t2_correct == 1) / n(), .groups = "drop")

# Attentional Blink Magnitude
ab_magnitude <- ab_summary %>%
  pivot_wider(names_from = lag, values_from = p_t2_given_t1, names_prefix = "lag_") %>%
  mutate(abm = (lag_7 - lag_3) / lag_7)

# Repeated-Measures ANOVA
anova_out <- anova_test(data = ab_summary, dv = p_t2_given_t1, wid = participant_id, within = lag)
print(get_anova_table(anova_out))

3. Python (pandas and statsmodels)

import pandas as pd
from statsmodels.stats.anova import AnovaRM

# Conditional accuracy per participant and lag
valid = df[df["t1_correct"] == 1]
cond_acc = (
    valid.groupby(["participant_id", "lag"])["t2_correct"]
    .mean()
    .reset_index(name="p_t2_given_t1")
)

# AB Magnitude
pivot_df = cond_acc.pivot(index="participant_id", columns="lag", values="p_t2_given_t1")
cond_acc["ab_magnitude"] = (pivot_df[7] - pivot_df[3]) / pivot_df[7]

# Repeated-Measures ANOVA
model = AnovaRM(cond_acc, depvar="p_t2_given_t1", subject="participant_id", within=["lag"])
print(model.fit().summary())

Frequently Asked Questions (FAQ)

The attentional blink results from a capacity bottleneck during working memory consolidation (Chun & Potter, 1995). While Stage 1 perceptual processing operates in parallel, Stage 2 conscious consolidation is capacity-constrained. While Stage 2 processes T1, T2 remains vulnerable in Stage 1 and degrades from backward masking.

Lag-1 sparing is the preservation of T2 accuracy when T2 immediately succeeds T1 (Lag 1, ~100 ms SOA). Sparing occurs because both targets fall within the initial temporal window of attentional selection before the gate closes, entering Stage 2 consolidation together.

Inattentional blindness occurs when an unexpected stimulus is missed because spatial attention is absorbed elsewhere. In contrast, the attentional blink is a purely temporal deficit at a single spatial location where both targets are explicitly expected, but T2 arrives during consolidation of T1.

Why is conditional accuracy (T2|T1) required instead of raw T2 accuracy?

Conditional accuracy is required because the attentional blink measures processing interference from T1 onto T2. If a participant misses T1, no consolidation bottleneck was triggered. Raw T2 accuracy conflates low-level sensory misses with true attentional competition.

Attentional blink magnitude is plastic and can be attenuated. Studies show that mindfulness meditation reduces redundant resource allocation to T1, accelerating capacity recovery. Action video game players also exhibit reduced blink depth.

Sources

The definitions, experimental paradigms, mathematical formulations, and empirical benchmarks on this page are derived from the following peer-reviewed literature and reference standards:

  • Chun, M. M., and Potter, M. C. (1995). A two-stage model for multiple target detection in rapid serial visual presentation. Journal of Experimental Psychology: Human Perception and Performance, 21(1), 109–127. https://doi.org/10.1037/0096-1523.21.1.109 (Checked 2026-11-09).
  • Dux, P. E., and Marois, R. (2009). The attentional blink: A review of data and theory. Attention, Perception, & Psychophysics, 71(8), 1683–1700. PMC2915904. https://doi.org/10.3758/APP.71.8.1683 (Checked 2026-11-09).
  • Raymond, J. E., Shapiro, K. L., and Arnell, K. M. (1992). Temporary suppression of visual processing in an RSVP task: An attentional blink? Journal of Experimental Psychology: Human Perception and Performance, 18(3), 849–860. https://doi.org/10.1037/0096-1523.18.3.849 (Checked 2026-11-09).
  • Shapiro, K. L., and Raymond, J. E. (2009). Attentional blink. Scholarpedia, 4(2):3320. http://www.scholarpedia.org/article/Attentional_blink (Checked 2026-11-09).

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · free to try

Ask about Attentional Blink: Measurement, Formula and Guide

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →