Flanker Task

Technical Manual

Script Author: Katja Borchert, Ph.D. (katjab@millisecond.com), Millisecond

Created: September 19, 2012

Last Modified: August 06, 2026 by K. Borchert (katjab@millisecond.com), Millisecond

Background

This script implements the original Flanker Test procedure by Eriksen & Eriksen (1974). The Flanker test is an inhibition test to assess the ability to suppress inappropriate responses evoked by task irrelevant information.

References

Eriksen, B.A. & Eriksen, C.W (1974) Effects of noise letters upon the identification of a target letter in a nonsearch task. Perception & Psychophysics, 16, 143-149.

Lisas Scores:

Vandierendonck A. On the Utility of Integrated Speed-Accuracy Measures when Speed-Accuracy Trade-off is Present. J Cogn. 2021 Mar 18;4(1):22. doi: 10.5334/joc.154. Erratum in: J Cogn. 2021 Nov 10;4(1):59. doi: 10.5334/joc.192. PMID: 33778368; PMCID: PMC7977027.

Duration

30+ minutes

Description

The Eriksen Flanker Test is a simple categorization task of 4 target letters (H, K, S, C) into "angular" vs. "curved" categories. Participants are asked to press "Q" if it's H or K, press "P" if it's S or C. Target letters can appear alone ("no-noise") or be flanked by noise letters ("noise") that can either be close together or far apart.

Procedure

Design:

3 (spacings) x 5 (noise conditions) tested within participants,
2 control conditions of nonoise targets tested in a mixed vs. blocked format within participants

Session Set-Up:

6 experimental sessions (see Eriksen & Eriksen, 1974, p.146)
1 experimental session: 3 blocks.Mixed (3*32 trials = 3*(28 noisetrials + 4 nonoisetrials)) + 1 block.Blocked (12 nonoisetrials)

= 6*3*28 noisetrials + 6*3*4 nonoisetrials(mixed) * 6*1*12 nonoisetrials(blocked)
= 504 noisetrials + 72 nonoisetrials(mixed) + 72 nonoisetrials(blocked)

=> nonoisetrials:
nonoisetrials per target letter (4 letters) in mixed and blocked format= 72/4= 18 (see Eriksen & Eriksen, 1974, p.146)

=> noisetrials:
noisetrials per letter spacing condition (3 spacing conditions)= 504/33 = 168
noisetrials per letter spacing condition per noise conditions (5+2* conditions) = 168/7 = 24 (see Eriksen & Eriksen, 1974)
noisetrials per letter spacing condition per noise conditions per letter (4)= 24/4 = 6

Spacing Conditions: Eriksen & Eriksen (1974)

1)about 0.06 degree visual angle separation (depends on viewing distance and font size) -
2)about 0.5 degree visual angle separation (depends on viewing distance and font size)
3)about 1 degree visual angle (depends on viewing distance and font size)
(4) distance from target to fixation cross: 0.5 degree visual angle)

!!! SPACING = NEEDS FINE TUNING
the default coordinates for the noise stimuli in this script are only
rough approximations. They need to be fine-tuned by each experimenter.
Specifically: fine tune
•distance target-to-fixation cross under EDITABLE VALUES (parameters.fixDistance)
•spacings between stimuli under EDITABLE VALUES (parameters.spacing1, parameters.spacing2, parameters.spacing3)
•fontstyle: Erikson & Erikson (1974) used Paratipe 18point Futura Bold Style,
=> this script uses Monospaced Courier New, 0.5cm: can be changed under EDITABLE VALUES (parameters.fontheight_innm)
If Monospaced Courier New is not available try a different monospaced fontstyle

Noise Conditions:

1) noise equal to target: all 6 noise letters = target letter (e.g, HHHHHHH)
2) noise similar to target: all 6 noise letters = the other letter in the target category (e.g. KKKHKKK)
3) noise other target set: all 6 noise letters = one of the letters of the opposite target category (e.g. SSSHSSS)
choice of letter of opposite target category is balanced across experiment
4) noise similar to target: noise = 3 letters that have similar characteristics (e.g. NWZHNWZ)
position of the three letters balanced across experiment
5) noise dissimilar to the target: noise = 3 letters with dissimilar characteristics (e.g. QJGHQJG)
position of the three letters balanced across experiment

•
Eriksen & Eriksen (1974) ran double the amount of trials in noise conditions 4 and 5 than in the remaining 3 conditions
(refer to Eriksen & Eriksen, 1974, Table 3, p.147), numberwise this is equivalent to running 7 conditions

(D) BLOCKS:
Each block.Mixed runs 32 trials: 7 H-trials, 7 K-trials, 7 S-trials, 7 C-trials, as well as
4 nonoise trials (one of each letter).
Trials are randomly chosen without replacement.

3 blocks make up 1 experimental session:
Every combination of letter x spacing condition x noise condition is run once (combinations of
noise condition 4 and 5 are run twice),
H: 1(=noise condition)1(=spacing condition),12,13,21,22,23,31,32,33,41,42,43,41,42,43,51,52,53,51,52,53
K: 11,12,13,21,22,23,31,32,33,41,42,43,41,42,43,51,52,53,51,52,53
S: 11,12,13,21,22,23,31,32,33,41,42,43,41,42,43,51,52,53,51,52,53
C: 11,12,13,21,22,23,31,32,33,41,42,43,41,42,43,51,52,53,51,52,53

Balancing of the different letter rotations/selections of noise conditions 3 (selection), 4 (rotation), 5 (rotation)
is achieved across all 6 experimental sessions (but not across a single experimental session).
This script runs all 6 experimental sessions in a row.

Each block.Blocked runs 12 nonoise trials, 3 of each of the 4 target letters. Letters
are randomly sampled without replacement.

(E) TRIALS:
•Fixation Cross => press Space Bar to bring on target (and noise letters)
•target appears directly above Fixation Cross
( Eriksen & Eriksen, 1974: 1/2 degree visual angle above,
the location in this script is only a rough approximation and needs fine-tuning under EDITABLE VALUES)
•noise letters appear flanking the target (in noise trials),
type of noise letters depends on noise condition, spacing depends on spacing condition

• In Eriksen & Eriksen (1974) the target and noise letters remained on the screen for 1s (regardless of whether
a response is given) and then disappeared together. In this script the response terminates the trial.
• In Eriksen & Eriksen (1974) error trials were repeated later in the session. In this script, error trials
are not rerun.

Practice:

By default, three blocks of 32 trials are given as practice. Number can be edited under section EDITABLE PARAMETERS.
During practice, participants receive feedback if they make a mistake.

Stimuli

target letters: H, K (angular), S, C (curved)

Instructions

Instructions are not original and can be edited in script flankertask_instructions_inc.iqjs

Scoring

Linear Integrated Speed-Accuracy Scores (Lisas): (Vandierendonck, 2021)

A LISAS score is simply the weighted sums of the RT and PE measures, with weights being calculated as the ratio of
the (correct) response time standard deviation to the error rate standard deviation.
LISAS output is expressed in units of time. It represents an individual's "error-corrected processing speed".

LISAS = meanRT(of correct responses) + (sdRT/sdErrors) * propErrors

=> A low LISAS means the participant was both fast and accurate; the minimum LISAS that can be assigned to a participants
equals their mean correct response time (this happens with error rates = 0);
=> A high LISAS means the participant struggled with accuracy or had to drastically slow down to maintain accuracy

in this script sdRT and sdErrors are based on the pooled standard deviation of all MIXED test trials

Summary Data

File Name: flankertest_summary*.iqdat

Data Fields

Name	Description
inquisit.version	Inquisit version number
computer.platform	Device platform: win \| mac \|ios \| android
computer.touch	0 = device has no touchscreen capabilities; 1 = device has touchscreen capabilities
computer.hasKeyboard	0 = no external keyboard detected; 1 = external keyboard detected
startDate	Date the session was run
startTime	Time the session was run
subjectId	Participant ID
groupId	Group number
sessionId	Session number
elapsedTime	Session duration in ms
completed	0 = Test was not completed 1 = Test was completed
propErrorNoise1-propErrorNoise5	Error proportion, separately for each noise condition
propErrorSpacing1-propErrorSpacing3	Error proportion, separately for each spacing condition
propErrorNoNoiseM	Error proportion of nonoise trials tested in a mixed (M) design
propErrorNoNoiseB	Error proportion of nonoise trials tested in a blocked (B) design
meanRTNoise1-meanRTNoise5	Mean latency of correct responses, separately for each noise conditions
meanRTSpacing1-meanRTSpacing3	Mean latency of correct responses, separately for each spacing condition
meanRTNoNoiseM	Mean latency of correct responses to nonoise trials tested in mixed (M) design
meanRTNoNoiseB	Mean latency of correct responses to nonoise trials tested in blocked (B) design
Speed-Accuracy Trade Off: Linear Integrated Speed-Accuracy Score, Lisas ("Error-Corrected Processing Speed"); Only Calculated For The Noise Categories In Mixed Trials
sdRTPooled	The pooled standard deviation of correct response times across all MIXED test trials
errorSDPooled	The pooled standard deviation of error rates across all MIXED test trials
lisasNoise0	LISAS score for Noise1 trials ("Baseline", no flankers in Mixed Trials)
lisasNoise1	LISAS score for Noise1 trials
lisasNoise2	LISAS score for Noise2 trials
lisasNoise3	LISAS score for Noise3 trials (Response Conflict)
lisasNoise4	LISAS score for Noise4 trials
lisasNoise5	LISAS score for Noise5 trials
Bias Scores For Noise: Based On Lisas Differences With Baseline Condition (No Noise Condition In Mixed Trials)
flankerBiasNoise1	Flanker Bias Score based on LISAS: Difference in LISAS scores ("error-corrected processing speed") from Noise1 and Noise0 Trials (lisasNoise1 - lisasNoise0) => positive difference, "error-corrected processing speed" was faster for Noise0 than for Noise1 ones => a positive difference supports that processing Noise1 trials required more cognitive effort, time, or accuracy sacrifices than processing Noise0 trials it's no unreasonable to expect to see a negative difference for this pairing, as it's likely that the identical flankers help with response decision.
flankerBiasNoise2	Flanker Bias Score based on LISAS: Difference in LISAS scores ("error-corrected processing speed") from Noise2 and Noise0 Trials (lisasNoise2 - lisasNoise0) => positive difference, "error-corrected processing speed" was faster for Noise0 than for Noise2 ones => a positive difference supports that processing Noise2 trials required more cognitive effort, time, or accuracy sacrifices than processing Noise0 trials
flankerBiasNoise3	Flanker Bias Score based on LISAS: Difference in LISAS scores ("error-corrected processing speed") from Noise3 (Response Conflict!) and Noise0 Trials (lisasNoise3 - lisasNoise0) => positive difference, "error-corrected processing speed" was faster for Noise0 than for Noise3 ones => a positive difference supports that processing Noise3 trials required more cognitive effort, time, or accuracy sacrifices than processing Noise0 trials
flankerBiasNoise4	Flanker Bias Score based on LISAS: Difference in LISAS scores ("error-corrected processing speed") from Noise4 and Noise0 Trials (lisasNoise4 - lisasNoise0) => positive difference, "error-corrected processing speed" was faster for Noise0 than for Noise4 ones => a positive difference supports that processing Noise4 trials required more cognitive effort, time, or accuracy sacrifices than processing Noise0 trials
flankerBiasNoise5	Flanker Bias Score based on LISAS: Difference in LISAS scores ("error-corrected processing speed") from Noise5 and Noise0 Trials (lisasNoise5 - lisasNoise0) => positive difference, "error-corrected processing speed" was faster for Noise0 than for Noise5 ones => a positive difference supports that processing Noise5 trials required more cognitive effort, time, or accuracy sacrifices than processing Noise0 trials

Raw Data

File Name: flankertest_raw*.iqdat

Data Fields

Name	Description
build	Inquisit version number
computer.platform	Device platform: win \| mac \|ios \| android
computer.touch	0 = device has no touchscreen capabilities; 1 = device has touchscreen capabilities
computer.hasKeyboard	0 = no external keyboard detected; 1 = external keyboard detected
date	Date the session was run
time	Time the session was run
subject	Participant ID
group	Group number
session	Session number
blockcode	The name the current block (built-in Inquisit variable)
blocknum	The number of the current block (built-in Inquisit variable)
trialcode	The name of the currently recorded trial (built-in Inquisit variable)
trialnum	The number of the currently recorded trial (built-in Inquisit variable) trialnum is a built-in Inquisit variable; it counts all trials run even those that do not store data to the data file.
trialCount	Counts the number of trials
responseCondition	Counterbalanced response key condition
noiseCondition	Stores the current noise condition (1-5) 1) noise equal to target: all 6 noise letters = target letter (e.g, HHHHHHH) 2) noise similar to target: all 6 noise letters = the other letter in the target category (e.g. KKKHKKK) 3) noise other target set: all 6 noise letters = one of the letters of the opposite target category (e.g. SSSHSSS) choice of letter of opposite target category is balanced across experiment 4)noise similar to target: noise = 3 letters that have similar characteristics (e.g. NWZHNWZ) position of the three letters balanced across experiment 5) noise dissimilar to the target: noise = 3 letters with dissimilar characteristics (e.g. QJGHQJG) position of the three letters balanced across experiment
spacingCondition	Stores the current spacing condition (1-3)
trialType	Contains all trialinformation Example: "K53B" -> target K, condition 5, spacing 3, B=refers to one of three possible rotations of the noise stimuli in condition 5
stimulusItem.1-stimulusItem.7	The presented stimuli
response	The participant's response (scancode of response buttons) 16 = Q (left) 25 = P (right)
correct	Accuracy of response: 1 = correct response; 0 = otherwise
latency	The response latency (in ms); measured from onset of stimuli

Parameters

The procedure can be adjusted by setting the following parameters.

Name	Description	Default
Sizing Parameters
fontHeightMM	Fontheight in mm	5
fixedDistanceMM	The distance of the target stimuli from the fixation cross in mm	10
spacing1MM	The distance between stimuli in spacing condition 1 in mm	3
spacing2MM	The distance between stimuli in spacing condition 2 in mm	10
spacing3MM	The distance between stimuli in spacing condition 3 in mm	20
Timing Parameters
trialTimeout	The max. trial duration (in ms)	1000
trialDelay	The delay (in ms) of the trial after pressing the spacebar	0
Design Parameters
maxPracticeBlocks	Determines how many practice blocks should be run equivalent to one experimental session)	3
Response Keys
leftKey	The assigned left key (on a QWERTY keyboard)	"Q"
rightKey	The assigned right key (on a QWERTY keyboard)	"P"

Flanker Task

Technical Manual

Background

References

Lisas Scores:

Duration

Description

Procedure

Design:

Session Set-Up:

Spacing Conditions: Eriksen & Eriksen (1974)

Noise Conditions:

Practice:

Stimuli

Instructions

Scoring

Linear Integrated Speed-Accuracy Scores (Lisas): (Vandierendonck, 2021)

Summary Data

Data Fields

Speed-Accuracy Trade Off: Linear Integrated Speed-Accuracy Score, Lisas ("Error-Corrected Processing Speed"); Only Calculated For The Noise Categories In Mixed Trials

Bias Scores For Noise: Based On Lisas Differences With Baseline Condition (No Noise Condition In Mixed Trials)

Raw Data

Data Fields

Parameters

Sizing Parameters

Timing Parameters

Design Parameters

Response Keys