PEBL Hub — IRB Fact Sheet — 2026-08-23
Operation Span Task (OSPAN)
working memory
executive function
dual-task processing
mathematical processing
Category:Memory
Duration:~15 minutes
Ages:18+
Populations:adults, research, clinical
Parameters:8 configurable parameters: feedbacktimeout (1500ms), distractfile (distraction.csv), batterythreshold (0.85), batteryFeedback (1), distracttimeoutminimum (1000ms), distracttimeoutmultiplier (2.5), requirethresholdprebattery (1), allowsecondchance (0), responsemode (none)
Version:2.4
Description
Complex working memory span task that requires participants to remember letters while solving math problems.
Participants solve math equations (e.g., '(3 * 2) + 4 = 11') and judge them as True or False, then see a letter to remember. After a series of math-letter pairs (set size 3-7), they recall the letters in order by clicking on a grid. The task includes practice blocks for letters only, math only, and combined math-letter trials. Math performance is monitored and participants must maintain 85% accuracy on math problems to ensure divided attention. The main test consists of 15 trials with varying set sizes (3, 4, 5, 6, 7) repeated 3 times each in random order.
Measures: Complex working memory span, executive function, dual-task processing, mathematical processing speed
Notes
- Task follows Unsworth et al. (2005) methodology with 3-7 letter set sizes
- Math accuracy threshold (default 85%) enforced to ensure divided attention
- Adaptive math timeout calculated from practice: mean + 2.5*SD with 1000ms minimum
- Practice includes: letters only (4 trials), math only (15 trials), combined (3 trials at size 2)
- Main test: 15 trials with set sizes [3,4,5,6,7] × 3 repetitions in randomized order
- Math problems loaded from distraction.csv file (arithmetic equations with T/F answers)
- Letter recall uses custom mouse-clickable grid with Clear/Del/Done buttons
- Real-time accuracy display shown in test phase (top right, red text, 50% larger font)
- Trial counter shows current trial / total trials during test phase
- Supports LSL markers for electrophysiological recording integration
- Supports online upload via InitializeUpload() and UploadFile()
- Uses HTML-lite formatted text with adaptive textboxes for instructions
- Includes 7 translations (en, es, de, pt, fr, nl, it) via JSON translation system
- Creates HTML report summarizing span score, math accuracy, and performance by set size
Key Parameters
| feedbacktimeout | Duration in milliseconds to display feedback after each trial (default: 1500) |
| distractfile | CSV file containing math problems (format: problem,answer,correctness) (default: distraction.csv) |
| batterythreshold | Minimum accuracy (0.0-1.0) required on math problems to proceed to main test. Typically 0.85 (85%) (default: 0.85) |
| batteryFeedback | Whether to show feedback during test battery (1=yes, 0=no) (default: 1) |
| distracttimeoutminimum | Minimum timeout in milliseconds for math problems during test battery (default: 1000) |
| distracttimeoutmultiplier | Multiplier for standard deviation when calculating math timeout: timeout = mean + (multiplier × SD). Unsworth et al. (2005) used mean + 2.5*SD based on extensive piloting (default: 2.5) |
| requirethresholdprebattery | Require meeting accuracy threshold before starting test battery (1=yes, 0=no) (default: 1) |
| allowsecondchance | Allow a second practice attempt if first practice fails to meet threshold (1=yes, 0=no) (default: 0) |
| responsemode | Response collection mode (currently only 'none' is supported, using custom mouse grid) (default: none) |
| testreps | Number of repetitions for each set size in the test battery (e.g., 3 means each set size appears 3 times) (default: 3) |
| testminsize | Minimum set size (number of letters to recall) in the test battery. Must be at least 2 (default: 3) |
| testmaxsize | Maximum set size (number of letters to recall) in the test battery. Cannot exceed 12 (the memory grid size: 3 columns × 4 rows) (default: 7) |
Scoring & Interpretation
| Validation | validated |
| Psychometric data | Available |
| Normative data | Available |
Example Data Output (5 sample rows)
| subnum | trial | probtime | memtime | length | numstim | stim | resp | memcorr | time | distcorr | numcorrdist | totnumdist | runningcorrect | runningtotal |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| example | 1 | -1 | 1000 | 2 | 3 | X|C | X|C | 1 | -1 | -1 | 0 | 0 | 0 | 0 |
| example | 2 | -1 | 1000 | 3 | 5 | F|X|H | F|X|H | 1 | -1 | -1 | 0 | 0 | 0 | 0 |
| example | 3 | -1 | 0 | 0 | 1 | NA | 2862 | 0 | 0 | 1 | 0 | 1 | ||
| example | 4 | -1 | 0 | 0 | 1 | NA | 3014.5 | 1 | 2 | 2 | 2 | 3 | ||
| example | 5 | -1 | 0 | 0 | 1 | NA | 2299 | 1 | 1 | 1 | 3 | 4 |
References
Turner, M. L., & Engle, R. W. (1989). Is working memory capacity task dependent? Journal of Memory and Language, 28(2), 127-154.
Unsworth, N., Heitz, R. P., Schrock, J. C., & Engle, R. W. (2005). An automated version of the operation span task. Behavior Research Methods, 37(3), 498-505.
Unsworth, N., Redick, T. S., Heitz, R. P., Broadway, J. M., & Engle, R. W. (2009). Complex working memory span tasks and higher-order cognition: A latent-variable analysis of the relationship between processing and storage. Memory, 17(6), 635-654.
Authorship & License
| Original author | D.L. Hegarty & Dufflecoat Enterprises (2014) |
| PEBL implementation | S.T. Mueller |
| License | GPLv3 https://www.gnu.org/licenses/gpl-3.0.html |