Decision Making
Feeling Forecast Task
Also called: affective forecasting task, impact bias task, emotional prediction task
Compare predicted feelings with reports after point gains and losses, immediately and after a short pause.
6-9 minUntimed
Free to start: 10 free participants per account, then $0.99 per participant for academic research, or any yearly plan. Try it here first, with no account.
- Papers, 2021 to 2025
- 209
- Evidence grade
- C
- Registered clinical studies
- 2
What it measures
Measures how predicted feelings compare with feelings reported after gaining or losing points. Participants forecast each possible outcome, then rate their feelings when the result appears and after a short pause. Scores show immediate and later forecast errors.
Constructs
- Affective forecasting
- Impact bias
- Predicted and experienced emotion
Research uses
- Emotion and decision-making research
- Forecasting accuracy research
- Within-study comparison
How the paradigm works
Affective forecasting compares predicted feelings with feelings reported after an event. This task uses point outcomes within one session. Each round starts with 100 points and adds or removes 20 or 60 points. The program fixes the outcomes before the participant answers, and the instructions say so. Each block contains one gain and one loss at each magnitude in seeded order.
More about the paradigm
Before seeing each result, participants predict their feelings for both possible outcomes: immediately and after a pause following their first feeling report. The result then appears with the first rating bar. After that report, a quiet pause lasts 10 seconds by default, followed by the same rating bar again. Every bar runs from very bad to very good, with the middle meaning neither, and has no initial value. The later score refers to this within-session pause, not lasting mood outside the session.
The main measure compares the forecast with the report matching the actual outcome. Its sign is reversed for losses so a positive impact bias means a more positive gain or a more negative loss was predicted than reported. Later bias, absolute errors, gain and loss bias and predicted versus reported change are also given. Only complete, correctly matched pairs enter each measure. The actual pause and elapsed time since the result are recorded.
Scores
- Immediate Impact Biasscore
- Mean forecast minus reported feeling after gains, and reported feeling minus forecast after losses, on a 0 to 100 feeling scale.
- Positive means a more positive gain or more negative loss was predicted than reported
- Later Impact Biasscore
- The same signed error after the pause following the first report.
- Positive means a more positive gain or more negative loss was predicted than reported later
- Immediate Absolute Errorscore
- Mean absolute gap between forecast and immediate report.
- Smaller means closer forecasts in this task
- Later Absolute Errorscore
- Mean absolute gap between forecast and report after the pause.
- Smaller means closer later forecasts in this task
- Gain Impact Biasscore
- Mean immediate forecast minus report on gains.
- Positive means the gain was predicted to feel better
- Loss Impact Biasscore
- Mean immediate report minus forecast on losses.
- Positive means the loss was predicted to feel worse
- Predicted Changescore
- Mean immediate minus later forecast, multiplied by the outcome sign, on rounds with both matched reports.
- Positive means feelings were predicted to move away from the outcome direction
- Reported Changescore
- Mean immediate minus later report, multiplied by the outcome sign, on rounds with both matched reports.
- Positive means feelings moved away from the outcome direction
- Matched Forecastscount
- Outcome- and time-matched forecast-report pairs, up to two per round.
- Informational
- Complete Roundscount
- Rounds with both immediate and later matched forecast-report pairs.
- More complete rounds support interpretation
- Check Correct0/1
- 1 when the question about programmed outcomes was answered correctly.
- 1 supports interpretation
- Median Rating Timems
- Median response time on forecasts and reports at or after 100 ms.
- Informational
- Ratings Under 100 mscount
- Forecasts and reports under 100 ms, excluded from the error scores.
- 0 expected
Use and evidence
209 papers from 2021 to 2025 name this task in their title or abstract, up 8% compared with 2016 to 2020. That ranks it 147 of 582 test families in our census. See the counts for every test.
Show the numbers
| Year | Papers |
|---|---|
| 2025 | 59 |
| 2024 | 42 |
| 2023 | 41 |
| 2022 | 39 |
| 2021 | 28 |
| 2020 | 38 |
| 2019 | 51 |
| 2018 | 38 |
| 2017 | 31 |
| 2016 | 35 |
| 2015 | 38 |
| 2014 | 29 |
| 2013 | 31 |
| 2012 | 37 |
| 2011 | 37 |
| 2010 | 36 |
| 2009 | 42 |
| 2008 | 29 |
| 2007 | 21 |
| 2006 | 19 |
| 2005 | 9 |
| 2004 | 11 |
| 2003 | 7 |
| 2002 | 8 |
| 2001 | 6 |
| 2000 | 3 |
Use grade C
100 to 299 papers from 2021 to 2025.
Evidence grade C
At least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies.
- Papers on test-retest reliability
- 3
- Papers on reliability or validity
- 26
- Registered clinical studies using it as an outcome
- 2
Measured on ConductCognition
Pooled across the studies that feed our platform norms, by device group. A figure needs 50 or more people.
No pooled figure yet.
Note on the counts: Programmed point gains and losses; paired forecasts and reports immediately and after a within-session pause; no norms..
How the grades work
Use: A is 1,000 or more papers from 2021 to 2025, B 300 to 999, C 100 to 299, D 30 to 99 and E fewer than 30. Evidence: A means at least 25 papers report its test-retest reliability and at least 25 registered clinical studies use it as an outcome; B, at least 8 test-retest papers, and either 5 registered clinical studies or 150 reliability or validity papers; C, at least 2 test-retest papers, 30 reliability or validity papers or 3 registered clinical studies; D, little published measurement evidence found. The counts measure use. They include every paper that names the task in its title or abstract. Grades describe the published research on this task, not ConductCognition's version of it.
Source: ConductCognition test census 0.39.0, counted 8 November 2026 from OpenAlex and ClinicalTrials.gov.
Task design
The trial sequence as the test runs in ConductCognition.
- Rounds
- 8 scored rounds by default, in 2 blocks of 4. Settings allow 1 to 3 blocks, or 4 to 12 rounds.
- Outcomes
- Each block has +20, -20, +60 and -60 points once. Every round starts with 100 points. Outcomes are programmed in advance and disclosed as such.
- Forecasts
- Four forecasts per round: gain and loss, each immediately and after the pause following the first report. Sign order is seeded.
- Reports
- Two reports per round: when the result appears and after the pause. The same 0 to 100 bar is used for all forecasts and reports, with no initial value.
- Pause
- 10 seconds after the first report by default; 5 to 30 seconds in settings. Actual pause duration and time since the result are saved.
- Practice
- One check question with feedback, then one practice round. Ratings have no response deadline.
- Scores
- Immediate impact bias first, then later bias, absolute errors, gain and loss bias, and predicted versus reported change.
This original repeated point-outcome procedure measures forecasting over a short pause within the session. It uses no monetary stakes or external events. Compare groups using the same settings. No percentile output.
Quality checks
Each session is marked valid, needs review or invalid, with the reason, before you see its scores.
Feeling ratings have no correct answer. Ratings under 200 ms on at least half the screens invalidate. Missing forecast-report pairs or a wrong check answer are flagged. A later report before the recorded pause ends is not scored.
Why these rules
There is no correct option to compare with chance, so accuracy is not judged against chance. An answer under 200 ms is faster than a simple visual reaction, so it cannot reflect the item. Following response-time effort screening for untimed tests (Wise and Kong 2005), the session is invalid only when at least half the answers are that fast. A forecast and report must refer to the same outcome, round and horizon. Identical ratings can describe a stable feeling and do not invalidate.
No percentile. Scores and quality flags only.
Running this test
Settings you can change
| Setting | Standard | Allowed |
|---|---|---|
| Blocks | 2 | 1 to 3 |
| Pause (seconds) | 10 | 5 to 30 |
Yearly plans can change settings within these limits. Free and pay-as-you-go studies use the standard settings.
Need a setting this test does not have?
Devices and languages
- Untimed test: runs on desktops, laptops, tablets and phones. Participant devices
- Languages: English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese.
- Available on: Every plan, including the free plan.
- No percentile. Scores and quality flags only. How norms work
- Recruiting online? See the setup guides for Prolific, Sona Systems, CloudResearch Connect and Qualtrics.
Cite and describe
Participants completed the Feeling Forecast Task in a web browser (ConductCognition task version 1.0.0, scoring version 1.0.0; Kermer, Driver-Linn, Wilson & Gilbert, 2006; Wilson & Gilbert, 2005), with the standard settings (blocks: 2; pause (seconds): 10). The main score was immediate impact bias (mean forecast minus reported feeling after gains, and reported feeling minus forecast after losses, on a 0 to 100 feeling scale).
ConductCognition versions: task version 1.0.0, scoring version 1.0.0. Every export records the versions each session used, for a methods section.
- Kermer, D. A., Driver-Linn, E., Wilson, T. D., & Gilbert, D. T. (2006). Loss aversion is an affective forecasting error. Psychological Science, 17(8), 649-653. https://doi.org/10.1111/j.1467-9280.2006.01760.x
- Wilson, T. D., & Gilbert, D. T. (2005). Affective forecasting: Knowing what to want. Current Directions in Psychological Science, 14(3), 131-134. https://doi.org/10.1111/j.0963-7214.2005.00355.x
Source papers
Questions
How long does the Feeling Forecast Task take?
About 6 to 9 minutes, as listed in the test library.
What device does a participant need?
Untimed test: runs on desktops, laptops, tablets and phones. The browser must be Chrome, Edge, Firefox or Safari.
What scores does it produce?
Immediate Impact Bias, Later Impact Bias, Immediate Absolute Error, Later Absolute Error, Gain Impact Bias, Loss Impact Bias, Predicted Change, Reported Change, Matched Forecasts, Complete Rounds, Check Correct, Median Rating Time and Ratings Under 100 ms. Each session is also marked valid, needs review or invalid, with the reason.
How widely is the Feeling Forecast Task used in research?
209 papers from 2021 to 2025 name it in their title or abstract, up 8% compared with 2016 to 2020. That ranks it 147 of 582 test families in the ConductCognition test census (use grade C). Source: ConductCognition test census 0.39.0, counted 8 November 2026 from OpenAlex and ClinicalTrials.gov.
How much reliability evidence does the Feeling Forecast Task have?
In the published research, 3 papers report its test-retest reliability and 2 registered clinical studies use it as an outcome (evidence grade C in the ConductCognition test census).
Does it report a percentile?
No. Scores and quality flags only.
Which settings can a researcher change?
Blocks (1 to 3) and pause (seconds) (5 to 30). Settings can be changed on yearly plans; free and pay-as-you-go studies use the standard settings.
Which languages can participants use?
Participants can take this test in English, Spanish, French, German, Italian, Portuguese (Brazil), Chinese (Simplified) and Japanese. English is on every plan, and the other languages come with yearly plans. Have a native speaker read the machine-translated text before launch.
Is this a clinical test?
No. ConductCognition is for research use only. It runs, scores and exports the test; clinical interpretation is outside the platform.
Every test: the full test list.
Run this test in your next study
Start free with 10 free participants on any test, then use participant credits or a yearly plan.