Function Acquisition Speed Test (FAST)

Background

The Function Acquisition Speed Test (FAST) is a behavior-analytic implicit measurement tool designed to quantify the strength of pre-existing associations between stimuli. This in turn can be used as a measure of a person's automatic, implicit attitudes or biases.

First introduced by Brian Roche and colleagues in 2012, it assesses how quickly a person can learn a new, arbitrary categorization rule. The core principle is that learning happens much faster when the new rule aligns with a person's pre-existing, subconscious beliefs. Specifically, items that are more strongly connected in one's mind are thought to be more easily sorted into the same categories, resulting in higher performance and faster sorting times. The FAST is suggested to be an alternative to the Implicit Association Test (IAT), the Implicit Relational Assessment Procedure (IRAP) or the Extrinsic Affective Simon Task (EAST).

The basic task in the FAST is similar to the one used in other implicit tests: Participants are asked to sort various words or images using keyboard responses into two categories. Unlike other tests, such as the IAT, that focus primarily on response times in milliseconds, the FAST paradigm focuses on a participant's learning curves in two different test blocks. Hence, the FAST does not provide the sorting rules like the IAT does; the sorting rules need to be learned via trial and error learning.

• Consistent Block (Consistent with prior learning): Concepts that are hypothesized to be more strongly connected in people's minds (based on prior learning/experiences) need to be learned - by trial and error- to be sorted by the same response keys. For example, a block might require grouping Male Names and Career Words on the left key, and Female Names and Family Words on the right key.

• Inconsistent Block (Inconsistent with prior learning): The rules are flipped for the Consistent Block. The participant must now learn to group Male Names with Family Words on one key, and Female Names with Career Words on the other.

Instead of tracking millisecond delays, the computer tracks the difference in correct and incorrect learning trials per block duration (fluency criteria) and calculates a per-minute Response Fluency Differential (RFD) score which reflects the degree to which the response fluency on the consistent block is greater than the one on the inconsistent block. Positive RFD scores indicate greater response fluency in consistent condition (consistent with hypothesis); negative RFD scores indicate greater response fluency in inconsistent condition (unexpected result). In addition, Roche and colleagues have begun looking at 'Learning Slope Measures' and traditional Signal Detection Measures as additional FAST metrics.

The Inquisit FAST script provides a generic template to run the FAST procedure either with words or images. The template scripts use the classic flower-insect categories as well as positive and negative attributes but these categories can easily be customized.

Task Procedure

The FAST procedure consists of 2 test blocks (consistent and inconsistent) that are run in random order. Each block runs 52 learning trials that consist of 13 positive attribute trials, 13 negative attribute trials, 13 flower trials and 13 insect trials. The trial order is randomized with the constraint that each group of four consecutive trials run one of the four conditions. During the consistent block, participants learn (through trial and error) to categorize flower stimuli and positive attributes with the same key (e.g. 'Z') and use the opposite response key ('M') to sort insect stimuli and negative attributes. During the inconsistent rules, the pairing changes and participants have to learn to use the same response keys for flowers and negative attributes ('Z') and insects and positive attributes ('M'), respectively. Each individual learning trial starts with a blank screen for 500ms, followed by the stimulus. The stimulus is presented for 3000ms or until a response is provided, whichever comes first. Performance feedback (a green ✓ or a red X) is provided for 500ms immediately upon responding.

What it Measures

The Function Acquisition Speed Test (FAST) measures implicit attitudes

Psychological domains

• Implicit Cognition: Cognitive Processes not under conscious control
• Social Cognition: People's thoughts and perceptions of themselves and others in the social world

Main Performance Metrics

• **fast RFD: a per-minute Response Fluency Differential (RFD) score; main FAST measure
• Learning Slope Difference: Difference in Learning Slope Measures (based on regression line); positive scores (steeper learning in Consistent Block) are expected
• Signal Detection Measures: measure of sensitivity and response criterium (use with caution, see User Manual for more information)

Psychiatric Conditions

The FAST has been used with the following patient groups

• Depression
• Anxiety
• Trauma and Phobias
• Substance Abuse
• Eating Disorders

Test Variations

References

Search Google Scholar for peer-reviewed, published research using the Inquisit Function Acquisition Speed Test (FAST).

---

O’Reilly, A., Roche, B., Ruiz, M., Tyndall, I., & Gavin, A. (2012). The function acquisition speed test (FAST): A behavior analytic implicit test for assessing stimulus relations. The Psychological Record, 62(3), 507–528. https:// doi. org/ 10. 1007/ bf033 95817

Cummins, J., Roche, B., Tyndall, I., & Cartwright, A. (2018). The relationship between differential stimulus relatedness and implicit measure effect sizes. Journal of the Experimental Analysis of Behavior, 110(1), 24–38. https:// doi. org/ 10. 1002/ jeab. 437

Cummins, J., & Roche, B. (2020). Measuring differential nodal distance using the function acquisition speed test. Behavioural Processes, 178, 104179. https:// doi. org/ 10. 1016/j. beproc. 2020. 104179

Watters, A., Cummins, J., & Roche, B. (2023). How to Build and How not to Build an Implicit Measure in Behavior Analysis: A case Study Using the Function Acquisition Speed Test. Perspectives on Behavior Science, 46(3–4), 459–492. https://doi.org/10.1007/s40614-023-00387-w