> For the complete documentation index, see [llms.txt](https://manymanys.gitbook.io/mm1-lab-manual/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://manymanys.gitbook.io/mm1-lab-manual/method/phase-length-and-learning-criterion.md).
# Phase length and learning criterion
Prior to any testing, each laboratory establishes a maximum phase length, carefully evaluating whether this trial load is feasible given the characteristics of their species, including attention span, fatigue, motivation, and session length constraints.
{% hint style="warning" %}
Phase length has direct implications for the overall trial demands placed on subjects. At the minimum phase length of 11 trials, completing a single condition requires at least 22 trials (one acquisition phase and one reversal phase). Across all three experimental conditions, this yields a minimum of 66 trials per subject.
{% endhint %}
Within each condition, subjects work through a series of trials organized into phases. To advance, they need to reach at least 80% correct responses within a phase. We also verify, using a binomial test, that this level of accuracy is statistically unlikely to have occurred by chance alone. The table below shows the number of trials required to reach learning criterion given a total number of trials.
Number of trials required to reach learning criterion
Total number of trials
Base phase length (same across all species/taxa)
9
11
...
12
15
...
16
20
...
20
25
...
24
30
Maximum phase length allowed (varies per species/taxon)
TBD
During testing, all subjects start with a base phase length of 11 trials, of which 9 must be answered correctly to meet criterion. If the subject achieves that criterion, it then moves to the next phase; if it does not, then ask whether this was your maximum phase length or not. If this was your chosen max phase length, then you stop the testing and the subject is dropped from the study; if this was NOT yet your chosen max phase length, then you will continue adding trials needed to reach the next phase length. Make sure to monitor the subject’s performance at each new phase length so you know whether to stop or not.
{% hint style="warning" %}
Please note that the learning criterion **varies across individuals in a lab.**
{% endhint %}
If a subject fails to reach criterion, it advances to the next phase length specified in the table below, which outlines the number of correct trials required to achieve 80% accuracy as a function of total trials administered. As phase length increases, the number of tolerated errors increases proportionally, this being an intentional accommodation for species that may require more trials to acquire a contingency.
```mermaid
flowchart LR
Z{"Start with the
base phase length
(11 trials)"}
Z--> A[Subject achieves
criterion]
subgraph one [" "]
A --> |no|C[Is the current phase length your maximum phase length?]
C --> |no| D[Add as many trials as needed to reach the
next phase length]
D --> E[Monitor subject performance over additional trials]
E -.-> |determine whether| A
end
A --> |yes|B[Phase completed,
start next phase]
C --> |yes| L[Subject is dropped]
style Z fill #cfe2f3, stroke : #111111
style A fill: #d9ead3, stroke : #111111
style B fill: #b6d7a8, stroke : #111111
style D fill: #cfe2f3, stroke : #111111
style E fill: #ffe599, stroke : #111111
style L fill: #f4cccc, stroke : #111111
style C fill: #ffe599, stroke : #111111
style one fill: #f3f6f4, , stroke : #f3f6f4
```
Following successful acquisition, subjects enter the reversal phase without interruption. In the reversal phase, reward contingencies are reversed without warning: the previously rewarded stimulus (S+) becomes unrewarded (S−), and vice versa. Subjects continue through trials with the goal of meeting the same learning criterion applied during acquisition.
Not all subjects will necessarily achieve the reversal criterion. Subjects that complete at least one full reversal contribute data to reversal learning analyses. All subjects, regardless of reversal performance, contribute data to acquisition analyses.
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://manymanys.gitbook.io/mm1-lab-manual/method/phase-length-and-learning-criterion.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.