Feedback Functions

A major concept in Pedal is a Feedback Function, which encapsulate the idea of a Feedback Condition and a Feedback Response pair. Pedal provides a large number of Feedback Functions and mechanisms to create new ones. As instructors, you have can control and override the default behavior of these feedback functions in various ways.

Some Composite Feedback Functions do not have keyword parameters, because of their complexity. However, we strive to universally provide these parameters!

This documentation describes Feedback Functions from the perspective of teachers who are interested in writing autograding scripts. If you are interested in extending Pedal, you might want to consult the Developers: Feedback Functions documentation page.

title: str: The descriptive name for this Feedback Function that will be shown to students as a “header” of sorts.

message: str: The actual text of the feedback that will be shown to the student. Depending on the current Formatter and Environment, may support HTML or Markdown. Will override the message_template if provided.

message_template: str

The string template used to generate the text to show the student if the message is not provided. Any field names wrapped in curly braces will be interpolated when the Feedback Function meets its condition. These fields can be further manipulated by incorporating the Formatter’s supported conversion flags:

# A message_template can have fields and conversion flags:
"Your code returned {student_answer!python_value}"

# Which will generate the HTML output:
"Your code returned <code>5</code>"

# Or the plain-text output:
"Your code returned 5"

Usually, you will override the message instead of the template, to set a specific message instead.

else_message: str: A string to render as the message if the Feedback Function’s condition is NOT met. Except for Feedback Functions that have a positive valence - then this message is shown when the FF’s condition IS met. Otherwise follows the same rules as the message field. The typical use case for this parameter is to display a positive message if the students happen to complete a specific subtask or task and a Resolver is chosen that will show multiple kinds of feedback. Note that because A) the default Resolver only shows one piece of feedback, B) that resolver prioritizes errors, and C) that resolve prioritizes the final Success over lesser successes, you are unlikely to see these messages in most cases as the actual feedback. However, the message will be provided as a Positive feedback, for environments that choose to render such things.

else_message_template: str: The corresponding string template parameter for the else_message parameter. Works the same as its corresponding message_template field, just for providing text to the student in the case of no else_message being provided. Usually not used.

category: FeedbackCategory: An internal name categorizing the feedback condition, such as "syntax" or "runtime". Usually you don’t want to change this.

kind: FeedbackKind: An internal name categorizing the feedback response, such as "hint" or "mistake". Usually you don’t want to change this.

fields: dict[str, Any]: Internal data for the Feedback Function used in constructing its message and justification.

field_names: list[str]: An explicit list of the field names for this Feedback Function. If not provided, then it will be inferred from the fields. Usually you won’t need to do anything with this.

tool: str: What Tool was responsible for creating this feedback function. Usually you don’t want to change this.

label: str: A unique name for this general class of feedback. You might override this to give a more specific identifier to the feedback for future analysis purposes.

justification: str | tuple[str, str]: An internal explanation for why this feedback was chosen (and/or not chosen), which can be shown to the instructor (but not the student). Usually you don’t want to change this.

priority

score

correct

muted

unscored

activate