{% extends "base.html" %} {% block title %}Criteria — FreeWeight{% endblock %} {% block content %}
List the qualities you want. Of each one, the wizard asks two questions — and these two questions are most of the value it adds.
If the answer to the second is "two", split it. "Not LinkedIn" is a vocabulary problem and a register problem; splitting it is what makes both measurable. The wizard will not split anything for you — the rubric is yours, and a model that edited it until it could measure it would be measuring its own edit.
{% if error %}{{ error }}
| Criterion | Weight | Rung | Split from | The two questions |
|---|---|---|---|---|
|
{{ criterion.name }}
{{ criterion.key }} {% if criterion.intent %} {{ criterion.intent }}{% endif %} |
{{ '%.0f' % (criterion.weight * 100) }}% |
{{ criterion.rung }}
{% if criterion.rung == 'judge' %}
{{ criterion.descriptors | length }} of {{ criterion.scale_points }} points described {% endif %} |
{{ criterion.split_from or '—' }} |
{% if criterion.needs_descriptors %}
This criterion is judged and its scale has no descriptors. A jury given "rate it 1 to {{ criterion.scale_points }}" and nothing else agrees with nobody — describe the top, the middle and the bottom in your own words. {% if criterion.one_quality is false %} You said this is two qualities. Split it, and each half becomes something a rule may be able to carry part of. {% else %} You said two people would grade this differently. That usually means the description names a topic rather than a quality, or fuses two qualities. Splitting it is the usual fix. {% endif %} |
No criteria yet. Add the first quality you want, in your own words.