qm

name	qm
description	Procedure for measuring outcomes that resist quantification.
output	{"format":"prose"}

Qualitative Outcome Measurement

Input: $ARGUMENTS

---

Overview

Many important outcomes can't be reduced to numbers: "Improved relationship quality," "Better work-life balance," "Increased confidence," "Stronger team culture."

This procedure provides rigorous ways to assess qualitative outcomes without forcing artificial metrics that miss the point.

Steps

Step 1: Define What "Better" Means

1. What outcome is being measured?
2. What would improvement look like concretely? (Observable behaviors, not feelings)
3. What would you notice if it got WORSE?
4. Who would notice the change? (Self, others, both?)

Step 2: Choose Measurement Approach

Approach	Best For	Method
Behavioral markers	Observable actions	Define specific behaviors that indicate the outcome
Rubric	Multi-dimensional quality	Create a descriptive scale for each dimension
Comparison	Before/after	Structured comparison against baseline
Narrative	Rich, complex outcomes	Structured storytelling with analysis
Indicator constellation	Hard-to-define outcomes	Multiple indirect indicators that together paint a picture

Step 3a: Behavioral Markers

Define observable behaviors that indicate the outcome:

BEHAVIORAL MARKERS for [outcome]:
Strongly present:
- [ ] [specific behavior you can observe]
- [ ] [specific behavior]

Moderately present:
- [ ] [behavior]
- [ ] [behavior]

Absent/declining:
- [ ] [behavior that would indicate absence]
- [ ] [behavior]

Score: Count of "strongly present" markers observed

Step 3b: Rubric

Create a descriptive scale:

Level	Description	Observable Indicators
5 - Excellent	[what excellence looks like in plain language]	[specific indicators]
4 - Good	[what good looks like]	[specific indicators]
3 - Adequate	[what adequate looks like]	[specific indicators]
2 - Needs work	[what this looks like]	[specific indicators]
1 - Poor	[what poor looks like]	[specific indicators]

Key: Each level must be distinguishable from adjacent levels by observable criteria.

Step 3c: Structured Comparison

Compare current state against baseline:

Dimension	Before	Now	Direction	Confidence
[aspect 1]	[description]	[description]	↑↓→	H/M/L
[aspect 2]	[description]	[description]	↑↓→	H/M/L

Step 3d: Narrative Assessment

NARRATIVE ASSESSMENT:
Situation before: [describe the starting state]
Actions taken: [what changed]
Situation now: [describe current state]
Key differences: [what's concretely different]
Evidence of change: [specific examples, quotes, observations]
What hasn't changed: [honest acknowledgment]

Step 3e: Indicator Constellation

When no single indicator captures the outcome:

Indicator	Direction	Weight	Current	Notes
[indirect indicator 1]	[desired direction]	[importance]	[observation]
[indirect indicator 2]
[indirect indicator 3]

Interpretation: If most indicators point in the same direction, the outcome is likely moving that way. Divergent indicators suggest complexity.

Step 4: Establish Measurement Rhythm

1. How often to measure? (Daily is too noisy, yearly is too slow — usually weekly or monthly)
2. Who measures? (Self-report + external observation is stronger than either alone)
3. Where to record? (Simple system you'll actually use)
4. When to review? (Regular intervals to spot trends)

Step 5: Guard Against Measurement Failure

Failure	Symptom	Fix
Goodhart's Law	Optimizing the metric instead of the outcome	Use multiple measures, change them periodically
Confirmation bias	Only seeing evidence of improvement	Include "what hasn't changed" in every assessment
Precision illusion	Over-interpreting small differences	Use coarse scales (5 levels, not 100)
Measurement fatigue	Stopping measurement	Make it simple enough to maintain

Step 6: Report

QUALITATIVE MEASUREMENT:
Outcome: [what's being measured]
Approach: [which method(s)]

Current assessment:
[Results using chosen approach]

Trajectory: [improving / stable / declining / unclear]
Confidence: [H/M/L]
Key evidence: [most convincing indicator of direction]
Caveat: [what this measurement can't capture]

Next measurement: [when]

When to Use

• Goal involves subjective or experiential outcomes
• Standard metrics don't capture what matters
• Success is "I know it when I see it"
• Multiple dimensions of quality matter
• → INVOKE: /qr (qualitative research) for deeper qualitative inquiry
• → INVOKE: /dot (outcome tracking) for tracking over time

Verification

• "Better" defined in observable terms
• Measurement approach matched to outcome type
• Scale is coarse enough to be reliable (not over-precise)
• Multiple indicators used (not single metric)
• Measurement failure modes guarded against
• Measurement rhythm established