
Some of the hardest messages I receive begin with the same conflict: “The apology felt real, but the behaviour came back.” That does not mean the person was foolish for believing it. A powerful apology can bring relief, tenderness and hope. What it cannot do—by itself—is prove what happens next.
Research on apology and repair suggests that accountability and reparative action matter. In everyday relationships, that creates a simple practical distinction:
An apology is a statement. Repair is a pattern of behavior after the statement.
The Apology Test is a 30-day observation tool. The name is intentionally memorable, but this is not a clinical test, a validated relationship scale or a method for diagnosing someone. It is a way to slow down the decision and compare words with behavior.
Do not announce that you are “testing” another person. Do not provoke situations. Simply record what occurs naturally. If observation itself could put you at risk, prioritize safety and speak with a qualified local service.
Before Day 1: Record the apology accurately
Date: ____________________
The specific behavior being addressed:
____________________________________________________________________
What the person said they understood:
____________________________________________________________________
What they accepted responsibility for:
____________________________________________________________________
What repair they offered:
____________________________________________________________________
What change they promised:
____________________________________________________________________
What support or practical step they proposed:
____________________________________________________________________
Do not improve the apology while writing it down. Record what was actually said, including qualifications such as “if you felt hurt,” “but you also…” or “I only did it because…”
The six elements to observe
1. Specificity
Can the person name the behavior and its impact without hiding behind vague language?
Specific: “I read your private messages without permission. That violated your privacy.”
Vague: “I’m sorry for everything.”
2. Responsibility
Do they take responsibility without transferring the cause to your tone, personality, past mistakes or reaction?
An explanation may add context. It should not erase agency.
3. Repair
Do they take a concrete step to address the harm? Repair might include returning money, correcting misinformation, replacing damaged property, restoring privacy, seeking appropriate professional help or accepting a reasonable consequence.
4. Respect for impact
Can they tolerate that you may still feel hurt, cautious or uncertain? An apology does not purchase immediate forgiveness, trust or access.
5. Changed behavior
Does the specific behavior stop—not merely for two calm days, but when stress, jealousy, frustration or disagreement returns?
6. Response to accountability
What happens when you refer back to the agreement? Genuine repair can include discomfort, but it does not require you to pretend the event never happened.
The 30-day tracker
Use one row per relevant event. Empty days do not need entries.
| Day/date | Relevant situation | Behavior observed | Matched the promise? | Response when addressed | Repair action taken |
|---|---|---|---|---|---|
| Yes / Partly / No | |||||
| Yes / Partly / No | |||||
| Yes / Partly / No | |||||
| Yes / Partly / No | |||||
| Yes / Partly / No | |||||
| Yes / Partly / No |
Weekly review
Complete this on Days 7, 14, 21 and 30.
Week ______
What improved without reminders?
____________________________________________________________________
What improved only after monitoring or pressure?
____________________________________________________________________
What repeated?
____________________________________________________________________
Was responsibility maintained, or was the history rewritten?
____________________________________________________________________
Were repair actions completed?
____________________________________________________________________
Do I feel more free to speak—or more responsible for protecting the apology?
____________________________________________________________________
Interruption patterns to notice
The following do not automatically prove manipulation, but they can prevent honest evaluation when they become recurring substitutes for change:
- a second apology every time the first promise is broken;
- gifts or affection with no connection to the specific harm;
- pressure to forgive quickly;
- anger that you have not “moved on”;
- a short period of ideal behavior followed by the same conduct;
- making you responsible for supervising the change;
- changing the subject to your mistakes;
- demanding credit for basic restraint;
- seeking public praise for private repair;
- claiming the apology itself proves the behavior cannot happen again.
Day 30: Compare promise with pattern
Answer from your records rather than from the most recent emotional moment.
- Did the specific behavior stop?
Yes / Partly / No / Not enough opportunity to observe
- Was the person able to name the behavior consistently?
Yes / Partly / No
- Was repair completed without repeated pressure from me?
Yes / Partly / No / Not applicable
- Could I discuss the impact without punishment or reversal?
Yes / Partly / No
- Did the change remain when stress returned?
Yes / Partly / No / Not observed
- Did my freedom, safety or trust increase?
Yes / Partly / No
- What evidence supports my conclusion?
_________________________________________________________________
- What decision follows from that evidence?
_________________________________________________________________
Three possible conclusions
“There is meaningful evidence of repair.”
The behavior stopped or substantially changed, responsibility remained intact, and repair did not depend on constant supervision. This does not force forgiveness or continued contact. It simply records evidence.
“The evidence is mixed.”
Some effort is visible, but important behavior repeats, accountability is unstable or change depends on monitoring. You may need more time, a firmer boundary or appropriate professional support. Do not convert uncertainty into a verdict merely to relieve pressure.
“The apology reduced tension but did not change the pattern.”
The same behavior returned, responsibility was withdrawn, repair was abandoned or retaliation followed accountability. The apology may have been emotionally convincing while remaining behaviorally ineffective.
When 30 days is the wrong framework
Do not wait 30 days to see whether violence, credible threats, stalking, strangulation, forced sex, dangerous driving, weapon threats or serious child endangerment will recur. High-risk behavior calls for immediate professional safety guidance, not a relationship experiment.
Thirty days is also not a universal deadline for rebuilding trust. Serious harm may require much longer evidence, and no one owes restored access because another person behaved well for a month.
How this resource was created
The Apology Test is an original Blueprint Breached educational tracker developed by Nadeem Riaz from independent research and repeated community questions about apologies that feel convincing but do not produce lasting change. It is not a validated clinical instrument.
Sources and further reading
- PubMed Central, Apology and Restitution: The Psychophysiology of Forgiveness After Accountable Relational Repair Responses.
- Scientific Reports, Experimental evidence that apologies promote forgiveness by communicating relationship value.
- Association for Psychological Science, The Psychology of Offering an Apology.
—
This article is for educational purposes and does not provide medical, mental-health or legal advice. Only a qualified professional can diagnose a mental-health condition.