When the Plan Isnt Working

When the Plan Isn't Working

How to recognize a stalled behavior plan and surface concerns effectively

For paraprofessionals and the teachers who supervise them

| |

| :-: |

| The frameBehavior intervention plans are not guaranteed to work. Some plans are well-designed but poorly implemented. Some are faithfully implemented but based on an incorrect function. Some worked initially and have since stopped working as circumstances changed. A para who sees that the plan is not producing results has important information -- but surfacing that information effectively, through the right channels and with the right data, is a skill. This brief covers how to recognize when a plan needs revision and how to raise concerns constructively. |

How to tell when a plan is not working

Data-based decision rules

Behavior plans should include explicit decision rules -- criteria that trigger a plan review. Examples: if the target behavior does not decrease by 20% within four weeks of implementation, review the plan; if the behavior increases during any two consecutive weeks, convene the team. If your plan does not have decision rules, ask the teacher or BCBA to establish them.

In the absence of explicit decision rules, general indicators that a plan may not be working:

The target behavior is not changing in frequency, duration, or intensity after three to four weeks of consistent implementation

The target behavior is decreasing but the replacement behavior is not increasing

The student is showing new problem behaviors not addressed by the current plan (behavior displacement)

The plan requires conditions that are not consistently achievable in the current setting

What the data should show

Progress in a well-functioning behavior plan typically follows a pattern: after an initial period of no change or slight worsening (as the extinction burst resolves or the new contingencies are established), behavior begins to move in the expected direction. Flat or worsening data after the initial period is meaningful information.

Track the data exactly as specified in the plan. Inconsistent or estimated data makes it impossible to tell whether the plan is not working or whether the data is simply inaccurate.

Fidelity check before plan revision

Before concluding that a plan needs to be changed, the team must determine whether the plan has been implemented as designed. A plan that has not been implemented with fidelity has not actually been tried. Common fidelity failures:

The reinforcement schedule is not being delivered consistently -- the para forgets, skips, or varies the timing

The planned ignoring is being broken by subtle reactions from one or more adults

The replacement behavior has not been taught -- the plan assumes the student has it, but they do not

The demand fading or graduated access schedule has not been followed -- the para is backing off when the student escalates

Different adults are implementing the plan differently, creating inconsistency

A fidelity check involves a supervisor or BCBA observing the implementation and scoring it against the written plan. If fidelity is below approximately 80%, the plan has not been given a fair test. Improve fidelity first, then evaluate effectiveness.

How to surface concerns

What to bring

When you have concerns about a behavior plan, bring data -- not impressions. The difference:

Impression: I don't think the plan is working -- he seems the same

Data: I have been recording frequency for three weeks. In week one, there were 12 incidents. In week two, 14. In week three, 11. The behavior does not appear to be changing.

Specific behavioral observations are also helpful: I noticed that he is now also engaging in \[new behavior\] during the times when we are using planned ignoring for \[original behavior\]. This suggests the team should also review whether the replacement behavior has been taught and practiced.

Who to tell

The chain of communication for behavior plan concerns typically runs: para informs the supervising teacher, who convenes the team or contacts the BCBA. In some schools, paras have direct access to the BCBA -- know your school's protocol.

Raise concerns proactively and specifically. Do not wait for a scheduled meeting if you are seeing significant deterioration or safety concerns -- those warrant immediate communication.

How to frame concerns

Frame your concern around the student's outcomes and the data, not around criticism of the plan or its designers. Effective: I want to make sure the plan is adjusted if it needs to be -- here is what I am seeing in the data. Less effective: this plan clearly is not working and never made sense to me. The goal is a productive team response, not validation of your skepticism.

When to escalate to a BCBA or specialist

Escalate immediately when:

The student or others are at risk of injury and the current plan does not appear to be reducing that risk

The behavior is increasing significantly despite consistent implementation

A new behavior has emerged that is more dangerous than the original target behavior

The student is experiencing a significant life change (new living situation, medical issue, medication change) that may require plan revision

Do not wait for a scheduled meeting for safety concerns. A phone call, email, or direct conversation with the teacher or BCBA is appropriate.

Scenarios

Scenario A: The flat data

A para has been implementing a DRA plan for a student's calling-out behavior for five weeks. She has been tracking frequency daily. Her data shows 12-15 incidents per day across all five weeks with no consistent downward trend. She brings her data sheet to the supervising teacher and says: I have been tracking the call-outs since we started the plan. The frequency has been flat -- 12 to 15 per day -- across all five weeks. I wanted to flag this in case the team wants to review. The teacher thanks her and sets up a team meeting with the BCBA. At the meeting, the BCBA observes an implementation session and discovers that the reinforcement is being delivered inconsistently -- the fidelity issue, not the plan design, is the problem.

Scenario B: The new behavior

A behavior plan was developed to address a student's property destruction. After three weeks of implementation, the property destruction has decreased, but the student has begun hitting peers -- a new behavior. The para notes the pattern and brings it to the teacher with her ABC data from the past two weeks showing the new behavior. This is behavior displacement -- the original function is still active but the replacement behavior has not been sufficient. The team reconvenes to address the new behavior within the existing function-based framework.

| | |

| :-: | :-: |

| Try this | Watch out for |

| Track data exactly as specified and bring it when raising concerns -- data is more actionable than impressions | Waiting weeks to raise a concern because you assumed someone else was monitoring |

| Check fidelity before concluding the plan design is wrong -- many plans fail because of implementation, not design | Attributing plan failure to design problems before checking whether the plan has been implemented with fidelity |

| Raise concerns proactively and through the correct channel, not in casual conversation | Raising concerns in a way that criticizes the designer rather than focusing on the student's outcomes |

| Escalate immediately when there are safety concerns -- do not wait for a scheduled meeting | Continuing to implement a plan unchanged when safety-relevant deterioration is occurring |

| |

| :-: |

| Bottom lineThe plan is a hypothesis, not a verdict. Data tells you whether the hypothesis is working. Your job is to implement the plan faithfully, track what happens, and surface what you see. When the data does not support the plan, the team needs that information to help the student. |

Page