When Support Depends on the Adult: What Inconsistent Behaviour Might Actually Be Telling Us

There is a particular conversation that seems to happen in schools whenever a child behaves very differently depending on who is supporting them. One adult describes a child who is cooperative, engaged and relatively easy to support, while somebody else describes refusals, arguments, withdrawal or behaviour that can become extremely difficult to manage. Eventually somebody says something along the lines of, “Well, they are fine with me.”

It is an understandable observation. If the same child can complete a lesson with one teacher but refuses with another, or accepts a boundary from one member of staff while becoming distressed when somebody else gives exactly the same instruction, it is difficult not to wonder why. Sometimes the conclusion is that the child knows exactly what they are doing. They are choosing who they behave for, testing certain adults, looking for attention or trying to find the person most likely to give them what they want.

Occasionally, there may even be some truth within parts of those explanations. Children learn very quickly that different people respond differently. Adults do too. We all change aspects of our behaviour depending on who we are with, what has happened previously and what we expect will happen next. A child discovering that one adult negotiates while another does not is not evidence of something unusual or particularly manipulative. It is evidence that they are learning from experience.

What I think becomes problematic is when we notice that behaviour changes between adults and immediately treat that variation as evidence of something about the child's character.

Perhaps the variation itself is the information we should be interested in.

The Same Child, but Not the Same Interaction

Imagine a child who completes maths every Tuesday when supported by one teaching assistant but regularly refuses the same work on Thursdays with somebody else. On paper, very little appears to have changed. It is the same child, the same classroom, perhaps even the same type of work. It becomes tempting to conclude that because the child can do it on Tuesday, they are choosing not to do it on Thursday.

But that conclusion depends on the assumption that changing the adult changes nothing important.

In reality, an interaction between a child and an adult contains hundreds of small variables that rarely appear in a behaviour plan. One adult may give an instruction and immediately begin explaining it again, while another gives the child ten seconds to process what has been said. One might naturally use fewer words. Another might ask several questions while the child is already becoming overwhelmed. One adult may notice the child's face change before behaviour escalates and quietly reduce the pace, while somebody else may understandably continue because nothing obvious has happened yet.

Tone matters. Timing matters. Familiarity matters. Expectations matter. So does what happened yesterday, last week or the last time the child struggled with that person.

None of this necessarily means one adult is good and the other is bad.

That distinction is important because conversations about relationships can very easily become conversations about staff competence. If a child appears more regulated with one adult, we can idealise that person and imply that everybody else simply needs to behave more like them. Equally, if a child shows their most difficult behaviour with a particular member of staff, that adult can begin to feel that they are somehow failing.

Human relationships are more complicated than that.

Sometimes the adult experiencing the greatest difficulty is carrying the hardest part of the child's day. They may be supporting the lesson the child finds most difficult, managing the transition they most dislike or maintaining a boundary that other adults have unintentionally allowed the child to avoid. Sometimes they are simply the person present when everything that has accumulated throughout the day finally becomes too much.

If we want to understand why behaviour changes between adults, we therefore have to look beyond the adult's identity and examine the interaction itself.

Children Learn People

From a behaviour analytic perspective, there is nothing particularly mysterious about children responding differently to different people. Behaviour develops through experience, and people become part of the context in which that learning occurs.

A child learns which adults give them time to answer and which adults repeat the question quickly. They learn who tends to notice when something is becoming difficult and who expects them to say so explicitly. They learn whose “not now” genuinely means later and whose “later” usually means the thing never happens. They learn who follows through, who changes their mind, who becomes louder when frustrated and who remains relatively predictable.

Most of this learning is not conscious in the way we sometimes imagine it to be. The child is unlikely to be sitting in maths conducting an analysis of each member of staff. They are simply responding on the basis of thousands of previous interactions, in much the same way all of us do.

We behave differently with different people because relationships contain histories.

I think this becomes particularly important for children who have experienced repeated misunderstanding from adults. A child who has learned that saying “I don't understand” results in the instruction simply being repeated more loudly may eventually stop saying it. Another child who has discovered that asking for a break is interpreted as avoidance may wait until they can no longer manage before trying to leave. With an adult who has responded differently in the past, those same children may communicate much earlier.

From the outside, it can look as though the child has decided which adult deserves good behaviour.

Another possibility is that the child has learned which forms of communication are likely to work with which people.

That is a very different thing.

Consistency Is More Complicated Than Everybody Doing the Same Thing

Schools understandably value consistency. Children generally benefit from knowing what adults expect and what is likely to happen next, and staff need enough shared practice that expectations do not change completely from classroom to classroom.

But I sometimes wonder whether we have confused consistency with uniformity.

If consistency means every adult using exactly the same words, following exactly the same sequence and delivering exactly the same consequence regardless of context, we may create something that looks consistent on paper while feeling remarkably inconsistent to the child.

Two adults can read the same script differently. The words might be identical while the pace, tone, facial expression and timing are completely different. One adult may deliver an instruction while standing over a child and another may say the same thing from across the table. One gives processing time and the other fills the silence. One notices that the child is becoming distressed and pauses, while the other continues following the agreed plan because consistency has been emphasised.

This does not mean plans are pointless or that adults should simply respond however they feel in the moment. In fact, unpredictability between adults can create significant difficulties of its own. A child should not have to work out an entirely new set of rules every time the member of staff changes.

Perhaps what we need is consistency around principles rather than perfect sameness in delivery.

The child should know that adults will listen when they communicate difficulty. Boundaries that genuinely matter should remain recognisable across people. Agreed adjustments should not disappear because a particular member of staff is absent. Adults should broadly understand what escalation looks like for that child and what helps before it reaches that point. Repair should happen when interactions go wrong.

Within that consistency, however, there still has to be room for human judgement.

What Does “They Are Fine With Me” Actually Tell Us?

I think “they are fine with me” becomes most useful when we treat it as the beginning of an investigation rather than the end of one.

If a child consistently manages better with a particular adult, there is something potentially valuable to learn. But we have to resist assuming we already know what that something is.

\Perhaps that adult has developed an exceptionally strong relationship with the child. Perhaps they communicate more clearly. Perhaps they naturally provide more processing time or recognise subtle signs of difficulty earlier. Perhaps the child has spent much longer with them and familiarity itself reduces uncertainty.

But there are less comfortable possibilities worth considering too.

Perhaps the child behaves differently because that adult places fewer demands on them. Perhaps tasks are being completed for the child rather than with them. Perhaps expectations that other adults maintain have gradually disappeared because avoiding conflict makes the day easier. Perhaps the child appears calmer but is becoming overly dependent on one particular person.

A child being calm with an adult does not automatically prove that the adult's approach is the right one.

In the same way, a child becoming distressed with an adult does not automatically prove that the adult is doing something wrong.

This is why observation matters. What specifically happens differently? How are instructions given? How quickly are demands presented? What happens when the child hesitates? What does the adult do when the first signs of frustration appear? How much support is being provided? Does the child actually complete more independently with one adult, or do they simply appear happier?

Once we ask those questions, “they're fine with me” becomes useful data rather than professional competition.

When the Trusted Adult Sees the Hardest Behaviour

There is another assumption that deserves examining because it sits almost at the opposite end of the argument. We sometimes hear that children show their most difficult behaviour with the adults they trust most because they finally feel safe enough to let everything out.

There are certainly situations where this appears to happen. Families often describe children holding themselves together throughout school and becoming distressed once they reach home, and similar patterns can occur within relationships at school. A familiar adult may become the person a child seeks when they can no longer manage, which means that adult naturally witnesses behaviour other people rarely see.

But I think we should be cautious about romanticising this.

Not every child who repeatedly becomes distressed with a particular adult is doing so because they feel exceptionally safe with them. Sometimes the relationship itself is difficult. Sometimes the adult unintentionally contributes to escalation. Sometimes expectations are unclear or communication styles simply do not work well together.

If we automatically interpret difficult behaviour as evidence of trust, we risk becoming just as incurious as when we interpret it as manipulation.

Again, the behaviour is asking us to investigate rather than assume.

There is something quite important in accepting that adults can be part of the context without being blamed for the behaviour. Those two ideas are often treated as though they cannot coexist. The moment we ask whether an adult's tone, timing or response contributed to an escalation, people can understandably feel criticised.

But examining our contribution to an interaction is not the same as accepting fault for everything that happened.

Children influence adults and adults influence children. That is what relationships are.

The Problem With Making One Adult Indispensable

When a child develops a particularly strong relationship with one member of staff, there can be an understandable temptation to protect it at all costs. If that adult can get the child into school, help them engage with learning or support them through difficult moments, keeping them close feels sensible.

In the short term, it often is.

The longer term question is what happens when that person is absent.

Staff become ill. Teaching assistants change roles. Teachers move schools. Children move classes. If a child's ability to access education depends almost entirely on one particular adult being present, what initially looked like successful support can become surprisingly fragile.

This is where generalisation genuinely matters, but I think we sometimes approach it backwards. We remove the trusted adult or deliberately change things because we want the child to become less dependent, then expect them to transfer skills into relationships that have not yet developed the conditions in which those skills were learned.

A better starting point may be to understand what the successful relationship is providing.

If the child responds well because that adult uses fewer words, other adults can learn from that. If processing time matters, it can become part of shared practice. If the adult notices particular early signs of distress, that knowledge can be communicated. If the relationship itself is important, other adults may need opportunities to build relationships with the child before they become responsible for managing the hardest moments.

The goal should not be to make every adult identical to the preferred adult. That would be impossible and probably undesirable.

The goal is to make good support less dependent on luck.

What We Ask of Staff Matters Too

There is another side to this conversation that I think can get lost when we focus entirely on what children need.

Adults are people as well.

A member of staff who has been shouted at, hit, spat at or repeatedly rejected by a child does not simply reset emotionally because the behaviour plan says they should remain calm. They may understand perfectly well that behaviour communicates something and still feel anxious when they see the first signs of escalation the following day.

That anxiety can then change the interaction. Perhaps instructions become slightly firmer. Perhaps the adult intervenes earlier because they are worried about what happened last time. The child notices the change, responds to it, and both people gradually become part of a pattern neither intended to create.

If we expect staff to provide calm, thoughtful and predictable support, then staff themselves need somewhere to think about difficult interactions without being shamed for having human responses to them.

Reflective practice matters because relationships can become stuck.

Sometimes the most useful question after an incident is not simply what the child should do differently next time. It is what everybody involved might do differently, including the systems around them.

That is not lowering expectations for children.

It is recognising that adults have considerably more power to change the conditions in which the next interaction takes place.

From Consistent Adults to Consistent Experiences

The more I think about consistency, the less convinced I am that the goal should be children behaving exactly the same way with everybody.

None of us does that.

We speak differently to our closest friend than we do to our manager. We ask some people for help more readily than others. Certain personalities make us feel relaxed while others require more effort. We are more willing to admit uncertainty around people whose responses we trust.

Children are no different in that respect.

What schools can reasonably work towards is reducing the extent to which accessing support depends on happening to encounter the right adult at the right moment.

A child should not receive an agreed adjustment only when the teaching assistant who understands them is working. They should not be allowed processing time in one classroom and repeatedly rushed in another. Communication that is recognised as distress by one adult should not automatically be treated as defiance by somebody else.

That kind of consistency does require shared understanding.

But shared understanding is different from everybody becoming the same person.

Good teams talk about what they notice. They compare experiences without turning those comparisons into competitions. If one adult has found something that works, the question becomes what can be learned from it rather than why everybody else cannot control the child as effectively.

Equally, if one adult is consistently experiencing greater difficulty, the response should not automatically be that they need to build a better relationship. Perhaps they do. But perhaps they are being asked to maintain expectations that other people are not. Perhaps they are working with the child at a particularly difficult point in the day. Perhaps they need more support themselves.

Context matters for adults too.

A Final Reflection

When a child's behaviour changes dramatically depending on the adult in front of them, I do not think our first task should be deciding who is responsible.

The variation itself tells us that something changes when the interaction changes.

Sometimes that difference will be relational. Sometimes it will be behavioural. Sometimes it will involve communication, familiarity, expectations, processing time or the child's history with that person. Often it will be several of those things at once.

The difficult part is resisting explanations that feel satisfying because they are simple.

“They know who they can get away with it with” is simple.

“They only behave like that because they feel safe with you” is simple too.

Human relationships rarely are.

Perhaps the better response when somebody says, “They're fine with me”, is genuine curiosity. What happens with you that might be helping? What are you doing before the child becomes distressed? What expectations are you maintaining? What support are you providing? What does the child do when they do not understand? What happens after something goes wrong?

Then we can ask the same questions about the interactions that are not going as well.

The purpose is not to identify the best adult in the room. It is to understand enough about the child's experience that good support does not depend entirely on which adult happens to be standing beside them.

Children do not need every adult to have the same personality, use the same tone of voice or build identical relationships with them. Nor is it realistic to expect children to respond identically to every person they meet.

What they do need is something more fundamental. They need the important parts of their support to survive changes in people. They need adults who share information, reflect on what works, maintain meaningful boundaries, recognise communication and remain willing to repair relationships when things go wrong.

Perhaps that is what consistency should mean.

Not creating identical adults or demanding an identical child, but building an environment in which neither the child's success nor their access to understanding depends entirely on who walks through the door that morning.