The Five-Factor Model: How Clinicians Actually Diagnose Parental Alienation

Love Over Exile is a plain-language research archive on parental alienation, written by Malcolm Smith — an alienated parent and author of the forthcoming book Love Over Exile — for non-specialist readers (other alienated parents, family members, therapists, lawyers) who want to understand the evidence base without a psychology degree or a journal subscription. This page is one entry in that archive.

What the Researchers Asked

When a child refuses to see a parent, the question every clinician, lawyer and judge then has to answer is disarmingly simple: why?

There are many reasons a child might reject a parent. Some are legitimate — abuse, neglect, a genuinely frightening parenting style. Some are developmental — a normal preference in adolescence, a phase of autonomy-seeking, a temporary reaction to a new partner. Some are coping mechanisms — avoidance of loyalty conflict, anxiety that attaches to the whole situation of separation rather than to one parent. And some are the product of influence — one parent systematically reshaping the child’s view of the other until the child will no longer see them.

These reasons are not interchangeable, and the intervention required for each is different. A child rejecting a genuinely abusive parent needs protection, not reunification. A child rejecting a parent under another parent’s coercive influence needs the opposite: swift, firm, professionally-guided restoration of contact. Get the answer wrong and the harm compounds.

The problem the 2022 paper addresses is that, until recently, the field lacked a shared, structured framework for answering the question. Clinicians used Gardner’s 1985 list of eight behavioural signs in the child; but Gardner’s list said nothing about the rejected parent’s history or the favoured parent’s conduct. It could identify a rejected-parent pattern without distinguishing an alienated child from an estranged one. Courts, asked to adjudicate between competing expert opinions, had no common vocabulary against which to measure them.

The question William Bernet and Laurence Greenhill set out to answer in their JAACAP commentary is this: what is the minimum set of criteria an evaluator should check before concluding that a child is parentally alienated, rather than refusing contact for some other reason? The answer is the Five-Factor Model.

What They Did — Methods in Plain English

This paper is a commentary, not an empirical study. The Journal of the American Academy of Child & Adolescent Psychiatry prints the word “COMMENTARY” at the top of the article itself — a signal that the authors are proposing a clinical framework, not reporting new data.

That distinction matters, because a commentary is a different kind of evidence from a controlled trial. A controlled trial tests a hypothesis against data. A commentary synthesises existing data and existing theory into a proposal the field is asked to adopt, refine, or reject. The strength of a commentary rests on three things: the credibility of its authors, the coherence of its argument, and the rigour of the work it cites.

On credibility, the paper is unusual. William Bernet is the most-published author on parental alienation alive today. He trained at Harvard Medical School, is triple-boarded in general psychiatry, child and adolescent psychiatry and forensic psychiatry, is Professor Emeritus at Vanderbilt University School of Medicine, and founded and now presides over the Parental Alienation Study Group — a professional society of around 220 clinicians and lawyers in 32 countries.

Laurence Greenhill is a past president of the American Academy of Child & Adolescent Psychiatry, which publishes this journal, and principal investigator of NIMH’s MTA Study on ADHD treatment. His contribution to the commentary, per the paper’s own author-contribution note, was editorial — “Writing — review and editing” — not conceptual.

Read the paper as a Bernet paper co-signed by a senior child-psychiatry editor whose role is to vouch for the editorial standard, not to add new PA research expertise.

On argument, the paper does three things in four pages. First, it defines parental alienation: a mental state in which a child — usually one whose parents are engaged in a high-conflict separation or divorce — allies strongly with one parent and rejects a relationship with the other parent without a good reason. Second, it sets out the five factors that must typically all be present before a diagnosis is warranted. Third, it explains how the model slots into the existing diagnostic codes in DSM-5 and ICD-11, even though parental alienation itself is not in either manual as a named disorder.

On the work it cites, the paper draws on four kinds of prior evidence:

• Prior conceptual work — Gardner’s 1985 paper, Bernet’s own 2020 article in Feedback that introduced a Four-Factor Model, and the 1,200-page Lorandos and Bernet textbook Parental Alienation — Science and Law.
• Instrument-validation studies — most importantly Baker’s 2020 reliability study of the Four-Factor Model (68 mental-health professionals rating 16 case vignettes), the PARQ-Gap paper (a measurement instrument that distinguished 45 severely alienated from 71 non-alienated children with 99% accuracy in its validation sample), and the Baker Alienation Questionnaire (96% classification accuracy in its own validation work).
• A 2021 custody-evaluator survey — Bernet, Baker and Adkins surveyed 119 custody evaluators and found 85% agreed or strongly agreed with the FFM’s definition.
• Epidemiology — Harman, Leder-Elder and Biringen’s 2019 US prevalence study, which we cover in its own article elsewhere in this cluster.

No new data are introduced. The authors concede as much: “more research needs to be done to further strengthen the reliability of the FFM.” What the paper contributes is structure. It proposes a decision framework the field can use today, and against which future empirical work can be measured.

One other thing the paper is explicitly not is a re-branding of Gardner’s Parental Alienation Syndrome. Bernet and Greenhill note that most writers have abandoned the word syndrome — partly because PA is not listed as a syndrome in DSM-5 or ICD-11, and partly because the “syndrome” label invited dismissal from critics who read it as medicalising conflict. The five-factor model keeps the eight behavioural signs Gardner identified in the child (these become Factor 5), but adds criteria Gardner’s list lacked. That additive move is the paper’s core contribution.

Three frameworks compared — Gardner (1985), Four-Factor (Bernet 2020), Five-Factor (Bernet & Greenhill 2022)

The Five-Factor Model did not appear in isolation. It is the third generation of a clinical framework that has been refined for nearly four decades. Reading the three side by side makes the FFM’s specific contribution visible.

	Gardner (1985)	Four-Factor Model (Bernet, 2020)	Five-Factor Model (Bernet & Greenhill, 2022)
Where published	Academy Forum	Feedback (Family Therapy Association of Ireland)	J. Am. Acad. Child & Adolesc. Psychiatry
Type	Clinical observation	Diagnostic framework	Diagnostic framework + DSM-5 placement
Focus	Behavioural signs in the child only	Adds criteria about both parents	Adds explicit threshold — actual contact refusal
What it measures	The 8 behavioural signs	(1) Prior positive relationship · (2) Absence of abuse by rejected parent · (3) Multiple alienating behaviours by favoured parent · (4) Gardner’s 8 signs	(1) Contact refusal · (2) Prior positive relationship · (3) Absence of abuse · (4) Alienating behaviours · (5) Gardner’s 8 signs
Differential diagnosis	None — child-only lens	Estrangement vs alienation rule-out	Estrangement vs alienation rule-out + threshold gate against over-diagnosis
Diagnostic placement	”Parental Alienation Syndrome” (proposed term, never adopted in DSM/ICD)	No DSM/ICD placement specified	Slots into existing codes: CAPRD (Z62.898), Parent-Child Relational Problem (Z62.820), Child Psychological Abuse, ICD-11 QE52.0
Status today	Foundational but no longer used alone	Largely superseded by FFM	Standard framework — 85% expert agreement (n=119 evaluators, 2021)

What They Proposed — The Five Factors

The decision tree at a glance

The five factors are not a flat checklist. They form a decision tree — a threshold gate, two rule-outs, and two positive-evidentiary criteria — walked in order. Each factor has a different role in the framework.

The five factors are not a simple checklist where each item scores one point and four out of five is good enough. They fall into three functional roles.

Factor 1 is a threshold gate — if it is not present, there is nothing to diagnose. Factors 2 and 3 are differential-diagnosis rule-outs — they test whether the refusal is something other than alienation. Factors 4 and 5 are the positive evidentiary core — what the favoured parent does, and what the child then shows.

A responsible evaluator walks the factors in that order, not as a checklist but as a decision tree. Each sub-section below explains one factor, what it is doing in the model, and where it typically goes wrong in practice.

1. Factor 1 — The child manifests contact resistance or refusal

This is the factor the 2022 paper adds to Bernet’s earlier Four-Factor Model, and it is the most important guardrail in the whole framework. Without observable contact refusal by the child, the FFM does not apply. A parent who feels undermined by the other parent, who sees concerning behaviours, who fears alienation is coming — that parent may be right, but without a child who is actually refusing contact, the diagnostic category is alienating behaviours (Factor 4 without the rest of the model), not full parental alienation.

The phrasing is deliberate: “contact resistance or refusal”. Resistance is the earlier, milder end of the spectrum — reluctance, withdrawal, episodes of distress before or after visits. Refusal is the harder end — the child will not attend at all, refuses communication, treats the rejected parent as someone they do not recognise. Both count. What is required is that the rejection is coming from the child’s own observable conduct, not the alienating parent’s report.

This matters because, before the 2022 paper, a version of the FFM existed in which contact refusal was implied but not explicit. The 2022 revision makes it explicit precisely to stop clinicians diagnosing “alienation” where the child is still in regular contact with both parents. The point is not to expand what counts; the point is to be honest about the threshold.

2. Factor 2 — A prior positive relationship existed between the child and the rejected parent

The second factor is a differential-diagnosis rule-in. Its purpose is to distinguish a child rejecting a parent they once loved — which is the pathology — from a child rejecting a parent they never warmly bonded with, which is a different clinical picture and a different intervention.

The evidence base for “prior positive relationship” is not the word of either parent. By the time a case reaches an evaluator, each parent has a narrative, and those narratives are often mirror-image.

Factor 2 asks for collateral evidence rather than either parent’s word. The kinds of evidence that count are:

• photographs and home video from the period when the relationship was still intact
• written correspondence (letters, cards, messages) showing warmth between child and parent
• third-party reports from people who saw them together — teachers, babysitters, family friends, clergy, prior therapists

A pattern of warmth documented before the separation, seen by multiple people, is the anchor.

Where Factor 2 fails in practice is in the edge case where a parent and child had little pre-separation contact — the parent was absent, or travelled constantly, or the child is very young. A strict reading of Factor 2 can then be used to argue the FFM does not apply. Critics reasonably note that one mechanism of alienation is precisely to prevent the warm pre-separation bond from forming in the first place. The paper does not resolve this tension cleanly, and an honest evaluator will hold the factor alongside the developmental context of the particular case.

3. Factor 3 — Absence of abuse, neglect, or seriously deficient parenting by the rejected parent

This is the factor that does the most important work in the whole model, and the factor critics pick apart most vigorously. It is the rule-out that distinguishes alienation from estrangement.

The distinction is this. A child who rejects a parent who genuinely abused, neglected, or seriously mis-parented them is an estranged child. The rejection is proportionate to what the parent did. The correct intervention is protection, not reunification.

A child who rejects a parent who did not do these things, under the systematic influence of the other parent, is an alienated child. The rejection is disproportionate. The correct intervention is careful, professionally-guided restoration of the relationship.

Factor 3 requires the evaluator to actually do this work. That means reading medical records, child-protection files, police reports. It means taking seriously any allegation of abuse, investigating it, and only ruling it out when the evidence genuinely does not support it. It means looking at proportionality — is the child’s intensity of rejection commensurate with what the parent actually did?

Critics (Garber and Simon in 2023, and Joan Meier in a longer-standing critique) argue Factor 3 is where the model is most exposed. “Abuse” and “neglect” are defined differently in different jurisdictions. “Seriously deficient parenting” is not operationalised. And the binary framing — abuse, or not — cannot represent the cases, which are common in practice, where both alienating behaviours and genuine parental shortcomings are present in the same family. The honest answer is that Factor 3 requires evaluator judgment that the model itself cannot fully script.

4. Factor 4 — Multiple alienating behaviours by the favoured parent

Factor 4 is the first of the two positive evidentiary factors. It turns the lens on the favoured parent — the one the child is aligning with — and asks what that parent is actually doing.

The reference catalogue is Baker and Chambers’ 2011 list of seventeen strategies, covered in detail on our Baker’s 17 strategies page. In condensed form, the alienating behaviours include: badmouthing the other parent to the child; limiting contact and sabotaging communication; telling the child the other parent does not love them; creating the impression the other parent is dangerous; asking the child to spy or keep secrets; calling a step-parent “Mom” or “Dad” as if to replace the other parent; withholding medical, school or social information from the other parent; and undermining the other parent’s authority.

Factor 4 requires multiple such behaviours. Not all seventeen. Not even most of them. The paper uses “multiple” deliberately, in the sense of “more than one or two isolated incidents” — a pattern, not an anecdote. The threshold is low for a reason: the catalogue is long because alienation is not a single tactic but a repertoire, and the pattern is what differentiates alienation from the kind of bitterness that is unfortunate but normal in acrimonious separations.

Important differential point that the paper itself makes: alienating behaviours (ABs) are not the same as parental alienation (PA). Citing Harman and colleagues’ 2019 US prevalence study, Bernet and Greenhill note that more than 30% of divorced US parents report being the target of multiple ABs, but only around 1.3% report being moderately or highly alienated from a child. Alienating behaviours are common. Full parental alienation, where the child has actually refused a relationship, is roughly twenty-five times rarer. The presence of Factor 4 alone does not constitute alienation; Factor 1 — actual contact refusal — is required for that.

5. Factor 5 — The child displays Gardner’s eight behavioural signs

Factor 5 is the second positive evidentiary factor and the oldest element of the model. Richard Gardner first described the eight signs in 1985, and the FFM reuses the list almost verbatim. See our dedicated signs of parental alienation page for the full picture; in brief, the eight are:

1. A campaign of denigration — the child speaks persistently and without nuance against the rejected parent.
2. Weak, frivolous, or absurd rationalisations — the child’s stated reasons for rejection, under probing, turn out to be trivial or implausible.
3. Lack of ambivalence — the child splits; one parent is all-good, the other all-bad, with no mixed feelings permitted.
4. The “independent thinker” phenomenon — the child insists they have reached their views on their own, using language and justifications that echo the alienating parent.
5. Reflexive support for the favoured parent in any conflict, regardless of merit.
6. Absence of guilt about cruel or disrespectful behaviour toward the rejected parent.
7. Borrowed scenarios — the child describes events they could not have witnessed, or uses adult language and concepts that do not match their development stage.
8. Rejection of the rejected parent’s extended family — grandparents, aunts, uncles, cousins, often including people the child was warmly connected to.

“Many” of these, per the paper, is the threshold. Not all eight. A responsible evaluator is looking at the constellation — the number of signs present, their severity, and the degree to which they form a coherent pattern rather than scattered symptoms. Factor 5 is where the paper is most-aligned with historical clinical observation, which is both its strength (half a century of observational consistency) and the limitation critics press on (no prospective longitudinal data; self-report questionnaires to validate the signs used children previously classified as alienated by the same signs).

6. How the FFM slots into DSM-5 and ICD-11

The sixth piece of the paper is not a factor but a placement argument — and it is the piece that resolves the longest-running objection to parental alienation, which is that “it is not in the DSM so it is not real”.

Bernet and Greenhill concede the narrow version of that objection openly: “the words parental alienation are not in DSM-5 or the International Classification of Diseases 11th Revision (ICD-11).” That is simply true. An earlier campaign, led in part by Bernet, to include “Parental Alienation Disorder” as a named DSM-5 entry failed in 2013. The 2022 paper does not try to reopen that fight.

Instead, the paper argues the concept is already covered by three existing DSM-5 codes, and one ICD-11 code, that a clinician can already use today:

• Child Affected by Parental Relationship Distress (CAPRD) — DSM-5 code Z62.898. Bernet co-authored the paper that introduced CAPRD as a DSM-5 category in 2016, published in this same journal.
• Parent-Child Relational Problem — DSM-5 code Z62.820.
• Child Psychological Abuse — DSM-5 codes V995.51 / T74.32XA.
• Caregiver-Child Relationship Problem — ICD-11 code QE52.0.

The FFM’s role, in this framing, is the decision framework that assigns a presenting case to the right existing code. A case that meets all five factors and involves sustained coercive control is a Child Psychological Abuse case. A case where the alienating dynamics are subtler may be a CAPRD case. A case where the rejection is real but the evidentiary core is weaker may be a Parent-Child Relational Problem. The clinician is not required to invent a new diagnosis; they are required to use the existing codes with a framework that makes the assignment defensible.

Why This Matters

The 2022 paper is a four-page commentary in a specialist journal. It introduces no new data. Why is it worth a full article on this site?

Because it is the piece of scaffolding that holds the rest of the diagnostic conversation together. Before the FFM, debates about parental alienation were frequently conducted across one another — one clinician would be pointing at Gardner’s eight signs, another would be insisting on the evidence against the favoured parent, a third would be warning that a child’s refusal might be justified estrangement, and none of them had a shared framework in which to integrate all three concerns. The FFM does not resolve the underlying disagreements, but it gives everyone — clinicians, evaluators, lawyers, judges — the same conceptual structure to disagree within. That alone changes the quality of the conversation.

Two downstream effects have followed.

First, the FFM is now the default framework in the field that studies parental alienation. The 2021 custody-evaluator survey Bernet and Greenhill cite found 85% of 119 evaluators agreed with the definition. Professional training in parental alienation assessment — PASG’s continuing-education courses, the training modules used by the AFCC (Association of Family and Conciliation Courts) — teaches the FFM as the standard. It has been translated into multiple languages and is used by evaluators across the US, UK, Canada, Australia, the Netherlands, Italy and Brazil. That adoption matters because it means the therapist or evaluator a parent consults is increasingly likely to be using this framework whether they cite the paper or not.

Second, the FFM has become the hook by which parental alienation enters the existing diagnostic codes. That is a subtle but important shift. Before 2022, a clinician writing a report often had to choose between calling a case “parental alienation” (untethered from any DSM code) or assigning it a vaguer label like “Parent-Child Relational Problem” (a real DSM-5 code, but one that said nothing specific about the mechanism). After 2022, the clinician can assign the DSM-5 code and note that the case meets the FFM criteria. The family court then has something concrete to look at: not a contested diagnostic category, but a defensible assignment to an existing code, supported by a named framework.

For the people this site serves — alienated parents, their families, the professionals who support them — the practical consequence is that when you next find yourself trying to explain what is happening in clinical language, you have a framework that is being used by people who take it seriously. You are not asking them to accept a category they may not believe in; you are asking them to check a case against a five-part framework and see which of the existing codes the case fits.

This is particularly important where institutions are slow to engage. In family court, the FFM provides a structured basis for expert-witness testimony. In safeguarding conversations, Factor 3’s rule-out insists that abuse allegations are investigated rather than dismissed — which is as important for genuinely at-risk children as it is for genuinely alienated ones. In the therapist’s room, the FFM gives the therapist something more than a binary “is this alienation or not” question: they have a structured assessment to run. The paper itself is short; the effect of its adoption across the field is what makes it worth reading.

What This Means for You

The practical weight of the Five-Factor Model depends on who is reading it. Four audiences, four different implications.

If you are a targeted parent. The FFM’s most immediate use for you is as a structured way to describe what is happening. When you next speak to a lawyer, a therapist, a school, or a GP, you are not saying “my ex has turned my child against me” — a phrase that lands as he-said-she-said and invites the listener to remain neutral. You are describing a framework, point by point: factor by factor, with dated examples of each.

Factor 1 — my child now refuses contact, starting from this date. Factor 2 — here are photographs and collateral accounts from this period when the relationship was warm. Factor 3 — I can provide my medical, child-protection and employment records; no abuse findings. Factor 4 — here are documented instances, with dates, of alienating behaviours. Factor 5 — here are the specific signs my child now displays, with examples.

That structure is qualitatively harder to dismiss than a narrative complaint. It also does the work of pre-empting the most common professional reflex, which is to look for an abuse finding that would flip the case into estrangement — Factor 3 names that possibility and addresses it.

If you are a lawyer or therapist reading this. The FFM is the framework your peers are increasingly using. Whether or not you personally accept parental alienation as a primary diagnostic concept, the five factors are a differential-diagnosis checklist that serves any child-custody case where contact refusal is in play.

Factor 3 in particular forces the thing the field has been rightly criticised for skipping: a proper investigation of abuse allegations before alienation is considered. A lawyer or therapist who walks a case through the five factors — in that order — will produce better-reasoned assessments than one who does not.

If you disagree with the model, engage with Garber and Simon’s 2023 critique and with Joan Meier’s longer-standing work. But do not ignore the model; it is the one most of your peers are using.

If you are a family member, friend, or colleague of an alienated parent. The FFM gives you language for something that may have seemed private and formless. When your friend or relative describes what is happening to them, you can listen for the five factors — not as an amateur diagnostician, but as a way of understanding that their experience has a shape, and a name, and a framework used by clinicians who take it seriously. That alone is a kind of support. A parent who feels their situation is being invalidated by the people around them carries a heavier load than one who feels understood.

If you are the alienating parent reading this with discomfort. The FFM is not a courtroom verdict. It is a diagnostic tool a clinician uses. What it may offer you, however, is a mirror more honest than the one you have been looking into.

Factor 4 is the factor that names what you are doing. If, reading through the list of seventeen alienating behaviours, you recognise more than a few — even if you have rationalised them to yourself as “protecting” the child, “being honest” with them, “letting them decide” — the FFM is asking you to notice the pattern.

Factor 5 is the factor that describes what your child is becoming. If you recognise those signs in your own child, the discomfort is information. What you do with it is still in your hands.

What the Paper Doesn’t Tell Us — Limitations

No paper is the final word. A responsible article about a framework has to name what the framework does not cover, and where the genuinely capable critics have pushed back.

First, the authors’ own acknowledgements. Bernet and Greenhill concede that “more research needs to be done to further strengthen the reliability of the FFM.” The empirical base behind the model is distributed across other papers — Baker’s 2020 Four-Factor reliability study, the PARQ-Gap validation, the 2021 custody-evaluator survey — rather than contained in this commentary. Independent prospective validation, in which children and families are followed over time and outcomes compared, is still sparse. The 2021 survey showed evaluators agree with the definition; it does not show the definition reliably identifies cases that go on to behave differently.

Second, the critics, fairly stated. The sharpest published critique of this specific paper is Garber and Simon’s 2023 Journal of Divorce & Remarriage article, “Looking Beyond the Sorting Hat”. Neither author is a general PA denialist. Their objections are structural.

They argue Factor 1’s terminology (contact, resist, refuse) is undefined and that healthy adolescents, autistic children, anxious children and trauma survivors can normatively resist parental contact — so the threshold is broader than the paper admits.

They argue Factor 2 is circular, because alienation itself can prevent a positive prior relationship from forming, which the FFM then uses as a rule-out.

They argue Factor 3 suffers from jurisdictional variability (abuse is defined differently state by state) and ignores vicarious trauma — a child traumatised by witnessing intimate-partner violence against one parent is not directly abused by the other, but the child’s rejection of the witnessing parent is not straightforwardly alienation either.

And they argue the binary framing — alienation or not — cannot represent the “hybrid” cases, common in practice, where alienating behaviours and genuine parental shortcomings coexist. Their proposed alternative is an ecological model that treats parent-child relationships as a system to be assessed in context rather than a sorting hat.

Joan Meier, at George Washington Law, raises a compounding concern at the policy level. Her empirical work — including a 2020 US study of family-court outcomes — finds that alienation claims are disproportionately credited against mothers who raise safety concerns, and that the family-court system is less likely to protect children when alienation is alleged.

That is not a critique of the FFM’s five factors as such; it is a critique of how PA evidence gets used in court. A rigorous Factor 3 investigation is the counter-measure the model itself proposes.

Whether Factor 3 is rigorously applied in practice — across thousands of contested custody cases, with evaluators of variable training — is an empirical question the paper does not answer.

Third, conflicts of interest. The paper discloses that Bernet has received royalties from Charles C Thomas (publisher of the Lorandos and Bernet textbook), is the president of the Parental Alienation Study Group, and has been paid for testifying as an expert witness in child forensic psychiatry. Greenhill declared no conflicts. These disclosures do not disqualify the paper — Bernet is the most-published author in this field, and serious work in any specialist area inevitably generates such affiliations — but readers deserve to see them. They also underscore the importance of engaging with external critics like Garber, Simon, Meier and Mercer: a field in which the leading framework is authored by the president of the field’s main professional society is, by construction, a field in which external scrutiny is particularly valuable.

Fourth, what remains open. Beyond independent prospective validation, the live questions are operational. How finely should “multiple” alienating behaviours in Factor 4 be defined? How should Factor 3 handle hybrid cases where Gardner-type signs coexist with real parental shortcomings? What is the right treatment approach in mild cases where the model’s five factors are partially met but not conclusively so? These are not reasons to set the model aside — it is the best structured framework the field currently has — but they are reasons to hold the model with appropriate humility, and to keep reading the critics as the evidence develops.