Behaviorism and the Shaping of the American Mind (part 2)

by Marc-André Cotton—Int. Psychohistorical Association

This article was published in the peer-reviewed publication The Journal of Psychohistory, Vol. 52, No 2, pp. 128-140 (Fall 2024).

Abstract: Historically founded on animal experimentation, behaviorism made withdrawal of parental attention—or ‘time-out’—one of the instruments of its Parent Management Training programs. However, the question of the effectiveness, or even harmfulness of this measure is still being debated. The aim of this second article is to examine the possible side-effects of ‘time-out’ for children’s psycho-affective balance, and the evolution of its social acceptability.

For decades, researchers conducting behavioral experiments with children did not consider it useful to evaluate their psycho-affective impact on the principal parties concerned—children—, except to enhance the supposed effectiveness of their conditioning methods (Parke, 1969). Yet, as early as the 1960s, a parallel stream of research based on the attachment theory of British psychiatrist and psychoanalyst John Bowlby (1907-1990) had produced interesting observations to which behaviorists remained indifferent. A study of the natural socialization of babies in contact with a mother sensitive to the child’s signals stated, for example: “[Our] findings suggest that a disposition toward obedience emerges in a responsive, accommodating social environment without extensive training, discipline or other massive attempts to shape the course of the child’s development.” (Stayton, 1971, p. 1065) In contrast with the suggestion that babies could be efficiently trained by ‘time-out’ to reduce the frequency of dangerous behaviors (Mathew, 1987), this early study argued that attachment “provides an obvious safeguard against the possible dangers of exploratory behavior.”

Reassuring parents

If they wanted to sell their programs, behaviorists had to continue their work of persuasion. An attempt to test the acceptability of behavioral treatments with undisturbed children was carried out by Mark R. Dadds’ team at the University of Queensland, Australia (Dadds, 1987). In a first experiment, two groups of ten children aged 4 to 8 were surveyed on the appropriateness of five maternal responses to four reported situations of disobedience: permissiveness, spanking, threat of consequences, being put away in a corner of the room and ‘time-out’ in an isolated room. Surprisingly, permissiveness was rated less acceptable than all the other interventions, which were rated from neutral to very fair on a three-point scale—and the children made little distinction between physical punishment and ‘time-out’. Presumably, these ratings were simply a reflection of the childrearing violence they had suffered in their own families. But the researchers deduced that young children “prefer parents to take active disciplinary measures” in response to misbehavior and saw this as an argument in favor of ‘time-out’.

A second experiment carried out by the same team involved exposing a group of six disturbed children to behavioral techniques—including ‘time-out’—that their parents had learned in prior training and were instructed to apply at home. Prior to treatment, the mothers completed a list of 40 problem behaviors observed in their children and did the same eight weeks later. The ‘time-out’ experience had no influence on the children’s evaluation of this measure. On the other hand, and as expected, the Parent reports indicated that the treatment had lowered their rates of non-compliance. The study’s conclusion was surprising: “Parents can be reassured that this treatment is acceptable to most disturbed children, and that their child is likely to resist being placed in a time-out.” Despite the small sample size and the variability in individual ratings that the study itself emphasized, although another study had previously noted that most discipline measures used by adults in their environment are generally deemed acceptable by children (Wolfe, 1982), Dadds and his team concluded, “Future research in child behaviour therapy needs to continue social validation of treatment procedures, canvassing child consumers grouped by sex, age, and level of social development.”

 

What do children really say?

This projective logic further illustrates the short-sightedness of a pseudo-science in search of a bright future. The acceptability of behavioral treatments would therefore stem from the fact that children know nothing other than physical and psychological violence—such as empathetic listening to their feelings, for instance. With a little persuasion on the part of parents, the isolation measures advocated by Parent Management Training (PMT) programs should appeal to a wider audience and enable their designers to pursue an old ambition: to make them their therapeutic auxiliaries. The first investigation into children’s self-perception and understanding of ‘time-out’ procedures was conducted in 2000 by Christine Readdick of the University of Florida and trauma specialist Paula L. Chapman of the James A. Haley Veterans Hospital in Tampa (Readdick and Chapman, 2000). Observations were conducted in 11 preschools in a North Florida community, where 42 children aged 2 to 4 were interviewed immediately after experiencing this disciplinary measure.

In addition to the children’s feelings and understanding of why they had been isolated, the researchers also asked about the experiences of those who were most often exposed to it. Most expressed very negative feelings about the ‘time-out’ and about themselves, many confiding that they had felt sad and frightened by the procedure. Damage to their self-esteem was evident in the fact that they felt alone, unloved by their teachers and ignored by their peers. Most children were isolated for reasons deemed trivial by observers, such as failing to obey an instruction, and not for the aggressive behavior the measure claimed to target. Fewer than half of them could recall precisely what behavior had earned them the punishment, which also cast doubt on its effectiveness in preventing future transgressions. They were even likely to withdraw into themselves afterwards, or act out in other, even more undesirable ways. But the psycho-affective repercussions of the ‘time-out’ were clearest in children who were subjected to isolation on a routine basis.

Of the sample interviewed, 8 children admitted to being in ‘time-out’ often, one of them declaring that he had been taken there “Lots, maybe a hundred [times].” Visibly ostracized by their educators, they felt more isolated, sad, scared, and unloved than their less-punished peers, confirming the hypothesis that the measure could have unexpected repercussions, particularly for children already presenting behavioral difficulties. For example, one little boy called his caregiver “Meany,” and a little girl cried throughout the entire episode of ‘time-out,’ imploring: “I want my mommy. I want my mommy.” While acknowledging some limitations to their research, the authors summarized:

“Furthermore, it appears that the consequences of time-out, for many young children, may be punitive rather than instructional. Systematic, fine-grained observations of caregiver application of time-out procedures over time, and documentation of children’s attendant responses to, and feelings about, time-out are needed to confirm these preliminary and potentially disturbing findings.” (Readdick and Chapman, 2000)

 

The importance of emotional regulation

In support of these recommendations, a growing body of psychological research has highlighted the importance of family context and parental socialization practices in the development of young children’s emotional competencies, and later their social skills (Eisenberg, 1998). Some parents, for example, believe that expressing negative emotions such as anger is unacceptable, while others consider it desirable to be in touch with one’s emotions and to experience them in a socially acceptable way. The former are likely to insist on controlling them, while the latter will support the child’s ability to regulate them. Although standards of emotional and social competence vary from culture to culture, converging evidence suggests that repression of emotional expression is detrimental to children’s physical and psychological health, as well as to their success (Gross and Levenson, 1997).

To understand the impact of parental socialization practices on children’s emotional and social skills, Nancy Eisenberg and her colleagues in the Department of Psychology at the University of Arizona reviewed a large number of studies on the subject (Eisenberg, 1998). Multiple studies confirmed, for example, that infants whose mothers were receptive to their emotional signals were more serene because they had confidence in the ability of this “security base” to meet their needs (Cassidy, 1994). In preschool and school-age children, parental responses that did not support the expression of negative emotions, even when these were not harmful to others, were associated with harmful consequences for the child. Worryingly, parents who reported high reactivity to their children’s negative emotions also tended to see their children as predisposed to such reactions, prompting further repressive responses on their part.

In another study by Sally R. Ramsden and Julie A. Hubbard, from the University of Delaware, the mothers of 120 10-year-old schoolchildren were tested on their ability to welcome their children’s emotions and its impact on the aggressiveness the latter displayed in class (Ramsden and Hubbard, 2002). They found that parental support had an impact on the emotional regulation that children were able to implement in the school setting, which could neutralize or, on the contrary, promote potential conflicts. Finally, a 2007 review of more than a hundred developmental psychology studies on emotional regulation unambiguously confirmed that the family’s emotional climate, parents’ reactions to their children’s emotions, the way they talk to their children about them and externalize their own emotions, as well as attachment and parenting styles, were all factors impacting their children’s ability to regulate their own emotions. Among other things, the authors made a pertinent remark in the debate about isolating children for punitive purposes: “Longitudinal analyses indicate that in general, parents’ negative reactions to children’s emotions are associated with low quality of social functioning and emotional regulation difficulties.” (Sheffield Morris, 2007, p. 368)

 

Weak methodological rigor

How did behaviorists react to this body of research, which was ultimately more convincing than their own? A critical examination of more recent ‘time-out’ studies suggests that they did not revise their conception of the child’s potential tyrannical nature, nor the means to achieve compliance—despite inconclusive results. A study by Alan E. Kazdin on the treatment of antisocial behavior, for example, begins with a list of extravagant statistics, claiming that “conduct disorder in children represents a major social as well as clinical problem” (Kazdin, 1987, p. 188). The author goes on to suggest that no therapeutic approach has really proved its worth, before highlighting PMT programs, which he would go on to spearhead. Although the effectiveness of these behavioral treatments was evaluated in the short term and based on external observations alone—thus resulting in a confirmation bias as noted in the first part of this article —the author concluded: “Several features make PMT one of the most promising treatments for conduct disorders.”

A meta-analysis, carried out in 1996 by Wendy J. Serketich and Jean E. Dumas, reveals the impasse in which the behavioral approach found itself, whose champions claimed results based on research that did not meet the minimum criteria of scientific acceptability (Serketich and Dumas, 1996). Of a total of 117 studies targeting at least one child behavior that is deemed antisocial, such as aggression, tantrums, or rule breaking, only 22 publications presented usable statistical results. But how could the relevance of these results be assessed, given the small size and variability of the groups of children subjected to these experiments? Using a computational device that would later become commonplace in psychology research—the effect size measure—the researchers found statistical validity in the correlations appearing in these small samples while admitting that the absence of complete data “may positively bias overall effect sizes” and that PMT programs “appeared to be more effective with smaller samples”—suggesting a statistical bias (Serketich and Dumas, 1996, pp. 176 and 180).

Finally, in at least three respects, Serketich and Dumas described the methodological limitations revealed by their meta-analysis as “striking.” Only a small percentage of the available studies had the expected methodological rigor; few evaluated PMT programs against other forms of intervention; and very few did any long-term follow-up. As the researchers noted: “This makes it difficult to draw strong conclusions about the effectiveness and clinical utility of [PMT programs], particularly in regard to [their] ability to generalize beyond the child’s behavioral problems at home.” Emphasizing again the small number of controlled studies, they concluded by calling for further research, more rigorous in its approach and more comprehensive in its statement of findings. Have there been any such studies since? It would appear not.

 

A final argument

However, an attempt was made ten years later by Brad Lundahl, from the University of Utah, with a meta-analysis aimed at comparing behavioral programs with non-behavioral approaches, evaluating their follow-up and isolating variables likely to influence their results (Lundahl, 2006). Using the same statistical computation, Serketich and Dumas examined 63 studies conducted between 1974 and August 2003, Lundahl and colleagues were forced to admit that the impact of behavioral treatments did not differ significantly from their non-behavioral counterparts. Again, according to the effect size criterion, PMT programs would have a moderate impact on children’s behavior immediately after treatment, and even “small in magnitude” one year later. The authors suggested: “Maintenance of child behavior change therefore requires parents to persist with a highly structured approach to child management, a yeoman’s task for a busy parent.” (Lundahl, 2006, p. 100)

Here comes the final argument: if behavior modification programs don’t work, it’s because parents aren’t applying them properly. Thus, echoing elements of the controversy now surrounding ‘time-out,’ a 2011 Australian review by Alina Morawska and Matthew Sanders began by pointing out that “effective use of time-out involves a number of steps” (Morawska and Sanders, 2011). Their authors emphasized the eight parameters defined by MacDonough and Forehand in 1973—a study whose validity was discussed in the first part of this article—and justified the procedure’s acceptability through various satisfaction surveys. Acknowledging that ‘time-out’ could be misused, they concluded by denouncing the influence of popular reality TV shows on childrearing and encouraged parents to seek “evidence-based parenting information” from professionals.

Aimed at pediatricians, a 2014 study also lamented that the information available on the Internet was “largely incomplete, inaccurate and inconsistent” and warned: “Without adequate guidance on the accurate implementation of effective [time-out] procedures, parents of children with behavioral problems may conclude that [the procedure] is ineffective and resort to harsh methods of discipline […].” (Drayton, 2014) The authors then described nine conditions for ‘time-out’ effectiveness and compared them to advice found by 6 search engines on 102 English-language webpages, before concluding: “Overall, the most striking finding is that no webpage included accurate information on evidence-based [time-out] parameters. In other words, the likelihood that a parent would find complete and accurate information about [time-out] by turning to the Internet is near zero.” One might add that a disciplinary measure requiring such variability of parameters to claim questionable effectiveness was itself bound to inspire growing distrust.

 

Is ‘time-out’ bad for children?

Subsequent pro-‘time-out’ reviews would therefore go to great lengths to reaffirm its principles, invariably citing the same studies and issuing the same warnings. In 2015, Lauren Borduin Quetsch and her colleagues at West Virginia University, for example, lamented that “unfounded” arguments against ‘time-out’ were widespread in public discourse, and set about denouncing the “myths” propagated by its detractors—among them the idea that the measure could be harmful to children (Borduin Quetsch, 2015). One argument, however, caught their attention: “A valid concern is that time-out procedures could very well serve as a trigger for previous abuse experiences, particularly those that involved the caregiver becoming physically aggressive during an escalated and coercive discipline exchange.” In response, they simply argued that—from a behaviorist perspective—repeated exposure to consistent ‘time-out’ procedures could mitigate this traumatic reactivation, and even prove “highly therapeutic”.

But the controversy only grew. In a study published in 2019, Mark R. Dadds and Lucy A. Tully from the University of Sydney set out to answer the question of whether treatments including ‘time-out’ could be harmful, particularly in children with traumatic symptoms (Dadds and Tully, 2019). After painting an idyllic picture of their theoretical underpinnings and through convoluted argumentation, they argued that its use was compatible with the latest theories in developmental psychopathology. To the defenders of attachment theory, for example, they retorted that, since the separation was temporary and predictable, the ‘time-out’ did not threaten the child’s security base. To the proponents of emotional regulation, they replied that calm instructions given before ‘time-out’ increased the likelihood of the child self-regulating. As for fears that the procedure might reactivate existing trauma, they invoked the absence of evidence pointing in this direction and concluded “[The] claims that it is harmful should be considered extraordinary, and thus require an extraordinary level of evidence to back them up.” (Dadds and Tully, 2019, p. 805).

 

Ideological debate on ostracism

Likely, however, no “level of evidence” would allow behaviorists to break out of a theoretical framework that had its roots in their reverence for Skinnerian behaviorism and had been reaffirmed by Serketich and Dumas twenty-five years earlier:

“Human behavior is a function of the contingencies of reinforcement and punishment to which individuals are exposed in the course of their daily exchanges with the environment.” (Serketich and Dumas, 1996, p. 172)

The debate thus left the realm of science and took an ideological turn, as is still evident in the discussions surrounding the social exclusion argument. In 2003, a neuroimaging study by Naomi I. Eisenberger and her colleagues at the University of California, Los Angeles, showed that the brain bases of social pain are like those of physical pain (Eisenberger, 2003). In other words, the highly negative feelings expressed by children subjected to ‘time-out’ had a physiological basis, which tended to demonstrate the harmfulness of ‘time-out’ for their psycho-affective balance—a detrimental consequence that behaviorists recognized as being associated with corporal punishment. A more recent meta-analysis examined the effects of ostracism in 120 experimental social psychology studies using the virtual game Cyberball, on the premise that humans are social animals and care a great deal about whether they are included or ostracized by others (Hartgerink, 2015). It confirmed that an episode—however brief—of ostracism constitutes a threat to basic needs, while acknowledging that, in a second stage, this effect can be moderated by reflection.

How did ‘time-out’ advocates react to this research? A previously cited review simply brushed aside the former on the grounds that Eisenberger’s study focused on college-age adults isolated by their peers (not children segregated by their parents) and cited an experiment conducted on rats (Borduin Quetsch, 2015). As for the numerous Cyberball studies on social exclusion, they were simply ignored, no doubt because, being a commonly accepted disciplinary strategy, ‘time-out’ could not be equated with ostracism in the eyes of behaviorists. However, this circular logic was to come up against the arguments of another body of research surrounding Adverse childhood experiences (ACEs) and the growing recognition of the epidemiological consequences of traumatic stress for society.

 

Can ‘time-out’ be traumatic?

In an original study of ACEs carried out between 1995 and 1997, 64% of respondents reported having been victims of at least one category of Adverse childhood experiences, ranging from psychological, physical, or sexual abuse, to living with drug-addicted, mentally ill, or suicidal parents (Felitti, 1998). Like other more recent studies, it established a strong gradual correlation between the extent of their exposure to abuse and the severity of their health in adulthood (Chartier, 2010). It was shown that repeated stress could have an impact on a child’s brain architecture and foster the development of maladaptive social and behavioral skills with subsequent consequences for their health (Oral, 2016). By their advocates’ own admission, the execution of seclusion procedures could involve various outbursts, particularly if parents did not comply with them to the letter. The question then arose as to whether the behavioral programs, or at least some of their components, implied any traumatic risk that would make them fit into a category of ACEs (Canning, 2021).

Unable to directly question the possible harmfulness of a measure whose merits they were otherwise striving to defend, behaviorists were nonetheless forced to acknowledge that ‘time-out’ was causing growing concern for children previously exposed to adversity. A clinical trial was therefore recently conducted by Alex C. Roach and his colleagues at the University of Sydney, with the aim of demonstrating that these fears were unfounded (Roach, 2022). The non-randomized sample included 205 children aged 2 to 9, diagnosed with various behavioral disorders, whose parents had consulted the Child Behaviour Research Clinic in Sydney, directed by Mark R. Dadds and David Hawes. Of these, 156 were divided into two groups with high or low adversity exposure and participated in a Parent Management Training program including ‘time-out’, while 46 were placed on a waiting list as a control group. Various psychometric scales were used before and after treatment to assess both the children’s exposure to ACEs and their behavioral problems.

Although they did not compare the results with a group subjected to non-behavioral treatment, nor determine the specific impact of ‘time-out’ in relation to the other components of the treatment, and although no long-term evaluation was established, the researchers concluded: “This study has shown that children with high adversity exposure display greater reduction in [childhood mental health] problems and internalizing symptoms compared with peers with low adversity exposure.” (Roach, 2022, p. 7). Here again, the desire to demonstrate the validity of ‘time-out’ seemed to outweigh the rigor and objectivity required to undertake such an approach. In the end, it was in the Journal of Applied Behavior Analysis (JABA)—the scientific journal founded by the inventor of ‘time-out’ Montrose M. Wolfe—that the most critical internal reflection on the abuses observed in the practice of this discipline appeared in 2022, in the form of a warning (Rajaraman, 2022).

 

A belated admission of bad practices

The background to this publication is worth recalling. For some years now, the concept of Trauma-informed care (TIC) had been making an impact across all disciplines and had come to play an important role in the development of public health policies (Harris and Fallot, 2001; Goddard, 2021). The reflection led by Adithyan Rajaraman and published by JABA therefore aimed to integrate this new paradigm into the practice of ABA, suggesting that failure to take it into account could not only be detrimental to its public perception, but also to the effectiveness of its procedures. Clearly, this discipline had not yet defined what it meant to be trauma-informed, not least of all because it focused on the immediate family environment. The authors therefore stressed the need to “avoid retraumatizing clients who may have experienced traumatic events” and to recognize that “many current behaviors may be ways of adapting to and coping with past traumatic experiences” (Rajaraman, 2022, p. 44).

To understand the revolution proposed by the Rajaraman study, consider the following quote:

“If a child experienced neglect at home in the form of extended seclusion or isolation, it seems reasonable to assume that well-meaning behavior analysts would consider past trauma and exercise caution in clinical decisions.” (Rajaraman, 2022, p. 45)

The authors remind us that less intrusive behavioral alternatives to ‘time-out’ exist, and could even replace seclusion procedures (Trump, 2019). Special mention is made of people with a diagnosis of intellectual disability or autism spectrum disorder—a group particularly vulnerable to disciplinary abuse—who, due to their severe communication difficulties, present “with a history of trauma that will remain unknown to the service provider” (Rajaraman, 2022, p. 45). To ensure an environment of safety and trust, the study ultimately calls for minimizing intrusive restraint procedures and encouraging client participation—including children—in the choice of behavioral interventions: “Doing so may lead to the development of best-practice guidelines regarding the provision of choice throughout the course of service delivery.” (Rajaraman, 2022, p. 50)

That same year, noting that ‘time-out’ is a form of punishment that neglects the child’s emotional needs and offers no space for communication, a pilot study proposed replacing it with ‘time-in’—a positive parenting technique that instead allows the parent to connect with the child in a warm, loving way (Holden, 2022). 17 mothers of children aged 3 to 5 were trained in the ‘time-in’ approach, then experimented with it in their homes for two weeks before completing an evaluation questionnaire. The participants reported using the ‘time-in’ approach on average once a day, and less ‘time-out’. They were also very confident in using this technique, which encouraged them to reflect, and observe changes in their children. One mother recognized: “I’ve become more aware of my parenting—like looking for my child’s cues, taking more little breaks throughout the day.” (Holden, 2022, p. 249) Yet, although ‘time-in’ has been mentioned in the literature for years, the technique has curiously not been the subject of empirical study. Without contributing directly to the discussion surrounding ‘time-out’, this pilot study finally suggests that a more child-friendly approach is possible. But will its authors be heard?

 

This second part looks at the possible harmfulness of ‘time-out’ for children’s psycho-affective development and its social acceptability. It is important to remember that this assessment cannot be dissociated from the context in which such a question is formulated. The harmfulness of corporal punishment is no longer debated in our countries, although many parents still resort to it. Certain ABA experiments carried out in the past are now considered unethical. Will the same ever be true of ‘time-out’? We have seen the efforts made by its promoters to defend its acceptability, but at the cost of suspicious contortions suggesting a form of indoctrination. They also admit that the measure can be harmful if it is not implemented in compliance with procedures tested in a clinical context. That it can act as a trigger and reactivate existing traumas is no longer in any doubt, which answers the question of whether ‘time-out’ can still be recommended for children with a history of maltreatment.

And what about others? Ignored by the behavioral sciences, studies on the ostracism effect—the act of being excluded or simply ignored by others—are probably the ones that show most clearly that isolation, however brief, is a physiological alarm signaling a threatening loss of security in any individual, a fortiori in a young child. Numerous studies have shown that feelings of belonging, control, self-esteem and existence diminish significantly after an episode of ostracism, as does the ability to regulate emotions. Its impact on antisocial behavior and aggression has also been studied, with ostracized individuals showing less empathy for the suffering of others (Cursan, 2017). These findings should reassure parents who distrust the promises of behavioral sciences, and perhaps invite others to be more attentive to their own feelings and to the distress signals emitted by their children.

*Marc-André Cotton, MA, the President of the International Psychohistorical Association, and an International Member of the Psychohistory Forum, is a teacher, independent scholar, and director of the French website Regard conscient, dedicated to exploring the unconscious motivations of human behavior. He authored the French psychohistorical book Au Nom du père, les années Bush et l’héritage de la violence éducative published in 2014 by L’Instant présent (Paris).

---