Monica Pignotti

Is longer-term psychodynamic psychotherapy more effective than shorter-term therapies? Review and critique of the evidence.

Background: In 2008, Leichsenring and Rabung performed a meta-analysis of 8 studies of longer-term psychodynamic psychotherapy (LTPP). The work was published in the Journal of the American Medical Association (vol. 300, pp 1551–1565), and they concluded that LTPP was more effective than shorter-term therapies. Method: Given that such claims have the potential to influence treatment decisions and policies, we re-examined the meta-analysis and the 8 studies. Results: We found a miscalculation of the effect sizes used to make key comparisons. Claims for the effectiveness of LTPP depended on a set of small, underpowered studies that were highly heterogeneous in terms of patients treated, interventions, comparison-control groups, and outcomes. LTPP was compared to 12 types of comparison-controls, including control groups that did not involve any psychotherapy, short-term psychodynamic psychotherapy, and unvalidated treatments. Additionally, the studies failed to protect against threats to bias, and had poor internal validity. Conclusion: Overall, we found no evidence to support claims of superiority of LTPP over shorter-term methods of psychotherapy.

Some comments on "Energy psychology: A review of the evidence": Premature conclusions based on incomplete evidence?

A review of the evidence on energy psychology (EP) was published in this journal (see record 2008-07317-008). Although Feinsteins stated intention of reviewing the evidence is one we support, we noted that important EP studies were omitted from the review that did not confirm the claims being made by EP proponents. We also identify other problems with the review, such as the lack of specific inclusion and exclusion criteria, misportrayal of criticism of EP, incorrectly characterizing one of the studies as a randomized clinical trial, and lack of disclosure regarding an EP-related business. We note that in the American Psychological Association, decisions on classification of therapies as empirically supported are most rightfully the function of Division 12-appointed committees of psychologists. It is not enough for any one individual or group of proponents of a particular approach to make such a determination. (PsycINFO Database Record (c) 2010 APA, all rights reserved).

Thought Field Therapy: A Former Insider's Experience

Thought Field Therapy (TFT) is a novel therapy that employs finger tapping on purported acupressure points. Over the past decade, TFT, promoted on the Internet and through testimonials of fast cures, has gained popularity with therapists, including clinical social workers. Although TFT claims to cure a wide variety of psychological and physical problems, there is scant evidence to support such claims. The following is an account of my 7-year experience as a leading practitioner, author, and teacher of TFT and includes my initial skepticism, what first interested me, my experiences training to the highest level of TFT (the proprietary Voice Technology) and becoming part of the inner circle of TFT, an experiment I did with Voice Technology, and factors that led to my doubts and ultimate disillusionment with TFT. The pseudoscientific aspects of TFT and how they can impair critical thinking are also discussed.

Holding Therapy and Dyadic Developmental Psychotherapy Are Not Supported and Acceptable Social Work Interventions: A Systematic Research Synthesis Revisited

This article re-examines material discussed in a recent systematic research synthesis by Craven and Lee. The authors find that two of the interventions for foster children discussed by Craven and Lee were erroneously classified as supported by evidence of efficacy, and one, holding therapy, is shown to be potentially physically harmful to children. Detailed evidence is offered to show that holding therapy and dyadic developmental psychotherapy are not appropriately categorized as supported and acceptable interventions.

Science and Pseudoscience in Developmental Disabilities: Guidelines for Social Workers

Individuals with a developmental disability can now be provided a variety of empirically supported treatments that have been shown to be useful in promoting educational attainments, social and vocational skills, and self-care, and in reducing behavioral problems. Unfortunately, a large number of pseudoscientific or bogus therapies continue to be offered to this population and their families. We review the characteristics of pseudoscientific and bogus treatments and provide several examples of unsupported or harmful interventions offered by contemporary social workers and other human service professionals, to the detriment of people with disabilities. We encourage social workers to identify pseudoscientific interventions and avoid providing these, in favor of using empirically supported treatments.

The Problem of Pseudoscience in Social Work Continuing Education

ABSTRACT The National Association of Social Workers requires social workers to obtain continuing education (CE) after they receive their social work degrees. A large and profitable industry that has emerged catering to this need for CE that is particularly focused on the needs of licensed social workers. Quality control mechanisms in place to monitor the content and delivery of CE approved by the field of social work is relatively lax, and as a result a considerable number of social work CE programs convey content that can be labeled as pseudoscientific. We provide illustrations of currently approved social work CE courses that provide training in a variety of bizarre and unsupported assessment methods and treatments that are pseudoscientific.

Missed Opportunity to Rectify or Withdraw a Flawed Metaanalysis of Longer-Term Psychodynamic Psychotherapy

as they described in their letter, merely ‘suboptimal’ or ‘ambiguous’, but wildly inaccurate. We provided a power analysis of the individual randomized trials entered into their metaanalysis, appropriately calculated not on the basis of the overall sample size, but on the smaller of the intervention or comparison groups. None of the smaller groups had over 30 patients, 1 had only 10. Perhaps we should have been clearer on the implications: if LTPP indeed had a moderate effect, none of these studies had an over 50% chance of detecting it, and 1 had a less than 25% chance of detecting it. Thus, even if all of the studies tested a moderately effective form of LTPP, most should not have obtained a significant effect. That most had significant findings attests to the likelihood of publication bias. Leichsenring and Rabung claimed to have ruled out publication bias, a pervasive problem in the psychosocial intervention [5] and medical [6] literatures. They entered their miscalculated effect sizes into calculation of a failsafe-N estimate, i.e. how many studies with null results would have to be left unpublished to unseat their conclusion concerning the superiority of LTPP to shorter-term therapies. Many sources including the authoritative Cochrane Handbook [7] recommend against acceptance of the failsafe-N. Nonetheless, here as elsewhere, Leichsenring and Rabung and prepublication reviewers should have been alerted by the implausible estimate that 921 null studies had to be reclaimed from file drawers in order to unseat claims based on 8 modestsized randomized trials and 23 observational studies. Leichsenring and Rabung used summary Jadad [8] scores to rate the quality of studies entered into their metaanalysis. Yet, the Jadad checklist was developed to rate the quality in which a study is reported, not the likelihood that study results are free of bias. Moreover, summary scale scores have been empirically shown to lack validity [9] . We used Cochrane Collaboration guidelines to rate the studies and found that they lacked internal validity and the usual safeguards against bias expected of quality evidence. For instance, only 1 of 8 studies met the important criterion of being low in bias toward completeness of outcome data. Such flaws can prove particularly decisive in studies with small sample sizes, serving to undo the benefits of the study being randomized. Leichsenring and Rabung are correct that we did not consider the quality of the 23 observational studies included in their larger metaanalysis. But if we had, the quality ratings would have been even lower. Some of the observational studies lost substantial numbers of patients to follow-up [10] , with results for the final sample not being generalizable to the initial sample. It is naïve to assume that combining such similarly small and flawed studies into a metaanalysis can produce meaningful conclusions [11] . Leichsenring and Rabung included studies in their metaanalysis that were clinically heterogeneous in terms of comparison treatments and patient populations. They claimed to have demonstrated the appropriateness of combining these studies by showing a lack of statistical heterogeneity using the Q statistic. We eagerly anticipated the response by Leichsenring and Rabung to our sweeping critique [1] of their 2008 metaanalysis [2] in which they claimed longer-term psychodynamic psychotherapy (LTPP) was more effective than shorter-term psychotherapies. Yet, we were disappointed that Leichsenring and Rabung [3] , like Ehrenthal and Grande [4] , dealt with so little of the specifics of our critique. Much of the substance of the response by Leichsenring and Rabung refers to an unfinished manuscript and another, still undergoing review in some unnamed journal, limiting the opportunity for independent evaluation by anyone open-minded but skeptical about each of our perspectives. The crux of our critique was statistical and technical because it is in these areas that serious flaws in their metaanalysis initially occurred, undermining any of their substantive interpretations. Perhaps our readers were as confused as Leichsenring and Rabung seemingly were, and thus we will use the 1,000 words allotted to us to briefly summarize our critique in substantive terms. A basic criticism was that Leichsenring and Rabung consistently miscalculated summary effect sizes in a way that was grossly inflationary and favorable to LTPP. We find no basis for such an approach in the extensive literature of metaanalyses. We generated a hypothetical set of 10 studies in which there were only trivial differences in effect sizes for intervention and comparison groups. Using the computational method by Leichsenring and Rabung, we nonetheless generated a large summary effect size favoring the intervention that would have been unprecedented in the psychotherapy literature. Apparently neither Leichsenring and Rabung nor prepublication reviewers of their paper noticed that the summary effect sizes Leichsenring and Rabung claimed for LTPP were greater than any of the effect sizes for individual studies. Moreover, they claimed one implausible between-group effect size of 6.9, equivalent to 93% of the variance explained, for a set of studies in which none reported an effect size of more than approximately 2. Such figures should have aroused suspicion: their estimates were not, Received: June 21, 2010 Accepted: June 21, 2010 Published online: November 18, 2010

The Negative Stereotyping of Single Persons Scale: Initial Psychometric Development.

Objectives The current study presents an initial psychometric analysis of a three-dimensional scale designed to measure stereotyping of people who are single, defined as not currently legally married. Methods Following expert review, proposed scale items are administered to 178 graduate and undergraduate students. Reliability analysis and Confirmatory Factor Analysis are conducted. Results Coefficient alphas range from .87 to .91 on the proposed subscales and CFA result in all fit indices meeting recognized criteria for a good fit. Evidence for convergent validity is mixed. Conclusions The present analysis indicates that the scale has excellent reliability and factorial validity. More research is needed to support convergent validity. Future research with different populations and more specific focus on older, non-cohabiting singles is recommended.