Sign in to follow this  
Followers 0

Online Essay Club No.32: Psychotherapy

5 posts in this topic

'Psychotherapy is inherently untestable'. Discuss this statement critically with reference to methodological issues and the available evidence.

Share this post

Link to post
Share on other sites

Model answer from marking scheme:


There are three components to this question:

1. Research methodology and issues of testability in general;

2. Methodological issues regarding psychotherapy;

3. Awareness of the research literature.

Candidates may focus more on one area than another if they wish but all areas must be covered to some degree for a pass mark.


Testability and general research methodology:

Candidates should show awareness of the underlying principles of research design, including hypothesis testing, randomisation, control groups, measurement of change, analysis of data, sources of bias, length of follow-up, generalisability. Candidates should show awareness that all research involves a 'trade-off' between 'internal validity' and 'external validity' - the efficacy 'v' effectiveness problem.

In relation to psychotherapy good candidates will know about the debate between an approach using normal empiracal methods and an attempt to investigate 'meaning'. However, the two approaches are not mutually exclusive and candidates should focus on the issue of 'testability'. One paradigm has been called the 'drug metaphor' i.e. the research question is essentially comparable to psychopharmacology in developing 'pure' interventions delivered in a measurable form for specific problems with defined outcomes. The research summarised below demonstrates that this approach has been fruitful.

However, this concept of testability is limited and good candidates will bring in the need to test an underlying theory or mechanism by investigating process. Attempts have been made, but they are less developed than the 'effectiveness and efficacy' paradigms.

Share this post

Link to post
Share on other sites

Particular problems in psychotherapy research

At least one of these points must be mentioned to gain a pass mark.

Limitations of the 'drug metaphor':

The following issues may have an appreciable effect on outcome, obscuring treatment-specific effects: Patient/ therapist 'fit', characteristics of the patient and therapist, patient expectations, ability of therapist to understand certain types of patients, and allegiance of investigator to type of therapy. Randomisation may affect attrition leading to small cell size.

Monitoring quality of therapy

The main issue to be discussed is the extent to which therapists adhere to a model and practice it competently. One approach has been to provide treatment manuals which state what is expected explicitly, but manuals are easiest to write for single, uncomplicated disorders, may affect the therapeutic alliance and may not generalise to routine practice and cannot be used easily in long term treatment.

Problems of measurement

To improve the plausibility of studies measurement should take several perspectives, e.g. therapist; patient; independent observers and possibly relative; friends; employer. Also, different domains should be covered (symptoms, well being, social and occupational function).

Ideally, measures should reflect the underlying principles of the therapy but be applicable across different modalities. This has not been achieved fully, and some measures may intrinsically favour particular modes of therapy. There are few reliable measures of personality function as these are traditionally seen as trait measures and are not designed to be responsive to change.


Success of therapy involves ability to improve the patient's functioning at the end of treatment and maintaining that improvement. Follow-up of patients is often inadequate.

Statistical and clinical significance

Studies may reject the null hypothesis but actually do not show clinically significant change. Good candidates will be aware of different approaches to clinically significant change:

measure by a criterion measure of change

use a categorical criterion of recovery rather than change over time

compare with normative samples.

Sample size

There are relatively few studies with adequate power to distinguish between alternate treatments. Equivalence of therapies in many conditions is a common finding.

Diagnostic Considerations

The commonest standardised criteria for psychiatric disorders are the DSM and ICD. Diagnosis is not an ideal way of developing homogenous treatment groups and the problem of comorbidity has been insufficiently studied. Diagnosis is a poor predictor of outcome.

Awareness of research literature

Candidates should show some awareness of the available literature as summarised in several reviews, particularly Roth & Fonagy 1996 'What works for whom?'

Strongest evidence for efficacy (large numbers of RCTs) Good evidence for efficacy (several RCTs) Some evidence for efficacy (more trials needed)

Phobic disorders (panic, agoraphobia, specific phobias, social phobia Hypochondriasis and somatisation (including IBS) Borderline, antisocial and avoidant personality disorders

Obsessive compulsive disorder Chronic fatigue syndrome Self-harm behaviours

Depression Post traumatic stress disorder Bipolar affective disorder

Eating disorders Schizophrenia Substance misuse

Sexual dysfunction

N.B. Candidates should know that much of the evidence on efficacy comes from CBT, but they should be aware that there is now evidence of efficacy from RCTs for other modalities including psychodynamic and interpersonal psychotherapy, and for systemic therapies.

Good candidates will draw attention to the few comparative studies and the difficulty interpreting the results, e.g. the NIMH Treatment of Depression, Collaborative Research Programme (TDRCP). Meta- analyses of outcomes have shown relatively few advantages for any specific treatments, although there are slight advantages to CBT.


The essay should have an introduction outlining the forthcoming argument. Logically the essay should discuss research methodology in general, apply it to psychotherapy research, and illustrate the points with some research examples.


A pass mark requires:

awareness of the underlying methodological issues

the research literature.

At least one of the points listed under 'Particular problems in psychotherapy research'.

Awareness that the research methodology has shown the value of psychotherapies in the treatment of psychiatric disorder.

Higher marks should be awarded if particular studies are cited or for greater awareness of methodological issues particular to psychotherapy.


Aveline, M and Shapiro, D (eds) (1995) Research foundations for psychotherapy practice. Wiley: Chicester.

Elkin, I, Shea, MT, Watkins, JT et al (1989) National Institute of Mental Health Treatment of Depression Collaborative Program: General effectiveness of treatments. Archives of General Psychiatry 46, 971-982.

Roth, A and Fonagy, P (1996) What works for whom? A critical review of psychotherapy research. Guildford: New York.

Shapiro, D, Barkham, M, Ress, A, et al (1994) Effects of treatment duration and severity of depression on the effectiveness of cognitive/ behavioural and psychodynamic/ interpersonal psychotherapy. Journal of Consulting and Clinical Psychology, 62, 422-534.

Share this post

Link to post
Share on other sites

Is there any degree of overlaping between this topic & c ritical appraisal? ??? ??? ::)

Share this post

Link to post
Share on other sites

tough topic to answer ,but if covered well ,,,full marks 8)

Share this post

Link to post
Share on other sites
Sign in to follow this  
Followers 0