We mostly achieve this in randomised controlled trials (RCTs), where patients are assigned one of two alternative treatments. You can then compare a placebo with active treatment or a new treatment with the best available alternative.
This method has been enthusiastically applied to psychological treatments. For children, studies tell us that specific anxiety problems are well treated by exposure to what the child fears, coupled with encouragement and positive reinforcement.
Conduct problems in young children improve if parents receive training in behavioural management, and the addition of problem-solving training for the children may also help.
Depressed youngsters may respond to psychodynamic therapy, family therapy, interpersonal therapy and cognitive behavioural therapy.
But it is vital to remember the limitations of RCT studies. Most are performed in university clinics, with only 4 per cent providing treatment in the kind of setting where children are actually likely to be seen. Nine out of 10 studies are carried out in America and their findings may not apply to the multicultural populations of British cities.
Many samples are drawn from radio and newspaper advertisements, while only 13 per cent of patients come from families who feel desperate enough to seek treatment. Perhaps no more than 1 per cent of the studies replicate anything like a real-life setting.
When some experimental treatments are compared with no treatment, the impact appears great, but when compared with conventional treatment, the results are less impressive. A recent UK study of the treatment of adolescent anorexia found no difference between specialised and costly cognitive behavioural therapy and the talking therapies usually offered.
Follow-up of the large RCT studies is often inadequate. The investigation into Ritalin, featured in last week's BBC Panorama programme, showed that medication deemed highly effective after one year proved much less so after three.
How different are the 500 or so therapies for children? Close scrutiny reveals that they share many common features, such as promoting competence and self-monitoring and addressing relationship skills.
And do specific techniques matter as much as the quality of the therapist? Certainly, the outcome and the risk of premature drop-out depend less on treatment than on the perception of the therapist as competent and committed. Children tend to rate their relationship with the therapist as the most important aspect of their treatment.
So while the evidence from RCTs is extremely helpful to guide doctors in individual cases, it is not yet complete or robust enough to change the shape of psychological services for children. We must not allow politicians and policymakers to make partly-informed decisions that we come to regret. If the Government's new-found enthusiasm for cognitive behavioural therapy were to mean reduced access to family therapy, for instance, that would be a great shame.
Lack of evidence should never be allowed to deprive a child or family of potentially effective help. In the absence of RCT results, we should insist on practice-based evidence - collecting data from children and families that tells us if the child is better off than they would be without treatment. Science is good for practice, but practice is also excellent for science
Peter Fonagy is Freud Memorial Professor of Psychoanalysis at University College London and chief executive of the Anna Freud Centre
Fonagy, P, Target, M, Cottrell, D, Phillips, J, Kurtz, Z (2002). What Works For Whom? A Critical Review of Treatments for Children and Adolescents. New York: Guilford.
Kazdin, AE (2004). Psychotherapy for Children and Adolescents. In M Lambert (Ed.), Bergin and Garfield's Handbook of Psychotherapy and Behavior Change, 5th Edition (pp. 543-589). New York: Wiley.
Weisz, JR, Doss, AJ, Hawley, KM (2005). Youth Psychotherapy Outcome Research: A review and critique of the evidence base. Annu Rev Psychol, 56, 337-363.
Target, M, Fonagy, P (2002). The Psychological Treatment of Child and Adolescent Psychiatric Disorders. In A Roth P Fonagy (Eds.), What Works For Whom? A Critical Review of Psychotherapy Research (2nd ed., pp. 385-424). New York: Guilford Press.