Wilson's recent critique of the authors' appraisal of meta-analysis appears to misunderstand them as claiming more for meta-analysis than they intended. The present paper seeks to clarify consequent confusions concerning selection of studies, the quality of the literature reviewed, the classification of therapies, and the non-identical results of different meta-analyses. It is acknowledged that no single meta-analysis is definitive.

This is a series of short invited contributions from therapists and research workers who have volunteered to answer the following question: “If your fairy godmother were to wave her magic wand and give you an immediate answer to a question or questions, without having to do the relevant research, what would you most like to know?”

Shapiro (1961) devised a scaling method based on pair comparisons to overcome the problems associated with numerical rating scales. Phillips (1963, 1966, 1970, 1977, 1979) developed the theory underlying Shapiro's method and extended it to increase the precision of measurement. In this paper we present a simpler justification of the Shapiro–Phillips method and propose improvements by sequentializing the presentation of pair comparisons. In this way, the total number of comparisons can be reduced to correspond with an acceptable level of reliability. Responses are recorded graphically producing an instant score and reliability check. Some possible generalizations are discussed.

A multiple baseline design across subjects was employed to study the effects of training for manuscript handwriting on cursive letter formation. Ten first grade students received training to increase their accuracy in manuscript writing. Following training, all 10 students demonstrated increased accuracy in untrained cursive letter formation. Cursive letters were practiced but untrained in that no feedback on correctness was given at any session.

Meta-analysis is a systematic and quantitative approach to reviewing empirical literature. Its claimed benefits of superior objectivity and dependability arise from its adoption of the methodological principles and methods of primary empirical research. This paper offers a critical appraisal of these claims, with special reference to the application of meta-analysis to the therapy outcome literature. Features examined include the numerical combination of results obtained in independent studies; search procedures and inclusion criteria; and the coding of objective and qualitative features of the source studies, which are then used to study the correlates of outcome via disaggregation and multiple regression analysis. The major criticisms reviewed include allegations of overgeneralization, indiscriminate inclusion of low quality data, and idiosyncratic and unacceptable conclusions. It is concluded that the continued application and refinement of meta-analysis in the field of therapy outcomes has an important contribution to make, although not to the exclusion of other methods of integrating data. It is predicted that methods and principles currently associated with meta-analysis will increasingly find a place in all reviews of extensive tracts of empirical literature.

Several clinical deductions drawn from a 3-systems analysis of obsessions were tested in an experiment on 12 patients, each of whom received four sessions of habituation training or thought-stopping. The major predictions received only weak and inconsistent support. Moreover, the intrasession changes tended to fade in the inter-session intervals, and there were few signs of therapeutic advance.

A test instrument is presented which utilises role-play techniques to assess parenting skills in terms of behaviour modification techniques. The validity of the test is confirmed by its ability to measure established parenting skills as well as those acquired in a parent training exercise. The test should prove to be particularly useful in evaluating parent training programmes.

Behaviour modification procedures have been subject to criticism by proponents of the cognitive evaluation theory using as an empirical base, the overjustification effect. In this paper some of the tenets of the theory as applied to behaviour modification and the parameters of the overjustification effect are examined. It is concluded that methodological problems in the reported research and the lack of comparability between overjustification experiments and commonly used behavioural procedures invalidate the claim that rewarding a behaviour will lead to a decline in intrinsic motivation to engage in that behaviour.

Shapiro and Shapiro (1982) argue the case that meta-analysis has made, and will increasingly make, an important contribution to the evaluation of the effects of psychological therapies. Briefly stated, they make the following claims for meta-analysis: (1) Meta-analysis provides systematic and quantitative methods for summarizing and integrating the voluminous and diverse literature on psychotherapy. (2) Meta-analysis reduces (eliminates?) the subjective bias that is said to vitiate traditional, qualitative reviews of the literature. Consistent with similar allegations by Smith et al. (1980), Shapiro and Shapiro (1982) suggest that the conclusions drawn in “some” of these traditional reviews “are a function of the assumptions, predispositions and consequent biases of the reviewer.” Meta-analysis, however, is put forward as a distinct alternative to this sorry state of affairs, since it purportedly overcomes these sources of personal and professional bias and results in more objective more accurate evaluations of the evidence. (3) Meta-analysis has “the same superiority over a traditional literature review as that enjoyed by a systematic therapy outcome study over a series of impressionistic case reports.” (4) The advent of objective, quantitative meta-analytic procedures does not abolish the need for traditional literature reviews subjective and qualitative as they might be. The latter have their limited place, just as impressionistic clinical case reports can complement well-controlled experimental outcome studies.

Twenty-four moderately asphasic stroke patients attending a rehabilitation centre were treated for 4 weeks with speech therapy and for 4 weeks with either an operant training procedure, based on that devised by Goodkin (1966), or an attention placebo non-specific treatment. Speech therapy was given first to half the patients, while a quarter received operant training first and a quarter non-specific treatment. Results indicated no significant differences between the treatments. These treated patients were also compared in improvements in language abilities with nine moderately aphasic patients from hospitals with very limited speech therapy services. No significant differences between “treated” and “untreated” groups were found over a 4-week period. All groups of patients, however, showed some significant improvements in language abilities.

The article by Shapiro and Shapiro (1982) published in Behavioural Psychotherapy is one of three, or possibly four, which they have written on the subject of meta-analytic approaches to the evaluation of psychotherapy. Commenting on each of their articles would be burdensome for critics and tedious for readers. Wilson and Rachman (1982) have already written an evaluation of the Shapiros' position, and as Wilson (1982) gives an incisive and comprehensive critique of their Behavioural Psychotherapy article, I will confine myself to a few brief remarks.

A number of problems associated with the use of the commonly employed interval and event recording procedures are discussed and it is suggested that these problems may be overcome by the use of real time recording techniques. An electronic system for recording ongoing behavior is described. The Data Acquisition in Real Time II (DART II) system consists of an open-format, microprocessor-controlled coding unit which permits the recording of the onset and offset of discrete behaviors as they occur in real time. It is completely portable, totally silent in operation, allows up to 60 discrete behavioral categories to be coded, provides comprehensive feedback to the user, and can store data almost indefinitely in a form which allows for direct transfer to computer for analysis. While the DART II system may be employed for interval or event recording, its use in a two-channel configuration to record the interactions between a target subject and several input persons is described. Advantages of the system are discussed.

The effects of a structured sex education programme on the knowledge and attitudes of a group of disturbed adolescent offenders with sexual problems were assessed. Sex education was found to improve sexual knowledge significantly, to lessen anxiety associated with sexual intercourse, and to produce a less anxious self concept. Sex education had no significant effect on sexual interest.

Several writers have been interested in the extent to which behaviour therapy techniques are derived from the principles of learning. The present paper reviews this issue with respect to stimulus generalization and response generalization in behaviour therapy. Generalization has been a problem because many therapists have reported that newly learned skills and behaviours have not transferred outside the treatment situation to settings in the patients' or clients' life. Reviewing the literature on stimulus generalization it is concluded that in research on learning theory, stimulus generalization is shown to occur along a single stimulus dimension while in the behaviour therapy literature researchers often attempt to produce generalization across many stimulus dimensions simultaneously and as a result generalization may be poor. It is suggested that subsequent therapy and research should be designed so that treatment situations and generalization situations differ on as few stimulus dimensions as possible, thus maximizing the probability of generalization. With respect to response generalization it is felt that the concept is weak as an explanation for behaviour change and it may be more useful to consider that behaviours are functionally related, so that changes in behaviour A will produce changes in behaviour B or that covert responses may account for changes in untreated behaviours. Finally, a recommendation is made that therapists be more precise about the stimulus dimensions along which they wish behaviour to generalize.

A scheme is presented in which the level of agreement between two observers carrying out duration recording can be more accurately estimated than by conventional procedures. Each observer operates a push-button connected to a microcomputer to signal the presence of the behaviour to be measured and the duration of its occurrence. At the end of the measurement period, the computer reports each observer's rating of the proportion of time for which the behaviour was present, as well as various measures of the level of agreement between observers.

Forty-four community mental health outpatients participated in an assertiveness training program in which three treatment methods were compared. Skill training aimed at training specific assertive responses by means of instruction, modeling, behavioral rehearsal, feedback, and homework assignments. Cognitive training focused on altering the irrational cognitions which theoretically inhibit the expression of assertive responses. A third group (Cognitive–Skill) received both forms of training simultaneously, and a fourth group (Waiting List) was included as a control. A behavioral roleplaying test (B.R.P.T.) and the Social Fear Scale were administered prior to and immediately following four individual training sessions occurring at one week intervals for all three treatment groups. Analyses of covariance of assertive content ratings made by independent judges for B.R.P.T. responses revealed that all three treatment groups evidenced significantly more assertive and effective responses on both refusal and behavior change situations when compared to the Waiting List control. Analyses of paralinguistic dimensions of the B.R.P.T. revealed training generally to be more successful for the Skill group.

Marks (1981) recently argued that behavioural psychotherapy had few practical, theoretical or historical links with theoretical and experimental psychology. This paper attempts to elaborate some of Wilson's (1981) arguments against this view and cites some additional British evidence in support of these arguments. Wilson's consideration of the issue of service delivery appears unsatisfactory and the suggestion is made that behavioural psychotherapy be taken out of the medical arena and placed within a rationalized therapeutic service independent of the present medical–sychological–social work structures.