Multivariate statistical methods are being used increasingly in an effort to clarify problems in psychiatric classification. Numerous references to their application could be given, but it is sufficient here to mention recent publications by Kiloh and Garside (1963), Carney, Roth and Garside (1965) and by Kendell (1968) which have received especial attention (e.g. Eysenck, 1970), as they are concerned with the longstanding controversy about the classification of depressive illness. But there is some confusion, not least amongst the critics, about the particular roles which different multivariate techniques, notably factor analysis, can play in classification problems, and in this paper an attempt is made to clarify the situation. Almost as a trial of the reader's endurance this attempt necessitates some preliminary discussion of the psychometric concept of a ‘unitary trait’ and of the statistical rules involved in defining it. But after this problem has been disposed of attention is confined to a discussion of the main multivariate techniques in question—factor analysis, discriminant function and canonical variate analysis, and finally cluster analysis. The attitude adopted in the paper is admittedly purist, but when controversy arises, as in the case of the classification of depressive illnesses, it is well to be clear about the precise properties and purposes of the statistical models we employ if confusion is to be avoided.

Anxiety neurosis has been the subject of intensive study during the past ten years. Renewed interest in this syndrome has arisen both from the stimulus of new methods of treatment and from attempts to classify anxiety more accurately. Recently Pitts and McClure (1967) reported that anxiety symptoms and anxiety attacks could be produced by a specific biochemical stimulus: sodium lactate. This work was prompted by the finding that 'standard exercise’ tends to produce an excess amount of lactic acid in patients with anxiety neurosis (Cohen and White, 1950; Jones and Mellersh, 1946; Linko, 1950; Holmgren and Strom, 1959). Pitts and McClure found that an intravenous infusion of 10 ml. of half-Molar sodium (DL) lactate per kilogram of body weight, given over a twenty minute period, produced an anxiety attack in patients suffering from anxiety neurosis. The symptoms began a minute or two after the infusion was started and developed rapidly, and some patients reported ‘exacerbations of their characteristic symptom profiles for two to five days after the sodium lactate infusion’. In their double-blind study, many fewer symptoms were produced when calcium ion was added to the lactate infusion, and almost no symptoms were produced by an infusion of glucose in saline of similar osmolarity. Many fewer and less severe symptoms were produced in normal controls than in the patient group, both in response to sodium lactate and to lactate with added calcium, and almost no symptoms during the glucose and saline infusion. Pitts and McClure postulated that anxiety symptoms may have a common determining biochemical end-mechanism, involving the complexing of ionized calcium at the surface of excitable membranes by lactate ion with resulting interference ‘with the normal functioning of calcium in transmitting nerve impulses' (Pitts, 1969). They concluded that ‘there may be something highly specific about lactate ion in producing the naturally occurring hypocalcaemic anxiety symptoms in human beings'.

In 1959, West and Dally reported the effects of iproniazid in depressive syndromes, drawing attention to the greater therapeutic effectiveness of this drug in ‘atypical’ or ‘hysterical’ depressive states than in endogenous depression. Later, Sargant and Dally (1962) found a group of patients with good previous personalities, showing obvious features of depression, whose anxiety symptoms responded to phenelzine alone, but in view of the response of other patients with anxiety they questioned whether monoamine oxidase inhibitors were really antidepressant, or whether they were essentially drugs acting against anxiety. These findings prompted this study, intended to compare other parts of the clinical pictures of anxiety states responsive to monoamine oxidase inhibitors with those of depressive illnesses.

Although a number of papers have been written emphasizing the need for conservative dentistry and describing various ways of dealing with patients fearful of dental treatment, little systematic attention has been paid to understanding the basis of this impediment. A fear of dental treatment is, indeed, so common that it can almost be considered normal unless of such a degree as to interfere with much needed dental care. Thus, in the U.S.A. Freidson and Feldman (1958) reported that 51 per cent of a large and representative sample did not visit a dentist regularly. Out of these 9 per cent reported that their reluctance to do so was based on fear. Similarly, Kegeles (1963) reported that fear of dental treatment significantly reduced the number of preventive dental care visits, independent of social class. Shoben and Borland (1954) found that unfavourable family dental experience and unfavourable family attitudes towards dentistry were the only factors which discriminated between fearful and non-fearful dental patients.

This paper describes a retrospective study of the influence of personal variables and treatment methods on the prognosis for a return to flying of service aircrew suffering from anxiety states precipitated by flying stress. The study was made preparatory to a prospective investigation of treatment of anxiety states in aircrew.

In this study, 42 school phobic youngsters aged 11 to 15 were investigated to find out if they were exceptionally dependent, particularly on their mothers, as had repeatedly been asserted in connection with similar cases, often without adequate supporting evidence of an objective nature (Berg, 1970). The Highlands Dependency Questionnaire (H.D.Q.) was administered to their mothers around the time they were accepted for admission to an adolescent psychiatric in-patient unit. The Questionnaire had previously been found to measure at least two unrelated sorts of dependency, represented by a sociability factor (I) and an immaturity factor (III) with satisfactory reliability and validity; this emerged when a principal component factor analysis was performed on the results of applying it to the mothers of a randomly selected sample of 68 secondary school children from the general population, stratified for age, sex and social class (Berg et al., 1971). Data on 14 variables was processed on the University of Leeds English Electric KDF9 computer, using a system of standard programmes (Hamilton et al., 1965). In addition to the two sets of factor scores which were calculated using actual raw score weights, two corresponding sets of subscale scores were calculated using approximate raw score weights; correlations between factor and subscale scores in the control group had been found to be: r = ·87 for sociability and r = ·84 for immaturity (Berg et al., 1971), whereas in 19 youngsters out of the school phobic group, looked at in another context, they were: r = ·51 for sociability (perhaps explained by a divergence between at least two tendencies which have different emphasis in the factor and subscale scores) and r = ·95 for immaturity. The criteria adopted for the diagnosis of school phobia had been given previously when 29 school phobic cases with similar clinical features were reported in some detail (Berg et al., 1969).

The trial was a double-blind cross-over study in which patients presenting in general practice with anxiety received haloperidol (Serenace) 0·5 mg. twice daily for two weeks and a placebo of identical appearance for two weeks in random order. Forty-three patients entered the trial. Of these, six failed to attend for follow-up and one proved impossible to assess. A further eight completed the first two-week period of assessment only (5 active drug, 3 placebo) but are included in the analysis. There were equal numbers of men and women, aged 24 to 71 years.

For the purposes of this paper the term ‘precipitant’ is defined as a new circumstance associated in time with the onset of the addiction. If the precipitant is merely the ‘final staw’ in a long chain of causal circumstances, no particular value can be attached to its identification. If, on the other hand, the precipitant brings into operation the background causal factors leading to addiction, it has important practical and therapeutic implications. It could be expected that this type of precipitant might persist during the addiction and its disappearance might be associated with remission of the addiction.

In a survey of 612 young people remanded in the London area during 1964, positive urine tests were found for amphetamine for 18 per cent of the 558 boys and 16 per cent of the 54 girls examined (Scott and Willcox, 1965). This paper presents the findings of a similar type of survey carried out in the same remand homes five years later.

Jellinek describes periodic or Epsilon alcoholism as the ‘least known species of alcoholism’ (1). Åmark (2), comparing them to the general population, found periodic drinkers to be older at time of study, to show increased heritability in offspring and to have a higher incidence of cycloid and cyclothymic personalities. His determination of cyclothymic personality traits reinforces the hypothesis of Dobnigg et al. (3) of a correlative between periodic drinking patterns and affective symptoms. This paper examines the usefulness of the diagnosis of ‘bender’ alcoholism.

A number of recent reports have drawn attention to the abnormal EEG findings obtained in subjects with Klinefelter's Syndrome (Dumermuth, 1961; Hambert and S: son Frey, 1964; Hambert, 1964; Nielsen, 1969; Nielsen and Pedersen, 1969; Pasqualini et al., 1957; Prader et al., 1958; Zuppinger et al., 1967). An increased prevalence of epilepsy within this group has also been noted (Hambert, 1964). Much less is known about the electroencephalogram in the XYY syndrome, the available data being limited to single case reports or reports involving only a few subjects (Borgaonkar et al., 1968; Bartlett et al., 1968; Court-Brown et al., 1968; Cowie and Khan, 1968; Forssman, 1967; Forssman and Hambert, 1966; Kessler and Moos, 1970; Leff and Scott, 1968; Mintzer et al., 1968; Nielsen and Pedersen, 1969; Nielsen and Tsuboi, 1969; Nielsen et al., 1966; Pergament et al., 1968; Persson, 1967; Weiner et al., 1968; Welch et al., 1967). This paper describes the initial findings of a comparative study of Special Hospital patients with sex chromosome abnormalities and a group of matched controls of normal chromosome complement from the same institution.

The paranoid—non-paranoid distinction among schizophrenics may be useful and meaningful (14): certain psychological and physiological responses have been shown to distinguish patients of these subgroups (5, 12, 16) and biochemical (2, 5) and morphological (7, 9, 17) differences have also been demonstrated.

There is little definite information about the psychological correlates of the XYY chromosome anomaly. Hope et al. (1967) made observations on 7 subjects relevant to Reactive-Narcissism (Ss scores lower), Dependence (higher), Over-activity (lower), HOQ (More obsessional). Other studies with inconclusive results have been made by Little (1968), Nielsen and Tsuboi (1969) 9SS, Street and Watson (1969) 22SS, and Forssman (1970) 11Ss.

The difficulty in calculating the incidence and prevalence of transsexualism is obvious. Those figures which can be arrived at are always minimal values, and a group of persons with transsexual symptomatology, which are most difficult to assess, will always elude the calculations.

Personality structure has been linked with cancer, heart disease, and duodenal ulceration. Eysenck (1965) hypothesized that the association of personality with these diseases is the result of an association between the genetic factors influencing personality predisposition and disease susceptibility. As briefly discussed by the present author elsewhere (Segraves, 1970), Eysenck's hypothesis seems tenable, since both these personality dimensions and these diseases have been shown to have hereditary components and to be related to bodily habitus in analogous ways.

In a review of the 1967 group psychotherapy literature, MacLennan and Levy (1968) record that nearly 300 papers were published in that one year. Unfortunately, this proliferation does not reflect any great increase in the knowledge of group psychotherapy processes. There are plenty of original ideas, but these are almost totally unsupported by systematic theory and research. Where theoretical ideas have been used, they have been lifted from the sphere of individual psychotherapy and applied unmodified to group psychotherapy. The research has largely been concerned with determining such things as the most efficient group structure; who talks to whom about what; and what grammatical forms are used.

Bannister and Fransella (1966) described a clinical test based on the theory and methods of Kelly (1955). This test has since been widely used in clinical practice, usually as a test of schizophrenic thought disorder, and has also been used in several studies attempting to define the schizophrenic condition more exactly (e.g. Bannister and Salmon, 1966; McPherson and Buckley, 1970; McPherson, 1969). The test was developed from the previous work of Bannister (1960, 1962) with schizophrenics and other psychiatric patients.

Eye-contact is a prominent Non-verbal cue in any Dyadic interaction and can be easily observed and measured. Argyle (1969) has postulated a relationship with Intimacy and Distance, Kendon (1967) described its expressive and regulatory function in conversation and Exline et al. (1965 and 1967) have related it to other variables, e.g. sex, dependency and affiliative needs.