|Home | About | Journals | Submit | Contact Us | Français|
The paper gives an outline, with examples, of various statistical methods which may be of special use in psychiatry.
(1) Actuarial data.—The simple accumulation of accurate figures on the ages of patients, their diagnoses and length of stay in hospital or under treatment for mental illness is of great value in understanding the scope of psychiatric problems. The age and sex incidences which correspond to different disease groups are very characteristic. Such material has value in the estimation of the results of therapeutic experiments but special methods have to be devised, as there is no exact prototype in standard vital statistics or in work on therapeutic trials.
(2) Biometric techniques.—Knowledge of ordinary statistical practice guards against elementary errors and aids in establishing significance or otherwise of metrical deviations from the normal found in mentally ill subjects. Also the range of variations may be much more marked in abnormal than in normal groups. Furthermore, abnormal reactions in themselves may be characterized by either too much or too little variety, i.e. by scatter or by stereotypy.
Discrimination between normal and abnormal reaction can be based on a single quantitative measurement, on difference in variance or on a compound measurement, i.e. pattern or profile. The discriminative approach has advantages over other methods because in this approach the initial factors are concrete and based on known classes such as males and females, children and adults, special clinical types, &c.
(3) Genetical analysis.—Actuarial data can be useful in genetic studies by leading to estimation of population frequency of genes and consanguinity rates. Moreover, combined clinical and genetical observations can reveal the existence of new clinical entities.