Statistics in Latin America

Statistics is a relatively new discipline in Latin America. At present only sparse information is available about the state of the statistical profession in Latin America. We present and summarize the available information about the job market; training, education, and research; professional societies and their activities; and so forth. We also share our hopes, concerns, and suggestions for the future.     

Multiple Co-occurring Behaviours among Gamblers in Treatment: Implications and Assessment

A substantial portion of gamblers in treatment may have co-occurring mental health disorders including substance use disorders, especially alcohol dependency, personality disorders, affective disorders, anxiety disorders and impulse control disorders. Co-occurring mental health disorders affect treatment seeking, the treatment process and its outcome, quality of life and functioning in many mental health disorders. Participants in this study were 78 adults enrolled in state-supported out-patient services for pathological gambling and current Gamblers Anonymous attendees. The majority of participants (76.6%) had co-occurring behaviours; 55.8% had multiple co-occurring behaviours. Participants with multiple co-occurring behaviours were more likely to report that a co-occurring behaviour increased the severity of their gambling symptoms. A dose/response relationship between number of co-occurring behaviours and severity of gambling problems was found. Co-occurring behaviours can interact with gambling behaviour. Brief, validated screening instruments are available for screening and diagnosing co-occurring behaviours among gamblers presenting for treatment.     

Strong laws of large numbers for dependent heterogeneous processes: a synthesis of recent and new results

This paper surveys recent developments in the strong law of large numbers for dependent heterogeneous processes. We prove a generalised version of a recent strong law for Lz-mixingales, and also a new strong law for Lpmixingales. These results greatly relax the dependence and heterogeneity conditions relative to those currently cited, and introduce explicit trade-offs between dependence and heterogeneity. The results are applied to proving strong laws for near-epoch dependent functions of mixing processes. We contrast several methods for obtaining these results, including mapping directly to the mixingale properties, and applying a truncation argument.     

Ban Estimation for Chi-Square Test Criteria in Categorical Data

This expository paper describes some recent work that further develops the theory of BAN estimators and the related chi-scuare test statistics.The extensions are in several directions: (a) the class of regular estimators is broadened by permitting extraneous random elements; (b) more general models are permitted under the constraint equations specification; and (c) BAN estimators are defined for general models combining features of two types of specification.In particular, WLS estimators are shown to be BAN.     

On the use of the scholz-sievers method for dlstribution-free inference about the difference in slopes of two regression lines

A simple adaptation of a distribution-free method due to Scholz (1978) and Sievers (1978) for inference in a single regression setting is proposed for inference about the difference in slopes of two regression lines. We assume that the data are obtained from a designed experiment with common regression constants. A comparison of the proposed method to its competitors-one due to Hollander and the other due to Rao and Gore-indicates superiority of the new method.