A framework for the comparison of maximum pseudo-likelihood and maximum likelihood estimation of exponential family random graph models

The statistical modeling of social network data is difficult due to the complex dependence structure of the tie variables. Statistical exponential families of distributions provide a flexible way to model such dependence. They enable the statistical characteristics of the network to be encapsulated within an exponential family random graph (ERG) model. For a long time, however, likelihood-based estimation was only feasible for ERG models assuming dyad independence. For more realistic and complex models inference has been based on the pseudo-likelihood. Recent advances in computational methods have made likelihood-based inference practical, and comparison of the different estimators possible.     

Information systems and managerial decision making

The role of an information system is to support managerial decision processes. Some of these processes can be completely specified in advance and, as long as the specification remains acceptable to management, resolution can be done automatically, within the framework of the information system. However, in other cases, managers are involved in all stages of the decision process, which cannot be completely specified in advance. The task of the information system in these cases is to provide information to the managers engaged in making the decisions. This paper considers these problems from the point of view of the information system designer and makes recommendations about design and implementation.     

Respondent-Driven Sampling: An Assessment of Current Methodology

Respondent-Driven Sampling (RDS) employs a variant of a link-tracing network sampling strategy to collect data from hard-to-reach populations. By tracing the links in the underlying social network, the process exploits the social structure to expand the sample and reduce its dependence on the initial (convenience) sample.The current estimators of population averages make strong assumptions in order to treat the data as a probability sample. We evaluate three critical sensitivities of the estimators: to bias induced by the initial sample, to uncontrollable features of respondent behavior, and to the without-replacement structure of sampling.Our analysis indicates: (1) that the convenience sample of seeds can induce bias, and the number of sample waves typically used in RDS is likely insufficient for the type of nodal mixing required to obtain the reputed asymptotic unbiasedness; (2) that preferential referral behavior by respondents leads to bias; (3) that when a substantial fraction of the target population is sampled the current estimators can have substantial bias.This paper sounds a cautionary note for the users of RDS. While current RDS methodology is powerful and clever, the favorable statistical properties claimed for the current estimates are shown to be heavily dependent on often unrealistic assumptions. We recommend ways to improve the methodology.     

Some initial specifications for a strategic information system

Much of the discussion and published literature on information systems to date has been about systems concerned with the internal administration and operations of an organization. Considerably less attention has been given to the information requirements of the more strategic managerial activities of planning and policy making, in which the future directions and objectives of the organization are decided. Internal information systems in many organizations have now grown to maturity. At the same time, managers have become aware of the need for an additional component of the overall information system to support their more strategic activities. This paper considers the broad characteristics of a strategic component of an organizational information system and derives some initial specifications from which a more comprehensive design can be developed.     

Parental involvement in reading: the influence of socio-economic status and supportive home visiting

SUMMARY: This paper describes the deployment of the technique for non-professional tutoring of reading known as ‘paired reading’ on a large scale in the Kirklees Local Education Authority. From the very large amount of data generated in multi-site field trials, particular attention is focused on the relationship between supportive home visiting (if any), an index of the socioeconomic disadvantage of the catchment area of each participating school, and outcomes in term of gains on norm-referenced tests of reading. Some surprising conclusions emerge.     

An LP model for assessing congeneration strategies

A linear programming model is described which is designed to assist in the assessment of optimum operational strategies for industrial plants with cogeneration facilities. A concise format is provided for linking together basic elements such as boilers, valves, turbines and processes with steam and electricity demands with steam, fuel and electricity balance equations. The model should prove particularly useful for assessment when electricity tariffs have time-of-use energy or demand charges or when tariffs have conditions which require consideration of strategies over an extended period. The model has been applied to a paper mill which utilises pass-out turbines for cogeneration of electricity and process steam. Results indicate that savings in fuel costs of the order of 400,000 dollars per year could be obtained by implementing optimal strategies for operating the present turbines. Savings of the order of one million dollars per year could be obtained by converting the turbines to give high pressure pass-out steam and then operating optimally.     

Simulating involvement in complex situations

A model for the process of resolution of complex issues involving many participants with differing preferences for outcomes has recently been proposed [3]. This paper describes the use of a simulation procedure based on this model as a means of preparing those involved in a complex issue for their part in its resolution. Details of the simulation procedure are given first. Practical experience with the use of the procedure is described in the latter part of the paper.     

Analysis in support of strategic policy making

Policy making involves choice between alternatives in situations that are usually very complex. The process of choice in such situations is not such that a straightforward analytical procedure can be prescribed yielding a uniquely rational solution. However, decision problems arising in policy making situations can be investigated using methods based on recent developments in game and metagame theory. This methodology can form the basis of a rational decision structure of the type envisaged by Diesing that can be incorporated in the day-to-day work of most organizations.     

Effects of Physical and Emotional Maternal Hostility on Adolescents’ Depression and Reoffending

This study examines whether (1) mothers vary in the way they express hostility toward their delinquent adolescent offspring, (2) different types of maternal hostility differentially affect adolescents’ depression and recidivism, and (3) adolescent depression serves as a mechanism through which maternal hostility predicts later reoffending. The sample consists of 1,216 male first‐time offenders, aged 13–17 years (M = 15.80, SD = 1.29). Confirmatory factor analysis supported the premise that maternal hostility could be distinguished into two subtypes: emotional and physical hostility. Adolescent offenders who experienced emotional or physical hostility by their mothers reported greater depressive symptoms and reoffending 6 months later. Further, the relation between maternal hostility (of each type) and adolescent reoffending was partially explained by depressive symptomology.