DIF Detection and Effect Size Measures for Polytomously Scored Items

Data from a large-scale performance assessment ( N = 105,731) were analyzed with five differential item functioning (DIF) detection methods for polytomous items to examine the congruence among the DIF detection methods. Two different versions of the item response theory (IRT) model-based likelihood ratio test, the logistic regression likelihood ratio test, the Mantel test, and the generalized Mantel–Haenszel test were compared. Results indicated some agreement among the five DIF detection methods. Because statistical power is a function of the sample size, the DIF detection results from extremely large data sets are not practically useful. As alternatives to the DIF detection methods, four IRT model-based indices of standardized impact and four observed-score indices of standardized impact for polytomous items were obtained and compared with the R ² measures of logistic regression.     

A flow and routing control policy for communication networks with multiple competitive users

This paper is concerned with deriving an optimal flow and routing control policy for two-node parallel link communication networks with multiple competing users. The model assumes that each user has a flow demand which needs to be optimally selected and routed on the network links. The flow and routing control policy for each user seeks to maximize the user's total throughput and minimize its expected delay. To derive such a policy, we consider a utility function for each user that combines these two objectives in an additive fashion with preference constants that can be adjusted to reflect the user's own preferences between maximizing throughput and minimizing delay. Additionally, we provide each user with the ability to make these preferences link-dependent to reflect the user's preferences for certain links over others in the network. A condition that depends on the link capacities and preference constants is derived to guarantee the existence and uniqueness of a non-symmetric flow and routing control policy solution which satisfies the Nash equilibrium of non-cooperative game theory. The Nash equilibrium is a desirable solution for such networks because it insures that no user can improve its utility by unilaterally deviating from its Nash control policy. We discuss in detail several interesting properties of this equilibrium and in particular, its relationship to the users’ preference constants. Two illustrative examples are also presented.     

ICT teacher training: Evidence for multilevel evaluation from a national initiative

This paper is one of a pair that re-examines the evidence from a national initiative to train all teachers in England to bring them up to the level of newly qualified teachers, who are required to know when to use and when not to use information and communication technologies (ICT) in their professional practice. Reanalysis of data gathered for the 2004 evaluation uncovers the complexity of such professional development. The effectiveness of contrasting approaches to ICT-related teacher training was analysed using the national survey of 496 trainees and experts' reports on 11 of the 47 training providers. multilevel evaluation of professional development was shown to be robust for ICT teacher training, including a significant correlation between the views of experts and those of teachers. The presence of the middle level of 'organisational support and change' emerged as a particularly discriminating factor, indicating that higher-quality teacher training supports change in the classroom and in the school. Therefore, we recommend that all five of Guskey's levels be consistently adopted for the evaluation of ICT teacher training, and that research should also adopt a multilevel model. A second paper delves deeper to describe and contrast the highest and lowest-rated approaches to ICT teacher training.     

No Joke: A Comparison of Substance in The Daily Show with Jon Stewart and Broadcast Network Television Coverage of the 2004 Presidential Election Campaign

This study examined substantive political coverage of the first presidential debate and the political conventions in 2004 on The Daily Show with Jon Stewart and the broadcast television networks' nightly newscasts. The study found the networks' coverage to be more hype than substance and coverage on The Daily Show with Jon Stewart to be more humor than substance. The amount of substantive information in The Daily Show with Jon Stewart and the broadcast network newscasts was the same, regardless of whether the unit of analysis was news stories about the presidential election campaign or the entire half-hour program.     

Student teachers in the classroom: their perceptions of teaching practice

This paper examines student teachers’ perceptions of the course of school experience in a teacher education programme in Turkey. Data were obtained through interviewing student teachers in Canakkale Onsekiz Mart University, Turkey. In order to analyse the data, Strauss and Corbin’s grounded theory methodology was adapted. The analysis revealed that although the current teacher education programme put emphasis on effective practical training, in some circumstances student teachers feel that their chances of gaining real teaching experiences are minimal. The study suggests that there is a need for further strengthening the existing school–university partnership in teacher education.