Principles and Practice of Scaled Difference Chi-Square Testing

We highlight critical conceptual and statistical issues and how to resolve them in conducting Satorra–Bentler (SB) scaled difference chi-square tests. Concerning the original (Satorra & Bentler, 2001 Satorra, A. and Bentler, P. M. 2001. A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66: 507–514. [Crossref], [Web of Science ®] , [Google Scholar]) and new (Satorra & Bentler, 2010 Satorra, A. and Bentler, P. M. 2010. Ensuring positiveness of the scaled chi-square test statistic. Psychometrika, 75: 243–248. [Crossref], [Web of Science ®] , [Google Scholar]) scaled difference tests, a fundamental difference exists in how to compute properly a model's scaling correction factor (c), depending on the particular structural equation modeling software used. Because of how LISREL 8 defines the SB scaled chi-square, LISREL users should compute c for each model by dividing the model's normal theory weighted least-squares (NTWLS) chi-square by its SB chi-square, to recover c accurately with both tests. EQS and Mplus users, in contrast, should divide the model's maximum likelihood (ML) chi-square by its SB chi-square to recover c. Because ML estimation does not minimize the NTWLS chi-square, however, it can produce a negative difference in nested NTWLS chi-square values. Thus, we recommend the standard practice of testing the scaled difference in ML chi-square values for models M ₁ and M ₀ (after properly recovering c for each model), to avoid an inadmissible test numerator. We illustrate the difference in computations across software programs for the original and new scaled tests and provide LISREL, EQS, and Mplus syntax in both single- and multiple-group form for specifying the model M ₁₀ that is involved in the new test.     

Filipino parents’ school choice and loyalty: a factor analysis

This quantitative study aims to ascertain the significant relationship existing between parents’ profile, and their school choice and school loyalty. Data were gathered using the researcher’s two-part made instrument. Respondents were first asked to fill in a robotfoto for purpose of profiling their baseline characteristics and were later asked to rate listed indicators, using a modified 8-point Likert scale. Using statistical tools such as the ANOVA, factor analysis, correlation, mean, and standard deviation, data were treated indepth. It is interesting to note that although parents are loyal to their alma mater, this does not have a much impact on their choice of school for their children. On the whole, parents’ choice of school is based on its indispensable qualities. Among the three interval-scale profiles of the parent respondents such as age, income, tuition fee payment and number of children, none of these relate to school choice, but age, income, and tuition fee payment negatively relate to school loyalty. From their nominal-scale profiles, significant differences were noted in their school choice when grouped according to civil status, place of work, and terms of payment. With respect to school loyalty, significant differences occurred in their responses when classified according to gender, terms of payment and their active participation in the Parent-Teacher Association (PTA).     

Determinants of Research Productivity in Spanish Academia

This article aims to widen the empirical evidence about the determinants of Spanish academics’ publication productivity across fields of study. We use the Spanish Survey on Human Resources in Science and Technology addressed to Spanish resident PhDs employed in Spanish universities as academics. Productivity is measured as the total number of publications in a three‐year period. We show how personal and academic variables explain differences in productivity within universities and fields of studies and across fields of research. Female workers report lower productivity than their male counterparts, but family responsibilities do not explain this gender gap. The type of contract and tenure or rank do not seem to have any influence on productivity. Researchers seeking professional promotion rather than altruism or personal satisfaction are more productive and young scholars publish more than their older counterparts. Additionally, we find a certain research‐teaching trade‐off and some nuances in the predictors of publication productivity across birth cohorts and fields of study. Finally, international cooperation is one of the most relevant determinants of the number of publications, regardless of the birth cohort. The institutional context in the Spanish research system as regards requirements for promotion and the assessment of research outcomes may contribute to the understanding and interpretation of our results.     

Using the S-STEM Survey to Evaluate a Middle School Robotics Learning Environment: Validity Evidence in a Different Context

Journal of Science Education and Technology - Numerous studies have been undertaken to design, develop, and provide validity evidence for using instruments to measure students’ attitudes...     

Effects of organizational pictures and modality as a feedback strategy on learner comprehension and satisfaction

One of the strongest instructional interventions documented in educational literature is the use of feedback to influence learning outcomes. However, there is lack of empirical research specifically pertaining to the use of multimedia in the feedback message. The purpose of this research was to test whether organizational pictures and modality as a feedback strategy had an effect on learner comprehension and satisfaction. The research design was a 2 Multimedia (Picture Present vs. Picture Absent) × 2 Modality (Narration vs. On-screen Text) × Trial (Trial 1 vs. Trial 2) with Multimedia and Modality serving as between-subject conditions and Trial serving as a repeated measure. One-hundred fifteen university students participated in the study. Findings show statistically significant increases in comprehension scores from Trial 1 to Trial 2 assessments for all four treatment conditions. Learners in the Picture present conditions were statistically more satisfied with the learning experience. Additionally, we used eye-tracking to verify the extent to which the pictures were used in the feedback message. A discussion and recommendations for future research and feedback design are provided.     

Tax incentives for R&D: a critical evaluation

A number of policy proposals have recently been presented that seek to address the problem of declining productivity growth. Many of these are aimed at encouraging private R&D spending through tax incentives. This paper first reviews some of these proposals, then evaluates tax incentives per se as a policy tool for encouraging R&D growth, and finally offers suggestions for redesigning R&D related tax incentive policies.     

Medical knowledge discovery systems: data abstraction and performance measurement

Knowledge discovery systems can be traced back to their origin, artificial intelligence and expert systems, but use the modern technique of data mining for the knowledge discovery process. To that end, the technical community views data mining as one step in the knowledge discovery process, while the non-technical community seems to view it as encompassing all of the steps to knowledge discovery. In this exploratory study, we look at medical knowledge discovery systems (MKDSs) by first looking at three examples of expert systems to generate medical knowledge. We then expand on the use of data abstraction as a pre-processing step in the comprehensive task of medical knowledge discovery. Next, we look at how performance of a medical knowledge discovery system is measured. Finally, the conclusions point to a bright future for MKDSs, but an area that needs extensive development to reach its full potential.