This paper explores three simple data manipulation techniques (synthesis, augmentation, curriculum) for improving abstractive summarization models without the need for any additional data. We introduce a method of data synthesis with paraphrasing, a data augmentation technique with sample mixing, and curriculum learning with two new difficulty…

This paper explores the effect of using multitask learning for abstractive summarization in the context of small training corpora. In particular, we incorporate four different tasks (extractive summarization, language modeling, concept detection, and paraphrase detection) both individually and in combination, with the goal of enhancing the target…

The vertical scales of large-scale achievement tests created by using item response theory (IRT) models are mostly based on cluster (or correlated) educational data in which students usually are clustered in certain groups or settings (classrooms or schools). While such application directly violated assumption of independent sample of person in…

A basic question of instructional interventions is how effective it is in promoting student learning. This paper presents a study to determine the relative efficacy of different instructional strategies by applying an educational data mining technique, learning decomposition. We use logistic regression to determine how much learning is caused by…

This paper helps researchers develop criteria for conducting surveys. Cost considerations frame the paper's suggestions. Specifically, the text advises that researchers consider the human and material resources needed, type of survey instrument, medium for conducting the survey (mail, telephone, personal interview, etc.), data analysis…

Two relatively new tools for analysis of data compiled in evaluation studies are presented. The National Test-Equating Study in Reading, known as the Anchor Test Study, produced tables of score-correspondence between the eight reading comprehension and vocabulary tests most widely used in the United States. Two types of tables from this report…

There are two distinct but related purposes for carrying out a "discriminant analysis": (1) discrimination, and (2) classification. The primary objective of this paper was to review the outputs of selected computer programs often used to carry out a "discriminant analysis" with respect to these two purposes. Information provided by the programs on…

The implications of data from a review of ten years of the American Educational Research Journal (AERJ) indicating that random sampling is rare and that there is increased use of quasi-experimental designs lacking in random assignment are considered. It is suggested that tests of significance could be abandoned or at least placed in a subsidiary…

Research with older persons suffering from physical illness presents numerous challenges to gerontologists. Issues of conceptualization pertaining to the definition of illness, its location in the research paradigm, and the context in which illness occurs must be addressed prior to dealing with methodological problems. Access to physically ill…

The special characteristics of the data from the National Assessment of Educational Progress (NAEP) that affect the validity of conventional techniques of statistical analysis are considered. In contrast to the assumptions underlying standard methods of statistical analysis, the NAEP samples are obtained via a stratified multi-stage probability…

This paper investigates the utility of various procedures used to locate and interview veterans as part of a longitudinal research study being conducted for the Department of Defense. The populations are comprised of below entry aptitude standards males who entered the military during the late 1960s, and potentially ineligibles who entered between…

Levin's (1975) sample-size determination procedure for completely randomized analysis of variance designs is extended to designs in which antecedent or blocking variables information is considered. In particular, a researcher's choice of designs is framed in terms of determining the respective sample sizes necessary to detect specified contrasts…

There has recently been a tremendous increase in the quantity of research published in gerontology. While the increase in the quantity of research has been regarded positively by gerontologists, there has been little discussion or evaluation of the quality of the research. Two articles that appeared in 1975 were both very critical of…

Five three-variable combinations were obtained from a variety of standardized test scores of 331 male children. These five groups were treated as unrestricted populations of the case III variety. Extreme univariate selection was produced in each group allowing for tests of differences between the correlation values of the two incidental variables…

The issues and some proposed solutions regarding Follow Through (FT) site variability are examined with a review of developments in FT evaluation. The role of adjusted site means with differences within sponsors and between sponsors and background characteristics is discussed to determine whether adjusted means are the preferred measures of model…