Two reading literacy/text comprehension tests with different demands (on-line comprehension vs. memory-based comprehension) were administered to 6,104 15-year-old students from all German states and school types. The combined and specific effects of proximal and distal variables from small-scale psychological research as well as from large-scale educational studies on each text comprehension measure were investigated.
Metacognitive knowledge, decoding speed, and the number of books at home (as an indicator for family background) were found to have specific and large effects on on-line comprehension and accounted for 46 percent of the variance with the highest effects for metacognition. Metacognitive knowledge was also highly predictive when the effects of specific prior knowledge and thematic interest on memory-based text comprehension were estimated simultaneously.
In addition, students who showed relative strength in building up a coherent representation of specific texts (memory-based text comprehension) were characterised by high amounts of prior knowledge and thematic interest thereby underlining the important of these student characteristics for learning.
In Early Modern English, double comparatives were often encountered in both spoken and written language. The present article investigates the redundantly marked comparative "worser" in relation to its irregular, but etymologically justified, counterpart "worse". My aim is to examine the diachronic development of the form as well as its distribution in the written language of the 16th and 17th centuries. Two detailed corpus studies are used to reveal the set of parameters underlying the variation between "worse" and "worser", which include system congruity, semantics, and standardization effects. However, the focus here is on the tendency to maintain an alternation of stressed and unstressed syllables, known as the Principle of Rhythmic Alternation. This prosodic principle (which has been argued to be particularly influential in English) turns out to be responsible for most of the results obtained in the analysis of the corpus data.