What do international student assessment studies measure? School performance, student abilities, cognitive abilities, knowledge or general intelligence?

What tests measure, like tests of international student assessments (PISA, PIRLS, TIMSS), is possible to describe by analysis of tasks and analysis of empirical relationships. Results of content-based task analysis show that items of different scales are very similar: Reading literacy items contain tables and graphs, science items extensive texts, math items common cognitive tasks with long texts. Conceptually most difficult to separate from intelligence is problem solving. Thinking, general knowledge, and school knowledge are necessary in different extent in each domain for solutions. Empirical relationships in form of correlations between used scales, with intelligence tests and results of factor analysis are difficult or even not possible to find in published reports. Correlations between scales on level of individual data are higher than known correlations between scales of intelligence tests. Correlations between scales on level of state data are extremely high; correlations between studies and with results of intelligence tests on level of state data are high. Discriminant validity is rarely found. All results indicate a strong g-factor of cognitive abilities. On level of macrosocial data a separation from intelligence is empirically not possible.