De Glopper and Horváth (1996) conducted a mother tongue validity survey, in which they collected data on the frequency, importance and perceived personal mastery of communicative tasks related to work, study and citizenship, in varied language functions, from people with a GCSE, from teachers in secondary education and from employers. The ways in which these three sub-samples evaluated literacy tasks by the above mentioned criteria show rather high agreements (r>0.78). Although little has been published of the reading-related fi ndings, it is known (Horváth, 1996) that self-reported reading problems include forms to be fi lled in, instructions and argumentative texts; that Hungarians tend to underestimate these problems when they compare them to their knowl-edge gaps in literature; and that employers attribute a higher importance to orality. (There are no data regarding whether the latter may be a cul-tural phenomenon or a consequence of the literacy problems mentioned.) The communication problems manifested from the responses were re-lated to the public sphere. Horváth observed that subjects attributed these to their own mastery problems.
Giving an overview of adult literacy studies, Vári et al. (2001) call for a variety of texts, including the introduction of non-traditional, non-lit-erary genres in the development of reading literacy at school. All these would lend support for the inclusion of reading purposes and topics to be included in diagnostic assessment, too, of course through texts that are appropriate for the given populations in both content and language.
Assessment standards typically focus on knowledge and skills, and not on aspects of motivation. Yet in the light of the above, it may be worth considering the inclusion of a reading motivation component in the standards, even if its assessment would be, at present, not feasible.
as expository texts). The notion of text types is not based on well-de-fi ned, clear-cut categories. Instead, the categories are fuzzy and they overlap. Whereas some authors distinguish only a few text types, others differentiate among 30 or more types (Biber, 1989). We fi nd it useful to distinguish text types according to the following functions: narration, report, exposition, persuasion, instruction, and entertainment.
Narrative texts convey information about events in time and space due to actions of individuals. They inform us about when something hap-pened, how, why and in what sequence. They are usually given by some-one from a specifi c point of view with subjective selectivity and empha-sis. Reports also convey information about events in time and space due to actions of individuals, and they, too, inform us about when, how and in what sequence something happened. However, the information is given within a more objective frame of reference. Expository texts provide information about the structure of an object or system, about how the elements are interrelated, what changes can occur in the system and why.
Persuasive texts convey arguments that explain why something should be considered as true or why something should happen. Instructional texts provide directions on what should be done and how it should be done; examples are recipes, how to give fi rst aid, manuals on how to operate software, apply social regulations or rules. Other documents such as novels, essays, letters, or reviews can be subsumed under these main text types or can be considered as combinations of them. Texts dealing with the interpretation of texts are basically texts about texts (‘meta-texts’) and may be necessary to assess the reader’s refl ection about a text. The various text types are not necessarily associated with different reading diffi culties. Nevertheless, the notion of text types is important to cover the full scope of reading literacy.
Formats and Combinations of Documents
Reading literacy refers to written documents usually consisting of verbal text, but often also augmented by graphics such as realistic pictures (in-cluding maps), diagrams or graphs. Texts, on the one hand, describe their content by referring to the intended objects or events by names, by further specifying these objects or events through attributes and by specifying
relations between these objects or events. Graphics, on the other hand, employ a fundamentally different representational principle: They dis-play structural analogues of their depicted content. Although they can also include verbal labels, these labels (usually names) only help to map the graphical entity on to the corresponding verbal signs in the text; but do not describe the depicted content. Description is given in the body of the text or in explanatory notes, often provided after the caption of the graphic.
Tables are a further format within documents. In the majority of cases, they use verbal symbols (words, phrases) for the specifi cation of col-umns and rows and as entries for the cells of the table. However, in some cases, these specifi cations (especially those within cells) can also be pic-torial (as, for example, in a table that indicates the appearance of differ-ent groups of animals). The differ-entries of cells in a table usually provide the building blocks for propositional structures. The relational framework for these propositions (their predicates) is usually defi ned elsewhere – often in the heading or as verbal explanation of the table.
In order to assess reading literacy, the documents to be read should include a variety of documents with varying formats, including pure texts with no other format, combinations of text and realistic pictures, maps, diagrams or graphics and tables3.
Reading in everyday life often requires usage of multiple documents (e.g., two texts with adjunct tables, diagrams or graphs). Although read-ing literacy in grades 1 to 6 is not yet at the level of grade 10 (the grade associated with PISA-assessments), it should be made clear in learning to read, that reading frequently requires the combined use of more than one document in order to fulfi l a specifi c task. Accordingly, some reading tasks in the assessment of literacy should involve more than one docu-ment. These texts have generally been generated independently and they also make sense independently. They are only put together accidentally or for some specifi c purpose (e.g., for the purpose of test), and they may complement each other or may contradict each other.
3 In the PISA 2009 framework, these combinations are referred to as ‘mixed texts’ (i.e. mixture of continuous with so-called non-continuous tests).
Technologies
Until recently, reading was (and still is) so much associated with reading printed documents that most people are not aware of alternative tech-nologies for reading. However, a historical view back to the use of clay tablets or papyrus scrolls a few thousand years ago reminds us that writ-ten documents bound together in the form of books is only one specifi c technology used for written communication, complemented by black-boards, whiteboards or fl ipcharts used for hand-written writing. Since the parts of a print document in sheets, brochures, magazines, books etc. are physically connected, their interrelation and associated access-structure to the specifi c parts of the text is highly transparent, because the infor-mation space is equivalent to the physical space. Printed text as the major part of printed documents suggests processing the presented information in a specifi c sequence corresponding to the linear surface structure of the text. Printed documents allow a higher amount of information to be pre-sented at one time and they allow the reader to underline or to make annotations without much effort.
Nowadays the ubiquity of new information technologies has made reading electronic documents displayed by computers and web-based technologies such a widespread phenomenon that it can no longer be ignored. In electronic documents, the interrelation of the different parts of the text is mostly much more complex, more fl exible, and less trans-parent. Accordingly, navigating in the information space and using the access structure of the document is much more demanding for the reader.
The borders between print documents and electronic documents are blurred. A document in an electronic medium can be printed out as a pdf fi le, and documents written on paper can be electronically encoded for transmission by fax and then recoded in print for the receiver. In order to defi ne print documents and electronic documents in a most distinctive way, we will apply the term ‘electronic documents’ only to hyper-docu-ments with navigation tools as mentioned previously.
Hyper-documents allow and require navigation. Whereas the sequence of reading itself is always necessarily linear (i.e. there exists no ‘non-linear reading’), readers have multiple ways of ‘non-linear reading at their disposal. Although electronic documents with a complex network-like structure do not suggest processing the presented information in a
spe-cifi c sequence, it is possible that some elements on the screen have a higher salience than others, which may suggest one should enter the document at this point (Kirsh, 2010). Due to the reduced visual percep-tibility of information, less information can be displayed at any time in electronic documents: screens contain fewer words than printed pages.
Accordingly, readers of electronic documents (everything else being equal) have to keep more information in working memory and have to retrieve more information from long-term memory for successful coher-ence formation than readers of print documents. Put more simply: readers of electronic documents must cope with reduced visual readability and with piecemeal presentation of information. Finally, electronic documents make underlining and annotations less easy or even, when the display software does not allow personal copies with included annotations, im-possible.