predicting the next group of words. They concern themselves with guessing the meaning
of the words or phrases.
Nuttall (1996) stated that readers might start predicting from the title of the
reading text, something that allows them to limit the scope of their reading. Additionally,
while reading, they may hypothesize the message the writer wants to convey and modify
their hypotheses according to what they read in the text. Comprehension begins with
higher levels of processing (making hypotheses), and proceeds to the use of the lower
The Interactive model
This model is built on the interaction of the bottom-up and top-down models.
Nunan (1990), Rumelhart (1977), and Grabe (1991) argue that efficient and effective
reading requires both top-down and bottom-up decoding. L2 readers, for example, may
use top-down reading to compensate for deficiencies in bottom-up reading. To achieve
meaning, they use their schemata to compensate for the lack of bottom-up knowledge
Stanovich (1980) argued that the interactive model is a process based on
information from several sources such as orthographic, lexical, syntactic, semantic
knowledge, and schemata. While reading, decoding processes can support one another in
a compensatory way. If, when reading word by word, readers with good bottom-up skills
do not comprehend the texts, they need to use their prior knowledge (schemata) to assist
them. Alternatively, readers who rely on the top-down model use textual clues and guess
wildly at the meaning, but they need to compensate for deficits such as weaknesses in
word recognition and lack of effective bottom-up processing.