About ten minutes after the eye-tracking experiment discussed in Section 7.1，the participants were required to take part in this test.They were shown 47 elements excerpted from my corpus of the English learning advertisements.Some of the elements（28）had appeared in the previous phase of eye-tracking experiment，and the others not.The participants were told to judge whether they had seen them or not.

7.2.1.1　Participants

The same group of participants as in the eye-tracking experiment in Section 7.1 took part in this test.

7.2.1.2　Material

Altogether 47 elements were used in this test.Most of them（34）were pictures，including photos and drawings.6 were diagrams and 7 were verbal paragraphs.Among them 19 were recognition foils，that is，they had not been presented to the participants in the eye-tracking experiment before the recognition experiment.The categorization of the experiment materials is shown in Table 7.6.All of them were excerpted from my corpus of the English learning advertisements.They were chosen out of the consideration of various variables including content，size，spatial placement，among others.

Table 7.6　Categorization of the experiment materials

The content variable includes two factors—the global coherence（the suitability with the discourse topic—English teaching and learning）and local coherence（semantic connection between the verbal and the visual）.All the visual signs used in this recognition experiment were presented one by one to 11 graduate students（mean age 26.6）at Peking University and their global coherence were rated by them on a five-point Lickert scale ranging from 1 to 5.The mean values were computed and used as the criterion of the global coherence of each picture.The local coherence values of these stimuli were computed on the basis of two independent raters'rating as mentioned in Section 7.1.6.The salience of each stimulus was decided by the overall generalization of several factors working for/against the highlighting of it in comparison to other semiotic resources in adjacency，including the relative size，the spatial location，and the contrast of color.Therefore，in order to obtain quantified data for salience，I and two other graduate students rated on the three items—size，color and spatial location—on a fivepoint scale.The inter-rater reliability was 0.71.The averages of the three items were used as the data for salience.

To minimize the effect of presentation order，three versions of the experiment were designed.Participants were randomly assigned to one of them.

7.2.1.3　Procedure

28 signs which had appeared in the experiment materials in the eye-tracking experiment discussed in Section 7.1 were re-presented discretely along with 19 recognition foils，one at a time，and the participants were asked to judge whether they had seen it or not for each one.The responses were made by pressing the“←”or the“→”key on the keyboard.Specifically，if they judged they had seen a particular sign，they should press the“←”key，otherwise，if they believed they had not seen it，they should press the“→”key.The key pressing ended the presentation of the particular sign and the program proceeded into the next sign.Before the experiment started，the participants were told to make their decisions as fast as possible and should not spend much time on judging the signs.