The reception of subtitling for the deaf and hard of hearing. Viewers' hearing and communication profile & subtitling speed of exposure

Viewers with a hearing loss need to assimilate, process and integrate the information conveyed by the subtitles and the images in order to gain full access to audiovisual content. This process involves a set of complex cognitive mechanisms and the extent of factors that may affect it is large and varied. This study investigates the reception of subtitles for the deaf and hard of hearing among three types of viewers with a different hearing and communication profile: hearing, deaf-sign language users and deaf-spoken language users. The main objective was to investigate whether and how participants’ profile and subtitling speed of exposure affect reception. Three versions of subtitles were analyzed according to its speed of exposure: high (with an average duration of between one and two seconds), medium (between two to four seconds) and low (between four to six seconds); with a different number of subtitles and segmentation and displayed at a rate of 15 characters per second. The analysis was divided into three stages and 72 participants took part in it by watching two videos and answering a set of questions. Participants’ eye movements were tracked and analyzed in terms of attention distribution between the subtitles and the image in order to identify patterns in their reading and viewing behavior (response). Their ability to understand and infer the narrative and to recall and process the verbal and the visual information was analyzed through questionnaires (reaction), as well as their self-assessment on the time and ease of reading the subtitles and viewing the image (repercussion). Background information was obtained through a set of prior and post questionnaires: the former collected data about participants’ hearing, education, language and communication aspects and reading and viewing habits –this part included a test to evaluate their reading skills— that allowed them to be allocated to a profile; the latter collected the opinion and preferences on current practices in subtitling, including speed. The two videos were selected from fiction films and showed different degrees of interaction between the verbal and the visual codes: they had a relatively high load of verbal content (continuous dialogue) but a different load of visual action (number and type of shots). The aim here was to explore how the different combination of verbal and visual information was processed and prioritized. The results obtained in this study confirm that the viewers’ profile affected all the stages of reception. Surprisingly, more differences were observed between the two groups with hearing loss than between them and the hearing group. The use of generalized linear models in the statistical analysis allowed the inclusion of the variable of adjustment “reading skills”, which turned out to play a crucial role on the overall process of reception. Results indicate that differences between profiles might be more due to reading-related factors than to hearing, language and communication aspects. The subtitling speed of exposure influenced the distribution of attention between the subtitles and the image but did not affect significantly reaction and repercussion. Differences in the reception of the two videos were observed in the three profiles and in the three stages of reception, suggesting that reception is highly dependent on the type of audiovisual content used.