Audio description training: A snapshot of the current practices

ABSTRACT

Audio description (AD) is a mode of audiovisual transfer which involves making audiovisual content and live performances accessible to people with sight loss. With the growing demand for this audiovisual translation (AVT) practice, a need has arisen to train professionals in the area. As a result, AD courses have sprung up, both in the form of university courses and professional in-house training. In this article we present a snapshot of the current AD training practices that were reviewed as part the ADLAB PRO project. We report on the results of a survey conducted among 86 AD trainers, involved in the delivery of a total of 192 courses. We first present some general information about the courses and the trainers, and then compare academic and non-academic courses vis-à-vis taught skills, conducted activities as well as competences and transferable skills that are deemed important. The results show that both course types are very practical and do not differ significantly. The only two statistically significant differences encountered were the group size and the importance of vocal skills. IT-related skills, such as using special software, were not prioritised in either of the two types of courses.

KEYWORDS:

• Audio description (AD)
• AD training
• AD skills and competences
• media accessibility training

1. Introduction

Audio description (AD) is a mode of audiovisual transfer which involves making audiovisual content and live performances accessible to people with sight loss. Although in the US and the UK AD has been around for over three decades, it broadened its geographical scope and gained recognition within the audiovisual translation (AVT) field only a little over a decade ago. Since then AD has been developing robustly both as a professional practice and research area. As the demand for AD is on the increase in some countries, partly due to EU legislation requiring its member states to provide a certain amount of accessible content (Audiovisual Media Services Directive of 2010 and the European Accessibility Act of 2019), professional audio describers have to be trained. Thus, new AD courses have been created both as university modules and in-house training.

This article presents the results of a questionnaire distributed internationally among AD trainers operating in both academic and non-academic contexts. The study was conducted as part of the European project Audio Description: A Laboratory for the Development of a New Professional Profile (ADLAB PRO, www.adlabpro.eu), with the main objective of eliciting information about current AD training practices. Below, we first explore AD training research to date, then present more information about the ADLAB PRO project and describe the questionnaire study. We also provide details about the AD trainers who responded to our survey and describe the various roles they perform in relation to AD. Since we elicited information about both academic and non-academic courses, we compare these two to find out if any discrepancies arise between how AD is taught at higher education institutions and in more vocational and business-oriented contexts.

2. Audio description training and ADLAB PRO

As already mentioned, AD is a fairly young discipline, both as a professional practice and an area of academic interest. This also means that AD training is not yet fully fledged, and AD training methodologies and pedagogical approaches are still underdeveloped. Until recently, the literature on the topic had mainly focussed on methods used by individual trainers in their AD teaching. For example, Matamala and Orero (2007) describe an AD module that is part of a postgraduate course on AVT, explicitly mentioning the competences that the module fosters, the expected learning outcomes, the course content and set-up as well as the assessment methods. The authors also propose a range of practical AD activities, from preparatory exercises (such as describing unknown kitchen utensils, fabrics and patterns, facial expressions, etc.) to audio describing video clips of a different nature. Remael and Vercauteren (2007, 75), for their part, focus specifically on teaching film AD and suggest ‘that students must first be taught how to identify crucial filmic clues meant for the original target public’ so that they can include them in the AD script, where appropriate. The authors believe that a good place to start is the exposition phase of films, as they are densely packed with information and, as such, are ‘especially interesting from a didactic point of view’ (Remael and Vercauteren 2007, 77). Snyder (2014), on the other hand, emphasises the importance of teaching students the four fundamentals of AD – i.e. observation, editing, language and vocal skills – and proposes some exercises to practice them. Other AD trainers who have shared their AD activities and exercises in publications include Sadowska (2014), who tackles the issue of describing press illustrations, and Jankowska (2015), who has incorporated the translation of AD scripts into the AD training curriculum.

When it comes to pedagogical approaches to AD, Chmiel, Mazur, and Vercauteren (2019) adopt the situated and contextualised learning models proposed by authors like Kiraly (2000). In their suggested course design procedure, the learning outcomes are aligned with educational materials as well as with learning and assessment methods. Jankowska (2017) describes an AD module that is taught in a blended environment format, with most tasks carried out online. The module is based on nine units, covering topics such as accessibility, regulations, translation vs. writing of AD scripts, guidelines and strategies. All units have a similar structure based on an introduction, lecture, reading assignment, quiz, discussion and tasks. As for the methodology, the activities are analysed from the perspective of Laurillard’s (2012) various types of learning: acquisition, inquiry, practice, production, discussion and collaboration.

Another area that has been discussed in the literature is that of AD skills and competences. Generally, skills can be divided into hard, i.e. related to a specific profession, and soft, i.e. that are not linked to a particular job and are transferable across different professional contexts, as further discussed in section 3.3. While a skill is specific and refers to the ability to perform a given task in a capable manner, a competence is a broader concept and is understood as ‘the proven ability to use knowledge, skills and personal, social and/or methodological abilities, in work or study situations and in professional and personal development’ (EC 2020, online).

In the context of AD, Fryer (2016) distinguishes between three main skills: writing, script preparation and delivery. Matamala and Orero (2007) provide an overview of AD skills and competences by drawing upon authors like Díaz-Cintas (2007), Hyks (2005), Matamala (2006), Navarrete (1997), Orero (2005) and Vidal (2004). Díaz-Cintas (2007, in Matamala and Orero 2007), for example, divides AD competences into four main categories: (i) linguistic competences, (ii) competences related to the content, (iii) technological or applied competences and (iv) personal and general competences. The linguistic competences include an excellent command of the mother tongue, creativity and linguistic sensitivity and the ability to revise texts in the mother tongue. Competences related to the content include knowledge of disability and accessibility issues, cinema language, legislation concerning AD, and the theory and practice of AD. Technological or applied competences refer to computer literacy, use of specialist AD software, ability to conduct appropriate documentation, and vocal and prosodic skills. Finally, personal and general competences encompass encyclopaedic knowledge, ability to analyse, summarise and interpret information, organisation, planning and information management skills, and willingness to work as part of a team.

Despite the various classifications of AD skills and competences, what still remains to be developed, and empirically validated, is an AD competence model akin to the translation competence model developed by the PACTE (2000) group, which includes the following six sub-competences: communicative, extra-linguistic, instrumental-professional, psycho-physiological, transfer and strategic. Obviously, the nature of the intersemiotic transfer that takes place in AD would have to be accounted for in the new model. For example, knowledge of semiotics, which is part of the above-mentioned competence related to the content, would have to be emphasised in lieu of knowledge of two languages, which is stressed in interlingual translation. PACTE’s assertion that the procedural or operative knowledge (i.e. ‘knowing how’) is more relevant in translation than the declarative knowledge (i.e. ‘knowing what’) can be said to be true for AD as it also involves a similar transfer process between the source and target texts, though such a claim would need to be tested empirically.

In 2016, academics and industry professionals from seven European countries joined efforts in a three-year Erasmus+ ADLAB PRO project, 2016–2019, to define the profile of the professional audio describer and to develop a comprehensive AD course. The project had six Intellectual Outputs (IOs, www.adlabpro.eu/results/intellectual-outputs): (IO1) assessment of current AD training practices; (IO2) AD professional: profile definition; (IO3) course design; (IO4) development of course content; (IO5) evaluation and testing; and (IO6) course evaluation, recognition and accreditation. The main deliverable of the project was a web-based course consisting of six modules: (1) general introduction, (2) screen AD, (3) AD of live events, (4) (semi-)live AD and recorded AD for static arts and environments, (5) additional services and (6) additional technical issues, developments and change. Each module is further subdivided into units that contain training materials such as introductory videos, core videos that give a general overview of a given topic (accompanied by transcripts and PowerPoint slides), tasks, reading lists, trainer’s guides and additional videos that zoom in on specific topics (www.adlabpro.eu/coursematerials). Freely available under the Creative Commons licence, the materials are mainly in English, though the core videos in the first module have subtitles in the languages of project partners (i.e. Catalan, Dutch, English, Italian, Polish, Slovenian and Spanish) and the core videos in the rest of the modules have only been subtitled in English. It should be noted that the ADLAB PRO online course is not a MOOC (Massive Open Online Course), as it has been created primarily for AD trainers or course organisers to be used in their respective training set-ups, although self-taught learners may also benefit from the materials. The modular structure of the course allows for flexibility and customisation: trainers may choose to do the whole course, one module or a combination of modules. They may use and/or adapt the materials for use in a classroom, e-learning or blended learning environment. Before designing the course, the ADLAB PRO consortium explored how AD had been taught to date, by obtaining information from AD trainers rather than from (online) descriptions of AD courses. To this end, a survey was conducted, whose design and results are discussed below.

3. Description of the study

The study was in the form of an online survey distributed among AD trainers in the spring of 2017.

3.1. Objectives

The primary objective of the study was to map out training practices in the area of AD and to create a snapshot of the current situation that could shed light on trends in AD training (both in academic and vocational contexts) and highlight any problematic areas or gaps. The study consisted of two parts, quantitative and qualitative, though in this paper we only focus on the former.

3.2. Methodology

The quantitative study draws its results from a survey based on an online questionnaire. In the sections below we discuss its design, distribution as well as the respondents and results.

3.3. Questionnaire design

The questionnaire had separate sections focused on academic and non-academic training.¹ By academic training we mean courses held at universities (as part of bachelor, master’s or specialised postgraduate programmes) that need not be of an academic nature (i.e. they may be vocational). On the other hand, non-academic training is construed to mean instruction held outside of a university setting, such as workshops, in-house training at a cultural institution or a broadcasting company, vocational courses offered by a user association, and the like. Each trainer taking part in the survey could provide information on up to three academic and three non-academic courses that they may have taught in the past three years. This means that a trainer could provide input on a maximum of six courses.

The questionnaire, designed to elicit information both about the instructors and the courses, included five parts: (1) introduction, (2) general questions about the trainer, (3) questions relating to academic as well as to (4) non-academic courses, and (5) general questions on AD-related skills and knowledge as well as transferable skills. The introduction stated the purpose of the questionnaire, specified its structure and gave a time estimate for its completion. The general questions aimed to find out about the AD trainers (researchers, university teachers, audio describers), how long they have taught AD, whether they cooperate with visually impaired persons in their classes and the type of AD they teach (for film, theatre, museum, opera or other live events).

In the case of academic courses, the questions asked about the level at which a given course is taught (BA, MA, specialised postgraduate courses), its nature (separate course/module, part of another course/module), the teaching mode (in-class instruction, remote, blended), its duration, average group size, main skills taught (drafting, translating, voicing, recording of AD, use of specialist software, awareness of the needs of the visually impaired), types of activities (analysis of exiting ADs, discussion of guidelines, presentation of theory, practical exercises) and course requirements (class participation, tests, homework, assessment projects). For the last three questions the respondents could also add their own answers.

A similar set of questions was asked about non-academic courses, the difference being its nature, where respondents could choose between workshop, vocational course, in-house training, one-on-one instruction or other, in which case they were asked to add their own answer. The question about assessment in academic courses was replaced by one concerning certification.

As for general questions, respondents were asked about the total number of students that had participated in the two training types over the last three years. Towards the end of the questionnaire, there were two questions about skills and knowledge related to the practice of AD, on the one hand, and transferable skills, on the other. When it comes to the former, the participants were asked to evaluate the importance of the following: use of mother tongue, choice of the most relevant information to describe, technical aspects (editing, timing, text compression), choice of the appropriate AD strategies (e.g. deciding when to name a character), reflection of filmic language (editing and camerawork) in AD, knowledge of cinematography, theatre, arts and/or semiotics of the image, awareness of the needs of the visually impaired, vocal skills, ICT skills (use of appropriate software). The question was formulated as a list of pre-selected options which the respondents were asked to rate using a 5-point Likert scale, where 1 meant ‘of no importance’ and 5 ‘very important’.

The other skills-related question focused on transferable skills, which are understood as ‘generic capabilities which allow people to succeed in a wide range of different tasks and jobs’ (TA 1990, 5). This means that they are not linked to a specific profession but are related to personality traits, personal attributes and communication and, as such, can be easily transferred from workplace to workplace. According to a report published by the European Commission (EC 2012, 7):

employees need to possess a set of transferable skills, which will equip companies with a more flexible, adaptable and mobile workforce, while at the same time facilitating worker mobility occupationally and geographically and making it easier for workers to develop their occupational pathways.

The question came with a list of pre-selected skills – namely, self-development, teamwork, ethics, efficient work organisation and time management – and the trainers had to assess their importance using a 5-point Likert scale. The aim was to verify whether AD courses equip students with skills that could be applied in other jobs or professional settings, thus helping bridge the university-job market gap and increasing the graduates’ employability (see section 4.1).

3.4. Questionnaire distribution and respondents

The questionnaire, in English, was disseminated online via Instant.ly, a web-based survey distribution tool. The survey was anonymous, but respondents had the option of providing their contact details for potential further collaboration. Translations of the survey into other languages as well as versions accessible for blind people were also provided when requested.

4. Results²

Responses were received from 86 AD trainers who reported on 192 AD courses they currently teach or have taught in the past. These included 93 academic courses (obligatory or elective in university curricula) and 99 non-academic courses (taught outside higher education institutions). Although we wanted to reach AD trainers from as many European countries as possible, the distribution of responses by country was not uniform. The majority of responses came from Spain (31%), followed by Poland (22%), the UK (20%), Italy (9%), 15 other European countries and 10 non-European countries from North America, South America, Asia, Africa and Australia. On the one hand, the reason for this distribution is that it was easy for ADLAB PRO project partners to approach AD trainers in their respective countries. On the other hand, the most represented countries in our pool of respondents are indeed the most active in AD development and, unsurprisingly, offer more AD courses.

When it comes to the language of instruction English ranks first (29%), followed by Spanish (24%) and Polish (20%). The dominant position of English might be because it is used as lingua franca when training multilingual groups of students.

4.1. Respondent roles

Despite the relative youth of AD, most respondents declared to have substantial experience in its teaching: 17% reported more than 10 years of experience, 57% reported 3–10 years and 26% less than 3 years. Yet, although translation, including AD, is a ‘typically vocational activity’ (Kearns 2008, 185), it is often taught in an academic context, which, to some scholars, risks creating a gap between the academically-based theory and the real-life professional practice (Aula.Int 2005; Katan 2011; Klimkowski and Klimkowska 2012; Mauriello 1999). One of the main objectives of the study was to ascertain the profile of the AD trainers, so the respondents were asked to select the labels that best described their role(s). The results are presented in Table 1, where the answers do not add up to 100% because the respondents could select all the roles with which they identified themselves.

Table 1. Respondent roles

Role	No. of responses (86 participants)	Percentage of respondents
University teacher	54	63%
Audio describer	48	56%
Researcher	43	50%
Representative of a company providing AD	25	29%
Representative of a non-profit organisation	12	14%
Voice artist	11	13%
AD user	11	13%
Representative of a broadcasting company	2	2%

Of all the participants, 18% identified themselves as researchers, university teachers and audio describers, while 44% saw themselves as both researchers and university teachers, 13% as university teachers and audio describers and only 3% as researchers and audio describers. This means that almost one third of the trainers have professional experience in AD and teaching, the best possible combination of competences in this context.

4.2. Cooperation with blind and visually impaired persons in AD teaching

Since AD is a service provided to blind and visually impaired audiences and it should take into account target users’ needs and preferences, participants were asked about any potential cooperation they had with blind and visually impaired persons when teaching AD, as illustrated in Table 2, where the total number of responses, 82, is lower than the number of questionnaires, 86, as some participants left the question unanswered.

Table 2. Cooperation with blind and visually impaired persons when teaching AD

How often?	No. of responses (82 participants)	Percentage of respondents
Always	19	23%
Often	20	24.5%
Sometimes	20	24.5%
Rarely	18	22%
Never	5	6%

Most trainers claim to have collaborated, to a greater or lesser extent, with visually impaired persons in the classroom, whereas only 6% acknowledge not to have ever done so. While indicative of the inclusive nature of AD, it also provides students with first-hand information on what the target audience expects from AD and helps them to learn about the needs and preferences of AD users. It would be interesting to investigate further what this cooperation entails and how often blind consultants are present during the actual training sessions.

4.3. Teaching various types of AD

AD includes not only pre-recorded AD for films and TV shows, but also a range of (semi-)live AD for dynamic performances and events (e.g. theatre plays and opera) and (semi-)live and pre-recorded AD for static arts and environments (e.g. museum exhibits). Table 3 lists the types of AD services taught in the classroom.

Table 3. AD teaching by type

AD type	No. of responses (86 participants)	Percentage of respondents
Screen	71	83%
Museums	48	56%
Theatre	35	41%
Other live events	27	31%
Visuals in teaching materials	18	21%
Opera	16	19%
Other	21	24%

As AD is more prominent in film and television (Reviers 2016), it was anticipated that this type would be the most frequently taught and, indeed, respondents taught AD for the screen most often (83%), followed by AD for museums (56%), theatre (41%) and other live events (31%). Over a fifth of the respondents (21%) taught the AD of visuals contained in teaching materials, which is a positive development in increasing accessibility to education for people with visual impairments. Other types of AD identified by the respondents included book and press illustrations, heritage sites, iconic buildings and football matches.

4.4. Comparison of academic and non-academic courses

As mentioned, respondents were given the option to answer a battery of detailed questions on up to six courses taught by them. To compare data for academic and non-academic courses, a series of chi-square tests were performed to find potential statistically significant differences. The courses did not differ as regards their duration, X² (3) = 3.48, p = .32, with approximately 37% of all courses lasting 10–20 hours and 33% of all courses being shorter than 10 hours. A difference on group size was found, X² (2) = 29.77, p < .001, with academic courses being taught to larger groups (10–20 students in 59% of cases) than non-academic ones (fewer than 10 students in 49% of courses).

Table 4 displays the results regarding the various skills taught in both types of courses.

Table 4. Skills taught in academic and non-academic AD courses

Skill	Academic courses	Non-academic courses
AD drafting	87%	91%
Raising awareness of the needs of people with visual impairment	76%	87%
AD voicing	29%	46%
Using AD software	27%	30%
AD recording	25%	35%
Translation of existing ADs into the students’ mother tongue	24%	16%

By far, the two most commonly taught skills are AD drafting and awareness of the needs of the visually impaired audience, included in the curriculum of more than three quarters of both academic and non-academic courses. These are followed by AD voicing, using AD software and AD recording, which were part of roughly a third of the sampled courses, with more presence in the vocational courses. Interestingly, around a quarter of courses teach the translation of existing AD scripts, usually from English into local languages. Given the large volume of AD scripts available in English, their translation has been considered a viable method of AD production, both in terms of costs and time (Jankowska 2015). This skill was taught more often at universities (24%) than outside them (16%), which could be explained by the fact that AD modules are frequently taught as part of (audiovisual) translation courses at universities.

No statistically significant difference was found in the distribution of skills: X²(5) = 5.74, p = .33, where AD drafting was the most frequently practiced skill, followed by awareness raising. Yet, some minor differences can be seen in AD voicing, recording and use of specialist software, which receive more focus in non-academic courses. Arguably, this could be due to the fact that non-academic courses are frequently in the form of in-house training, where participants may have more access to professional recording studios and software to hone these practical skills. In general, numerical differences show that non-academic courses focus more on skills than academic ones, most probably because the latter ones also include a stronger theoretical component.

Other skills mentioned by individual respondents in the open question about skills were specific to theatre and museum AD, such as guiding the blind patrons through space, organising touch tours, using touch objects, live describing spaces, managing technical systems or finding background information on art history.

The activities exploited by trainers in both courses were also compared, with the results presented in Table 5.

Table 5. Activities exploited in academic and non-academic AD courses

Activities	Academic courses	Non-academic courses
Practical exercises	91%	94%
Presentation of AD theory	84%	58%
Discussion of AD guidelines	84%	78%
Analysis of existing ADs	91%	72%

Numerical differences between these activities are not statistically significant, X² (3) = 3.73, p = .29. It comes as no surprise that practical exercises dominate in the courses, followed by the analysis of existing ADs, which can serve as examples and model solutions for would-be audio describers. Discussion of AD guidelines is also frequently exploited as an activity in both types of courses while the inclusion of AD theory sets them apart, with many more academic courses including a theoretical component.

Other activities not listed in the questionnaire but added by the respondents when discussing academic courses included action research projects with partners and stakeholders, translation of existing ADs, peer correction of AD scripts, blindfolded activities and learning to work in small teams. When it comes to vocational courses, additional activities were film analysis, audio introductions, guiding skills, feedback from visually impaired trainers, and working in small groups with blindfolds to appreciate what it is like to rely solely on verbal information.

As for the AD-related skills and knowledge as well as transferrable skills deemed important by trainers, Table 6 compiles the results, based on a 5-point Likert scale, where 1 means ‘unimportant’ and 5 means ‘very important’. The total score represents the significance of individual skills or knowledge types as scored by all 86 trainers participating in the survey. The scores for academic courses are calculated on the basis of data given by 26 lecturers with experience in offering academic courses, whereas the scores for non-academic courses are calculated on the basis of data obtained from 25 trainers. A series of t-tests were performed to compare the scores for each competence and the p-values in the last column are marked with * if statistically significant (the significance level being < .05).

Table 6. Scores for importance of competences

Skill or knowledge	Total score	Academic courses	Non-academic courses	t value	p value
Choosing the most relevant information to describe	4.87	4.80	4.84	−0.36	.72
Perfect use of mother tongue	4.70	4.84	4.46	1.92	.06
Knowledge of the needs of the visually impaired	4.65	4.16	4.69	−0.53	.59
Choosing appropriate AD strategies (e.g. deciding when to name a character)	4.46	4.38	4.46	−0.48	.63
Technical aspects (editing, timing, text compression)	4.32	4.11	4.50	−1.75	.08
Knowledge of cinematography, theatre, arts and/or semiotics of the image	3.84	3.65	3.92	−1.18	.24
Vocal skills	3.52	3.07	4.00	−2.70	.009*
ICT related skills (using appropriate software, etc.)	3.51	3.30	3.46	−0.63	.52
Reflecting filmic language (editing and camerawork) in AD	3.22	3.03	3.15	−0.36	.72

As the results show, trainers identify the importance of the listed skills and knowledge types on a similar basis for both types of courses. The only statistically significant difference pertains to the significance of vocal skills, which are deemed to be more important in non-academic (4.00, i.e. important) than in academic settings (3.07, i.e. neither important nor unimportant, according to the applied Likert scale). Five other competences are judged as (very) important in both types of courses, scoring above 4, namely: choosing the most relevant information to describe, perfect use of mother tongue, knowledge of the target audience’s needs, choosing the appropriate AD strategies and technical aspects (such as editing, timing and text compression). The rest of items scored between 3 and 4 and none were identified as unimportant.

Acquiring vocal skills is considered to be one of the least important skills, despite Snyder’s (2014) claim that it is one of the four AD fundamentals. This may be due to the fact that, in some countries, it is not customary for audio describers to voice their descriptions, which is usually done by professional voice talents, especially in the case of AD for films and television productions (Chmiel and Mazur 2014). Vocal skills may be more useful in other AD types, such as theatre or opera, where describers are often present during the live performance and read out their descriptions.

ICT related skills received one of the lowest scores (3.51) of all the listed competences, irrespective of the type of course. In the case of non-academic settings, and according to Bernd Benecke (personal communication), a professional audio describer employed by the Bayerischer Rundfunk in Germany, this might be due to the fact that working with specialist software is not very common in the industry, whether freelancing or being part of an in-house team. In the case of academic courses, universities might not have access to licenced professional software (such as Starfish Technologies and Miranda Softel Swift ADePT) because of their high cost, though free web-based services already exist (YouDescribe and LiveDescribe). Another reason why the use of AD software is not considered important by many trainers is that the aforementioned systems work well when producing screen AD but not so with other types of AD, especially static or live ones. It will be interesting to monitor potential changes in the near future as technology becomes increasingly more pervasive, and projects like ADLAB PRO offer users the chance of experiencing online audio description editors.

The lowest score, given to the skill of reflecting filmic language in AD scripts (3.22 of 5), may be due to the lack of consensus in the profession as to how to go about it, as well as to the fact that this topic only applies when teaching the AD of films and TV productions.

As for the transferable skills, the total score is based on answers obtained from all trainers (N = 86), the score for academic courses is based on answers given by trainers with academic experience (N = 26) and the score for non-academic courses is based on replies from trainers with vocational training experience only (N = 25). Table 7 presents the scores and statistics, where no p-values are statistically significant.

Table 7. Scores for the importance of transferrable skills

Transferable skill	Total score	Academic courses	Non-academic courses	t value	p value
Efficient work organisation and time management	4.45	4.30	4.50	−0.89	.37
Observance of ethics	4.29	4.26	4.23	0.17	.86
Self-development	4.15	4.23	4.15	0.35	.72
Teamwork	4.11	3.96	4.11	−0.52	.60

None of the four transferable skills listed in the questionnaire differ statistically between academic and non-academic instruction. All trainers ranked the skills in the following order: efficient work organisation and time management, ethics, self-development and teamwork. The first two are essential, as describers often work under extreme time pressure to meet tight deadlines. Teamwork is also important and, in countries like Germany, especially for ‘big productions’ (Benecke, personal communication), audio describers work in small groups, including a colleague with sight loss and a second describer. They must thus be able to deal with criticism, accept other points of view and defend their AD choices where necessary. Nonetheless, it is often the case that AD remains a solitary activity and self-development activities are crucial for the describer to stay abreast of the latest developments. Our results show no major gap or difference between teaching AD in academic institutions or in more vocational contexts outside the academia.

5. Limitations of the study

Although surveys are an effective tool for gathering large amounts of data from a whole range of participants, they are not devoid of disadvantages. Dillman and Bowker (2001) mention four types of errors that may affect this methodology: coverage, sampling, measurement and non-response. A coverage error occurs when some members of a population are excluded from the sample, whereas a sampling error is incurred as a result of the characteristics of a population being calculated based on a subset (sample) of a population. A measurement error is due to flaws in the way data are collected (e.g. imperfect scales, wrong wording of questions). A non-response error occurs when the selected respondents in a sample fail to reply.

All these errors can, to a greater or lesser degree, be applicable to our survey. First, although the survey was distributed internationally (in Europe and beyond), most responses were elicited from project partner countries, thus potentially giving rise to a coverage error. Second, a sampling error could have occurred as well, since the selection of AD trainers was random and not guided by their specific profiles. Third, a non-response error also occurred as some of the participants that had been contacted did not fill in the questionnaire and others filled it in only partially. Finally, the fact that some of the questions, including those on skills, were based on a pre-determined list of answers and a Likert scale was used to grade the responses could lead to measurement errors, as the answers were based on the respondents’ subjective judgement. Yet, the Likert scale remains the standard measurement for this type of questions and a list of pre-selected items, though limiting, serves the purpose of facilitating the quantification of the responses and making subsequent generalisations. As many datapoints used in the statistical analysis came from subjective judgements provided by the respondents, their significance might not fully delineate differences between academic and vocational courses.

Despite their limitations, surveys do remain an effective tool for gathering data from selected populations. Online surveys like ours offer the added benefit of reaching many participants simultaneously and in different geographical locations. For Plumb and Spyridakis (1992, 626), a survey also reinforces ecological validity as it ‘asks questions of real people in real situations’. All in all, a fairly representative cross-section of AD trainers has been included in the study and valid and interesting data have been elicited from them, thus helping to paint a picture of the current AD training situation.

6. Conclusions

The results obtained in the survey indicate that the AD courses on offer are very much practice-oriented, fairly short and serving rather small groups of students. Overall, the two course types, academic and non-academic, were not that much different and only two statistically significant differences were found: groups tended to be smaller in the non-academic training set-up, which could be explained by the in-house nature of such training in many cases, and the greater importance given to vocal skills in non-academic courses, in the form of AD voicing and recording. The latter could be explained by the more vocational nature of non-academic courses and, again, the fact that they are conducted for in-house purposes. What came as a surprise, especially given the ubiquity of technology nowadays, was the little attention paid to ICT related skills, both in the academic and the non-academic setups. This may be because teaching AD drafting is considered more important, especially taking into account the short duration of the courses. Additionally, universities may not have access to expensive professional software, or it may be the case that AD trainers do not know themselves how to use the software. However, as technology becomes more commonplace and affordable, it is expected that it will feature more prominently in future AD training.

The survey described in this article was part of the first Intellectual Output of ADLAB PRO and its results were used to inform subsequent stages of the project, especially the development of the course structure.³ Drawing on the findings, the decision was taken to design the ADLAB PRO course focusing on screen AD, (semi-)live AD of dynamic performances and events, and semi-live and recorded AD for static arts and environments. Furthermore, the skills and knowledge deemed most important by the trainers helped draft the main competence framework and domain-specific sub-competences for the course and, on the basis of the little prominence given to technology in existing AD courses, the consortium decided to feature a free online AD editor on its website to make it easier for trainers to incorporate the technological component into their classrooms. It should, however, be noted that the ensuing course is not directly grounded on the trainers’ feedback, as a separate design procedure was followed (Chmiel, Mazur, and Vercauteren 2019). What is more, the course is not meant to replace other existing training provision, but rather help instructors to enhance the courses they already teach or, as the case may be, create new ones, by taking advantage of the ADLAB PRO course’s flexible structure and the wealth of customisable educational materials on offer. The hope is that in so doing they can better equip future professionals with the skills necessary to successfully operate in the rapidly evolving AD market.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work has been supported by the Erasmus+ programme of the European Union under Grant [2016-1-IT02-KA203-024311]. The European Commission support for the production of this publication does not constitute an endorsement of the contents, which reflect the views only of the authors, and the Commission cannot be held responsible for any use which may be made of the information contained therein.

Notes

1. The questionnaire can be accessed at: https://www.adlabpro.eu/wp-content/uploads/2017/12/20161104_uam_io1_questionnaire_en.pdf

2. Due to limited space we present a selection of the results. All results are available in a report at: www.adlabpro.eu/wp-content/uploads/2017/12/20170608_uam_io1_report.pdf

3. www.adlabpro.eu/wp-content/uploads/2018/11/Simplified-course-structure-IO3-1.pdf