Font Size: 

INTRODUCTION

Computer Assisted/Aided Qualitative Data Analysis Software (CAQDAS) have supported qualitative research and expanded its frontiers for three decades. Traditional qualitative research has dealt mainly with text-based data or transcripts of other media into text format. CAQDAS assist researchers in the process where they interrogate data through iterative procedures of exploration, organization, integration and interpretation ^[1].  From several tools available, presently the most well-known and used are: ATLAS.ti 7, DEDOOSE, DRS, HyperRESEARCH, MAXQDA-11, MiMeG, NVivo10, QDA Miner, Qualrus, Transana-2_6 ^[2]. Many of these tools have broadened their capabilities to work with other units of analysis alongside text: still images, video, audio, web pages and new media.

The last decade has seen a wave of technological innovation that stimulates development of CADQDAS from different flanks. Web 2.0 generates a stream of new content to be studied and shared, in new manners: web based CADQDAS, QDA-wikis, tag clouds, visualization of big data integration and analysis through mashups ^[3].

Another significant challenge for CADQDAS tools is Multimodal studies. These aim at building a unified framework for the study of meaning production under a single theory and methodology, so the multiple resources used in human communication may be interrogated simultaneously. The diversity of multimodal data also promotes choice and development of new tools.

Despite theoretical and technical innovations that have extended study of human communication, there still prevails a description of all modes in terms of their relationship to the linguistic mode and not as means of communication in their own right. A paradigm shift, moving away from a focus on “monomodality” of language towards an interest in analyzing how different ways of interacting may be combined into a "multimodal whole”, reflect the changing nature of communication in the digital era.

New approaches that expand the meaning of what can be recognized as a text, no longer limit to a single message in written language, due to its inability to fully grasp the whole message in many cases. For example, a magazine or any other compound object of writing and images —still or moving— or a TV show, can not be understood solely by a written text; or even to a much greater degree, the phenomenon of a politician’s speech given from a stage at a football stadium in front of thousands of spectators and a great display of special effects, will not be fully comprehended in its meaning only by the script of the speaker. This new understanding of a text includes the definition of practical, everyday instances, where ordinary people build and exchange meanings. For Kress and Van Leeuwen these should now be understood as living texts ^[4].

Rather than using the traditional concept of data coding used in qualitative research, multimodal studies refer to the concept of transcription, a state previous to codification, the process of selecting information from observed reality transforming it into a text. The selection of the unit of analysis is the crux from the multimodal perspective, how to simultaneously transcribe and integrate multiple modes.As already mentioned, the traditional ways of presenting data in writing appear to be insufficient, other dimensions that comprise a physical and social context (such as image, time, space, activities, environments, interactions, objects and participants) must be incorporated to the text so it can be organized, reflected, explored and integrated.

Some argue that there is an interconnection between the tools of representation and other tools that are part of the human repertoire for the creation of meaning ^[5]. They state that the construction of meaning always revolves around social practices in a given context. Just as Schön and others ^[6] have also done, they ^{[5] [7]} analyze the interactions that occur in the context of training architects. For example, the multimodal discourse of a teacher exercising a design critique during a student’s architectural design presentation. The basic unit of analysis in sociocultural research is the notion of mediated action, this implies that the link between people and cultural tools should always be present in the research; people always act through the use of external resources. Therefore, the images produced by the design student are subjected to highly specialized forms of analysis mediated by professional language, face to face interactions, architectural educational context and devices such as: sketches, models, posters, and digital slide-show technologies.

CAQDAS often bring biases from traditional approaches ^[4], such as being oriented solely toward coding (wording) or support exploration of —mostly— hierarchical structures or represent the temporal dimension only in a linear mode (excluding iterative modes). These restrictive features generate tensions. For example, in studies which require analysis of the visual, spatial or temporal content, where loads of information occur simultaneously and unfold in time in a network of interrelationships. Considering new contents and perspectives, some CAQDAS incorporate timelines when the content has a temporal dimension, such as audio or video. They also include features for spatial information, such as sectioning a 2D image into subunits of meaning or visual semiotics and graphic concept modeling.

Other studies reviewed ^[4] describe analysis matrices that remind of film-making tools, such as technical scripts and storyboards, in an intent of simultaneously conveying several dimensions along a common timeline. Orthogonal rows correspond to single units of analysis (time unit, image frame, action, etc.) and columns correspond to the dimensions of analysis in the study (visual frame, visual description, kinetic action, metafunctional interpretation, etc.). While these arrays are a contribution and improvement for sense making, because they allow organizing the many dimensions that a transcription of a multimodal text examines into a common structure, they suffer from orthogonal rigidity that cannot fully account for the expressiveness of each mode and a complex network of interrelationships.

Other studies have gone further and entirely disrupted the orthogonal matrices, introducing tags, pictograms, typographic cues and spatial relationships that incorporate codes from infographic communication ^[4] to generate a much richer, although complex text. There is still much room for even more experimenting, for example with flexible mindmapping structures, with interactive search functions, improvement of user experience or web 2.0 practices ^[3].

New emergent CAQDAS visual resources may also raise questions about laborious preparation, the effectiveness of their communication or the viability of their dissemination in the scientific literature. Therefore they are regarded as developing methodological approaches that will surely inspire the construction of new software tools CAQDAS expanding the existing line.

However, as a synthesis it may be noted that there is still a gap between the What (domain of visualization and dataset) and Why (task of visualization) on the one hand, and the How (visualization and interaction idioms, plus the algorithms) on the other.

Previous research ^[8][9][10] has presented the need for a display mode that could enhance the processes of organization, reflection, exploration and integration of the transcript from an asynchronous multimodal discourse produced by the interactions between participants in a Community of Inquiry over a time period (2-3 weeks each), in the context of two Industrial Design Studio courses. Contents generated by these interactions are hosted in a blog used to mediate the completion of two design assignments. Posts and comments from teachers and students during the assignments use text, images, digital presentations and other digital resources.

 

What new visual idioms in a CAQDAS can enhance the processes of organization, reflection, exploration and integration of the transcript from an asynchronous multimodal discourse produced by the interactions between participants in a Community of Inquiry over a time period, in the context of a Design Studio Course?

 

In an initial phase, this research aims at exploring and defining the domain, dataset and tasks that would be implied in a further development of new visualization idioms that could enhance or complement a CAQDAS.

 

METHOD

Research will be guided the Nested Model for Visualization Design and Validation ^[11] (see Fig.). According to this framework there are four levels nested into each other that account for a process of design and validation of an idiom or characteristic mode of visualization. Domain Situation Level, describes phenomena where data is originated, its context and target users. Task and Data Abstraction Level, defines the proper domain-specific questions and what the visualization aims in achieving for its target users. These two outer levels are concerned with What and Why. A systematic review of existing CAQDAS will describe the affordances of existing CADQAS, strengths and unfulfilled needs for multimodal discourse analysis according to case study.

 

RESULTS

The full paper with present the analysis of the Domain and Dataset/Task Levels that will set basis so future developments may engage with Visual coding/Interaction Level and Algorithm Level.

 

DISCUSSION

The research will raise the background and theoretical framework for the future development of a visualization resource to assist qualitative researchers sense making, from the perspective of multimodality, that current CAQDAS still work out predominantly in a reductionist manner, forcing the semiotics other modes and media, to domination of written language.

CAQDAS; Multimodality; Visualization; Discourse Analysis