CAPÍTULO 3: Diseño de la solución
3.1 Aspectos generales del diseño
Consent to text messages in the present study was obtained at two stages: firstly, through initial informed consent from participants who agreed, either verbally or through the act of submitting them, to submit messages;1 and, secondly, through written consent obtained not only from initial participants but from all message senders where possible (see Appendix 4.2).2 The consent form produced for the current study3 meets specifications outlined by the Data Protection Act and AoIR‘s recommendations.
1 And to select those which they did not want to submit, as explained in section 4.3.4 above.
2 Kan (pers.comm) points out that in their study (How and Kan 2005), only messages sent by participants were used, thus ameliorating the need to gain consent from other senders.
3
A draft of the consent form was drawn up by Malcolm Kendall of the University of Birmingham‘s Information Services team.
Figure 4.1 The consent form used in the current study
I consent to the contents of my SMS text messages between myself and others participating in the research project organised by Ms Caroline Tagg, sent during the period November 2005 until October 2007, being used by Ms Tagg for the purpose of linguistic analysis as part of her research activities at the University of Birmingham. I also consent to the use of any of my personal data necessary for the purpose of analysis in connection with this research. I understand that linguistic findings pertaining to the text messages may be used for a commercial purpose, but that individual messages or personal data will not.
I give my consent on the understanding that all such personal data relating to myself shall be processed in accordance with the Data Protection Act 1998 by Ms Tagg, that the personal data shall be destroyed by the end of 2010, and that, if the contents of the messages are ever published in a dissertation, thesis or other format for any purpose, all such content will be anonymised by Ms Tagg such that neither myself or my correspondents can be identified from that content.
SIGNATURE DATE
NAME
In the majority of cases, it was possible to obtain written consent, but exceptions included messages taken from the AOL website, and those from senders with large circles of fluctuating interlocutors who could not recall who had sent which messages. Hard af Segersteg (2002: 79) similarly notes that consent was obtained in ‗as many instances as possible‘. The argument that message receivers can be deemed responsible for obtaining consent from those who have sent them messages (who are likely to be, according to the literature, their intimates), or of judging whether messages be submitted, can be justified with reference to the guidelines outlined by the AoIR (as mentioned in section 4.4.3 above).1 However, in the present study, personal information regarding participants was not obtained without consent from the individuals themselves. This was ensured by including the questionnaire used to elicit personal information in the same document as the declaration giving explicit consent (Appendix 4.2). In other words, where written consent was not obtained from someone who had sent messages (for reasons explained above), this was justified in two ways: firstly, that consent had either been obtained from the message receiver or that the messages had emanated from a public website; and secondly, that the messages were anonymous in that different senders were not distinguished and no personal
1
This is also the stance adopted by Hard af Segersteg (2002: 210) in the second stage of her study, when participants forwarded all messages, sent and received, to her phone for analysis and where responsibility for obtaining the senders consent was left firmly with the participants: ‗To forward messages that they had received, they [the participants in her study] had to have the original senders‘ permission‘. The intimate relationship assumed between participants and senders also meant that the latter could be briefed by the participants and could raise any doubts with them.
information obtained. With consent obtained from those receiving messages and without accompanying bibliographic information (or even names), it is believed that participants‘ ethical rights are protected.
According to the European Union Data Protection Directive (1995),1 not only must data subjects unambiguously give consent for personal information to be gathered and processed, but they must be informed as to why material is collected. My research goals were explained to the initial group of participants sufficiently for subjects to understand that linguistic form, rather than content, is the focus, as reiterated in the consent form. Documentation of the text message data with personal information ‗necessary for the purpose of analysis in connection with this research‘ was also explicitly consented to, and it was made clear to whom the data will be made available: that while linguistic findings may be used for commercial purposes, neither the text messages themselves nor personal information would be. Senders were also informed that the personal data will only be held until ‗the end of 2010‘. In other words, consent was informed and participants were aware of the research purpose and boundaries.
Participants were also informed that messages would be anonymised so that no individuals could be identified (see section 4.4.3 below). Subjects need to be informed of procedures such as anonymisation not only because it forms part of what they are are consenting to but also, as Rock (2001: 7) points out, because consent is vital to the facilitation of the anonymisation process in that it allows researchers to gauge informants‘ perceptions of what anonymisation entails (Rock 2001: 7).2 As discussed below in section 4.4.3, however, the wording of the consent formmasks certain assumptions regarding the ease with which data can ever truly be anonymised.
The European Union directive also states that data subjects must be able to correct erroneous data and opt-out of data collection. The AoIR similarly frames the obtaining of consent as an ongoing issue, suggesting that, in order to protect their rights to privacy, confidentiality and autonomy as well as informed consent, subjects should be approached at the beginning of research projects and informed if research goals change significantly over time (see Rock 2002: 6). In the present study, to enable participants to correct data or opt
1 Official Journal of the European Communities of 23 November 1995 No L. 281 p. 31.
2
In her own study, for example, Rock (2001: 7), ‗uses consent forms which detail items for anonymisation and invite participants to nominate additional items for anonymisation as they see fit‘.
out, message transcripts were sent to receivers for correction once the corpus had been compiled, and research drafts made available. Although senders were not directly approached, a channel of communication was available through the participants.