Tainá de Almeida COSTA; Eunice Almeida da SILVA e Régia Cristina OLIVEIRA
Estudos de Sociologia, Araraquara, v. 28, n. esp. 1, e023008, 2023. e-ISSN: 1982-4718
DOI: https://doi.org/10.52780/res.v28iesp.1.17375 13
was used because some posts published in the groups were about other subjects. Using this
filter, it was possible to define the sample to be analyzed more assertively, ensuring that the
posts made some kind of mention of vaccines; 3) On the left side of the page, a filter was
selected with the year referring to the data to be collected, in the 'date of publication' field. In
the case of this study, the year analyzed was 2020.
For data collection and recording (text, images and videos), it was necessary to open the
post, copy the link to the post and paste it into the Google spreadsheet where the collected data
was stored. Then, the following information was included in the spreadsheet: collection date,
post text, publication date, post engagement numbers (comments, reactions and shares), type of
media used in the post (image, video, link), text in the comments, name of the author of the
publication, link to the author's profile and specific reference to some type of vaccine. In
addition, the spreadsheet had two other columns that were used later in the content analysis of
the posts: key narrative and observation. For this content analysis (BARDIN, 2016), the
following phases were covered: pre-analysis; material exploration and treatment of results;
interference and interpretation.
The first phase has three “missions”: the choice of documents to be submitted for
analysis, the formulation of hypotheses and the elaboration of indicators that support the final
interpretation. This was done during the first exploration of Facebook groups, through floating
reading, as called by Bardin (2016), who highlights that this process aims to enable the
researcher to “know the text by allowing himself to be invaded by impressions and orientations”
(BARDIN, 2016, p. 126, our translation).
Only publications containing the keywords 'vacina' (vaccine) or 'vacinação'
(vaccination) were considered, without exception, in compliance with the exhaustiveness rule
(BARDIN, 2016). The sample was determined by Facebook's own search tool, following the
representativeness rule. Posts found within anti-vaccine groups on Facebook were chosen,
following the rule of pertinence. According to this rule, the retained documents must be suitable
as a source of information, in order to correspond to the objective that raises the analysis. All
these rules are part of content analysis (BARDIN, 2016).
After this step, the indexes were referenced, which served as the basis for phase two,
material exploration. The indexes used in this study followed the typology elaborated by Smith,
Cubbon and Wardle (2020, p. 9, our translation) that considers the following anti-vaccine key
narratives: 1) development, provision and access: posts related to continuous progress and
challenges of vaccine development. This also includes posts related to the testing (clinical trials)