Analyzing Large-Scale Political Discussions on Twitter: The Use Case of the Greek Wiretapping Scandal (#ypoklopes)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript "Analyzing large scale political discussions on Twitter: the use case of the Greek wiretapping scandal (#ypoklopes)" examines Twitter content from late summmer 2022 to early 2023.  This is a descriptive study that examines volume, influential accounts, and extent of polarization in tweets about the Greek wiretapping scandal.  The analysis includes using political followership to characterize partisanship of accounts, identifying account types corresponding to journalists, political accounts, and bots (the latter according to existing software), and performing network analyses such as community detection and node centrality. Among the findings are that the Twitter discussion about the wiretapping scandal was significantly polarized, that omitting journalists, politicians, and central nodes in the analysis results in increased polarization.

 

The dataset and analyses are interesting.  Some comments:

 

--Data access to Twitter has changed significantly since this study was conducted.  In particular, academic access is currently very limited.  I still think the manuscript is useful, but following the authors' analysis as a template is not possible due to these changes in data access.  

 

--Similar approaches have been used previously for analyzing Twitter content, including assessing partisanship, polarization, influential accounts.  For example, see the analysis by Tien et al (2020) on a Far Right rally in the United States; Gallagher et al (2021) for role of a small set of influential accounts in driving Twitter content; Monaci and Persico (2022) on misinformation relating to Covid-19 regulations in Italy; others.  It would be helpful to have additional discussion of previous work, and how the work of the authors fits in with the existing research literature.  In particular, is the approach taken by the authors novel, or is it an assembling of previously used techniques to study a novel topic on Twitter?  It is fine if it is the latter, but this should be clarified.

 

--Some details on the network construction and analysis would be helpful.  For example, are the edges weighted or unweighted (e.g. if A retweeted B 100 times vs. A retweeting B a single time, are these represented differently in the network)?  Are edges directed or undirected?  Line 350 suggests undirected ("...the network is constructed based on bilateral interactions among users") -- if so, what is the rationale? 

 

--The authors mention several centrality measures (PageRank, betweenness centrality, NetShield), and opt to use the least well-known of the mentioned measures.  Is there a rationale for this?  Can the authors give a brief description of NetShield so that someone who does not know this centrality measure can understand it?

 

--Can the authors give some description of the Friedkin and Johnsen polarization metric used in their analysis?  As currently written it is hard to interpret the PI values that the authors report (e.g. does a PI value of 0.4-0.5 reflect a highly polarized environment?).   

 

--There are potential issues with 'hashtag drift' over time (e.g. as described in Tufekci 2014), as the study involves search based upon a set of phrases over an extended period of time.  Some discussion of this would be helpful.

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors argue that the article intends to propose a “methodology for collecting data and analyzing patterns of online public discussions on Twitter”. Nevertheless, there is no discussion about what are the methodological gaps concerning this type of study. Also, they seem to focus more on the case study than on the methodology itself. The article should emphasize methodological aspects instead of results, clearly pointing out what problem these procedures are solving, why they are valid, etc.

Thanks for the opportunity to read your work, I hope my suggestions are useful.

Comments on the Quality of English Language

The authors could make a review to do minor editing concerning English.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors The article brings up an interesting research topic by addressing the public debate on Twitter about a relevant political scandal in Greece. However, there are points where the article could be adjusted so that it can make better contributions to the area.  The first point concerns the literature review, in which it would be important to construct a separate topic in which the state of the art on international literature in relation to discussions on Twitter is presented. As a suggestion of relevant texts in the area, the following are mentioned: Urman, A. (2020). Context matters: political polarization on Twitter from a comparative perspective. Media, Culture & Society, 42(6), 857-879. https://doi.org/10.1177/0163443719876541; Robertson, C. T., Dutton, W. H., Ackland, R., & Peng, T. Q. (2019). The democratic role of social media in political debates: The use of Twitter in the first televised US presidential debate of 2016. Journal of Information Technology & Politics, 16(2), 105–118. https://doi.org/10.1080/19331681.2019.1590283 Literature is also relevant to another point that needs to be addressed in the text: a more solid discussion of results section, in which the author debates how the findings contribute to reiterating or refuting the contributions of literature. The current version of the text is very descriptive, with few interpretations in light of theory. At this point, it would be very beneficial if there was a brief theoretical presentation of the importance of Twitter as a space for public debate, that is very studied in the area of deliberative democracy.  In methodological aspects, the work clearly details how the analysis was constructed and can be very useful for future research.Regarding political inference, however, it is not clear how left and right followers were categorized for each user. This point deserves a better explanation considering the high number of users surveyed.  To sum up, the article needs to focus on improving the literature review and discussion of results based on this literature so that it becomes a very productive contribution to the field of communication.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Thanks for considering my comments. The section you included in the article, "related work", helps to understand how your research dialogues with previous investigations. Nevertheless, I think it should be repositioned and come before your methodology section, not after the results. Furthermore, in the discussion section, you should debate your findings (and your methodology, since you want to contribute to this aspect too) in relation to these previous studies.

Author Response

Please see the attachment

Author Response File: Author Response.pdf