Available via license: CC BY-NC 2.0
Content may be subject to copyright.
Observatorio (OBS*) Journal (2023, Vol 17, nº2), 29-51 1646-5954/ERC123483/2023 29
Copyright © 2023 (Bacaksizlar Turbic and Martins Rosa). Licensed under the Creative Commons Attribution-
NonCommercial Generic (cc by-nc). Available at http://obs.obercom.pt.
Political Presence on Facebook during and beyond Election Campaigns: The
Portuguese Case
N. Gizem Bacaksizlar Turbic*, Jorge Martins Rosa**
* Postdoctoral Researcher, Chair for Computational Social Sciences and Humanities, RWTH Aachen University,
Germany (gizem.turbic@cssh.rwth-aachen.de)
** Associate Professor, Communication Sciences Department, NOVA FCSH and ICNOVA, Portugal (jmr@fcsh.unl.pt)
Abstract
Motivated by the conjecture that traditional political communication patterns and strategies are broken
within social network sites, this paper investigates the presence of Portuguese political parties on
Facebook during a year with two major elections (European and Legislative), both from the posting
activity aspect and the user engagement. The following research questions are formulated: Is it possible
to identify distinctive styles adopted by different parties or to demonstrate that their audience
acknowledges them in a different manner? If such distinction is possible, can it be correlated with the
party’s foundation time, relative size, or positioning within the political spectrum? Our analysis shows
that the common assumptions that older parties do not behave like newer ones fail when tested against
the collected data. We also observe higher levels of user engagement attracted by newly founded,
minor, and right-leaning parties, compared to their older, major, and left-leaning counterparts,
respectively.
Keywords: social media, European elections, national elections, user engagement, Facebook
Introduction
Since the emergence of Web 2.0 in the early 2000s, a whole range of online platforms, such as microblogging
services (e.g., Twitter) and social network sites (SNS) (e.g., Facebook) have allowed Internet users to
express their opinions and emotions (Stromer-Galley, 2003; Stieglitz and Dang-Xuan, 2013). With the
continuous rise in the use of online communication by the general public, political parties are becoming
increasingly present on social media platforms (Macnamara, 2011; Enli and Skogerbø, 2013). Particularly
after Barack Obama and his Chicago-based campaign team used social media and technology as a part of
the arsenal for his campaign in the 2008 United States elections, SNS came into the attention of political
actors across the world as a viable, if not indispensable, channel to spread their messages to the public.
This presence has given new roles not only to political parties but also to the citizens. The Internet users
began sharing their insights on micro-blogging channels, most notably on Twitter, and Facebook has turned
into a medium for sharing socio-economic, political, and healthcare related concerns and emotions (Kapoor
et al., 2018).
30 OBS* Journal, 2022, 17(2)
The online conversations and interactions on social media provide an exhaustive source of data on political
engagement, opinion dynamics, and online community structures compared to the limited traditional
sources, such as interviews and surveys. Manual analysis of social media data becomes impossible due to
its massive scale; hence scholars have started looking into the tools and methods from Computer Science
for inference from large data sets (Kapoor et al., 2018). Researchers aim to grasp the extent, the peculiarities
and eventual repercussions of social media in the political sphere, either as empowering devices for the
common citizen, as marketing tools for politicians, or even as unpredictable agents of change in democracy
(Bennett, 2012; Lilleker et al., 2015).
The topics of political communication and public engagement on Facebook, and content analysis of posts
during the political campaigns, have come into the focus of the emerging field of social media research
(Kapoor et al., 2018). As the most recent examples of studies on these topics related to European politics,
Larsson (2016) investigated the traces of permanent campaigning in Norway and Sweden, two politically
similar countries, while Kalsnes (2016) looked into the 2013 national elections in Norway, where the data
obtained from interviews with political communication directors were combined with those extracted from
Facebook. Taking on a similar subject, Magin et al. (2017) used Facebook data to analyze the 2013 national
elections campaigns led by German and Austrian parties. Boulianne and Larson (2021) explored the
complementary dimension of the users’ engagement with posts during the 2019 Federal Elections in Canada.
More recently, Jensen and Schwartz (2022) give an overview of research on social media and elections in
the last decade.
All of these studies concentrate only on the electoral campaigns and analyze social media data, i.e.,
interactions and posts from political parties and politicians, while neglecting the data from before and/or
after these electoral periods. On the other hand, Larsson and Kalsnes (2014) studied the general Facebook
use by the parliamentary politicians in Norway and Sweden, not related to political campaigns. Similarly,
Heiss et al. (2019) collected Facebook data from the national political figures in Austria for a non-electoral
period of six months, and Serra-Silva et al. (2018) studied the Portuguese parties’ Facebook engagement
over a seven-year span. However, the number of studies expanding their analysis to the pre- and post-
electoral campaign periods is limited. Also, the studies on the online presence for the political parties in the
Southern Europe are limited. The research field has been mostly applied in the Western Europe and North
America (Stieglitz et al., 2012). Moreover, except for the work of Serra-Silva et al. (2018), the political
communication on social media in Portugal has remained largely unacknowledged. Therefore, we aim at
addressing this gap by conducting an analysis of social media presence of Portuguese political parties on
Facebook.
In this paper we investigate the activity on social media by the active Portuguese parties before, during and
after the two major elections in 2019, i.e., the European and Legislative Elections in Portugal. The differences
and similarities in the behavior of the parties from different positions in the political spectrum and with
different foundation times are analyzed based on their Facebook posts and the corresponding interactions
from the users. The collected Facebook data are tested against the following three hypotheses, the first two
being previously considered in literature.
• (H1) Older political parties have a harder time engaging on social media platforms (Effing et al.,
2011).
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 31
• (H2) Major parties do not depend on social media as much as minor ones, as they receive more
mainstream media attention (Vergeer and Hermans, 2013; Larsson, 2016).
• (H3) The rise and political success of right-leaning parties is correlated to more social media
presence and user engagement.
These hypotheses are addressed both from the perspective of the
supply side
(the starting point of the
communicative exchange, herein corresponding to Facebook posts) and the
demand side
(the follow-up, i.e.
user interactions in the form of reactions, comments, and shares).
The first hypothesis (H1) is established following recent studies (Effing et al., 2011; Lilleker et al.,
2015; Magin et al., 2017) reporting that older parties have a more conservative approach to social media,
being lower on their priority tools to convey political messages and goals. Testing this hypothesis from both
aforementioned perspectives yields two underlying (sub)hypotheses:
• (H1.a) Newer parties post on Facebook more frequently than the older ones.
• (H1.b) Newer parties draw higher levels of engagement from the users.
The second hypothesis (H2) is formulated based on studies (Vergeer and Hermans, 2013; Larsson, 2016)
arguing that while the major parties do not use social media tools as much as the minor ones, i.e., relying
on the mainstream media attention being already directed towards them, the reproduction of content from
traditional media on social platforms works in their favor. This is commonly referred to as the “normalization
hypothesis” in literature (Klinger, 2013). Conversely, smaller parties – and especially fringe parties (Norris,
2003; Cardenal, 2013) – can compensate for their comparably limited access to traditional media outlets by
taking advantage of the online services to a higher degree, thereby reinforcing what is known as the
“equalization hypothesis” (also Klinger, 2013), which is the one we test here. Our hypothesis H2 is thus
composed by two underpinning ones:
• (H2.a) Minor and fringe parties post more often than the major ones.
• (H2.b) Minor and fringe parties draw higher levels of engagement from the users.
Complementing these two hypotheses, and considering the recent rise of right-wing populist parties
(Hameleers et al., 2017; Engesser et al., 2017, Stier et al., 2017), the third hypothesis (H3) can be
represented by the following:
• (H3.a) Right-leaning parties post more frequently compared to the left-leaning ones.
• (H3.b) Right-leaning parties receive more engagement from the users.
The rest of this paper is structured as follows. The next section provides the conceptual framework and
detailed information on the Portuguese political panorama, followed by a section on Facebook data
collection. We then articulate the methodology adopted in this work, and present the analysis results. Finally,
in the discussion section we outline our main findings.
32 OBS* Journal, 2022, 17(2)
Conceptual framework
Social networking platforms allow, at least by design, what some previous studies have described as a
decentralized communication of many-to-many (Janssen and Kies, 2005), although one can argue that the
very same design, built on the hierarchy post-comment-reply, can be as asymmetrical as one-to-many
communication (Lilleker et al., 2015; Theocharis et al., 2016). In fact, several studies across countries such
as Italy, France, Canada or the UK have observed that one-way communication is more prevalent
(Skovsgaard and Van Dalen, 2013; Kalsnes, 2016; Sobaci and Hatipoğlu, 2017).
Considering the recent studies on political communication, one of the most popular SNS is Facebook, which
provides one-to-many communication for political parties. We investigated how the communicative
exchange takes place on Facebook, in order to have a more accurate description of what happens both on
the
supply side
and on the
demand side
. The activity on the
supply side
can be measured by the number of
posts from the page owners, i.e. political parties, while from the
demand side
it can be measured by the
number of comments, shares, Likes and other reactions (Love, Haha, Wow, Sad or Angry) from the Facebook
users (Effing et al., 2011). These interactions, commonly explained and analyzed as user engagement
(Stromer-Galley, 2003; Stetka et al., 2019), are the major metrics used in literature.
In this paper, which is a part of a more comprehensive research project on the use of social media for
political communication in Portugal, our main goal is to apprehend how the official political parties with
current representation in the Portuguese parliament (i.e., following the Legislative Elections on 6 October
2019) use Facebook to communicate with the citizens. Particularly, we focus on the 2019 Legislative
Elections in Portugal and on the earlier elections for the Portuguese representatives in the European
Parliament (26 May 2019). Essentially, we want to understand how the Portuguese political parties used
Facebook to engage with the public in the imminence of the two nation-wide elections that took place in
2019 (namely before, during and after the electoral campaigns), and how the public responded.
The Portuguese political panorama
Among a plethora of other communication channels (e.g., news media, official bulletins, outdoors, etc.),
political actors nowadays also interact with citizens through social network platforms. These actors may be
institutional (governments, ministries and other governmental subdivisions, parties, their regional or local
chapters, etc.), semi-institutional (e.g., coalitions, official movements for more ephemeral events such as
referendums), or individuals (presidents, ministers, mayors, members of parliament or of other institutions,
candidates, etc.). All of these can be found in the Portuguese panorama, and they all compete for a place
in the political space and for the attention of citizens, with the leading role being on political parties instead
of individuals.
In order to justify some of the methodological options adopted in this work, in what follows we contextualize
the contemporary Portuguese political panorama until the legislature that started in late 2019, near the cut
date of our dataset. Since the 1974 Carnation Revolution (Palacios Cerezales, 2017) that overthrew a long-
standing dictatorship in Portugal, the number of official active parties has been between 15-20 in total, with
a clear majority of left-leaning ones. However, the dominant left orientation does not apply to the seats in
the parliament, as the electoral system aims to achieve proportional representation of votes. For the
Legislative Elections, this is done through a division of regions into districts or constituencies (called electoral
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 33
circles), which elect a number of members of the parliament (MPs) proportional to their relative population,
with the d’Hondt method being used to convert the votes of each circle into mandates (Leston Bandeira,
2009).
In the first elections of 1975, only seven individual parties earned enough votes to be represented, and until
recently the overall highest number of represented parties was nine. This occurred only once, in 1980, and
in the subsequent years fewer parties were in the parliament, partially because of the reduction in the
number of seats from 250 to 230 (in 1991). Since 1975, PS (Socialist Party) and PPD/PSD (Social-Democrat
Party) have been the major political contenders. These are followed by several medium-sized or minor
parties, including CDU (Unitary Democratic Coalition) – a coalition of PCP (Portuguese Communist Party)
and PEV (Ecologist Party) formed in 1987 –, CDS-PP (Popular Party), and finally BE (Left Block), which was
established in 1999 as a merge of several smaller left-wing parties.
This panorama of six parties in the national parliament lasted for five consecutive legislatures (1999, 2002,
2005, 2009, 2011), until the pattern was broken in 2015 with the arrival of the ecologist-animalist PAN
(People-Animals-Nature), with only one MP. Finally, in the 2019 Legislative Elections, the reversion of the
six-party trend became even more apparent with the establishment of other new parties: L (Free) on the
left, and on the right IL (Liberal Initiative) and C (Enough). Despite being founded in the 2010s, these three
parties managed to elect one MP each, while PAN enlarged its representation to four MPs. With this change
in 2019, the total number of individual parties with parliamentary representation became 10 (Table 1).
Table 1. Percentage of votes and the elected MPs count for the European and Legislative Elections from
2005 to 2019.
European Elections
Legislative Elections
2009
2014
2019
2005
2009
2011
2015
2019
Major
parties
Socialist Party (PS)
% of votes
26.5
31.5
35.9
45.0
36.6
28.1
32.3
38.2
MPs count
7
8
9
121
97
74
86
108
Social-Democrat Party
(PPD/PSD)
% of votes
31.7
27.72
23.6
28.8
29.1
38.7
36.93
29.2
MPs count
8
6
6
75
81
108
89
79
Minor
parties
Unitary Democratic
Coalition (CDU)1
% of votes
10.6
12.7
7.4
7.5
7.86
7.9
8.25
6.66
MPs count
2
3
2
14
15
16
17
12
Left Block (BE)
% of votes
10.7
4.6
10.6
6.3
9.81
5.17
10.2
10.0
MPs count
3
1
2
8
16
8
19
19
Popular Party (CDS-
PP)
% of votes
8.4
-2
6.7
7.3
10.4
11.7
-3
4.44
MPs count
2
1
1
12
21
24
18
5
Fringe
parties
Free (L)
% of votes
-
2.2
1.9
-
-
-
0.73
1.14
MPs count
-
0
0
-
-
-
0
1
People-Animals-Nature
(PAN)
% of votes
-
1.7
5.5
-
-
1.04
1.39
3.49
MPs count
-
0
1
-
-
0
1
4
1
CDU is a coalition between Portuguese Communist Party (PCP) and Ecologist Party (PEV). In this table, the number of
MPs of both parties is aggregated.
2
PPD/PSD and CDS-PP ran as a coalition (Aliança Portugal)
3
PPD/PSD and CDS-PP ran as a coalition (Portugal à Frente), except in Madeira and Azores (not included in %).
34 OBS* Journal, 2022, 17(2)
Liberal Initiative (IL)
% of votes
-
-
0.9
-
-
-
-
1.35
MPs count
-
-
0
-
-
-
-
1
Enough (C)
% of votes
-
-
1.64
-
-
-
-
1.35
MPs count
-
-
0
-
-
-
-
1
Total MPs
22
215
21
230
230
230
230
230
Source: Author’s elaboration with data from Comissão Nacional de Eleições [National Electoral Committee]
In Table 1 and Figure 1, the Portuguese political parties are categorized according to their foundation or
legal recognition dates, position in the political spectrum, and size – considering the results in the most
recent nationwide elections, from 2009 to 2019. Older parties are the ones established before 2000, the
year considered as the start of an information age on the Internet for the political institutions and actors
(Fuchs, 2017). Intuitively, newer parties are those established or legally recognized after 2000.
Figure 1. Mapping of parties according to the party’s foundation time, relative size, and position on the
political spectrum.
Source: Authors’ elaboration
Data collection
Regardless of being founded before or after the year 2000, all Portuguese official political parties have active
Facebook pages, which are the main source of information for our research. All of these Facebook pages
are public; hence, anyone including nonregistered users can view their content. For the purpose of this
exploratory study, posts from the Portuguese political parties were collected using Facepager, a tool for
4
C did not run for the elections, but the party’s president was the main candidate in a coalition between two small right-
wing parties (PPM and PPV/CDC).
5
The two remaining seats were occupied by MPT, a right-wing ecologist party.
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 35
fetching publicly available data from various websites (Jünger and Keyling, 2019).
As we investigate the activity of Portuguese political parties’ on Facebook around the periods of the two
major elections in 2019, all posts from that year were collected. The party’s name, Facebook page link,
account creation date, and the number of followers for each party’s page are listed in Table 2. It should be
pointed out that the information on the number of followers is not used in this study, since it reflects the
state at the time of the data extraction (May 2020), and the evolution of this metric across time is
unobtainable through the API. However, these numbers give us a general idea about the relative rankings
between parties, with PAN, PPD/PSD, and BE having the highest amount of followers at the time the data
were collected. On the other hand, CDU, CDS-PP, and C have the highest numbers of posts, and C, PS, and
IL have the highest numbers of comments during the considered period. Table 3 also shows the type of
political party, the total number of collected Facebook posts for each party and the performance of each
party page with comments, shares and reactions.
Table 2. The Portuguese political parties’ Facebook pages information from 2019.
Political Party
Facebook Page URL
Creation Date
Followers on
May 2020*
Portuguese Communist Party (PCP)
https://www.facebook.com/pcp.pt
2016-01-20
20K
Unitary Democratic Coalition (CDU)
https://www.facebook.com/CDUPCPPEV
2011-04-19
17K
Ecologist Party (PEV)
https://www.facebook.com/PartidoEcologistaOsVerdes
2011-02-10
12K
Left Block (BE)
https://www.facebook.com/esquerda.net
2011-11-13
102K
Free (L)
https://www.facebook.com/LIVREoficialpt
2013-11-16
28K
Socialist Party (PS)
https://www.facebook.com/SedeNacionalPartidoSocialista
2011-01-06
87K
People-Animals-Nature (PAN)
https://www.facebook.com/PANpartido
2015-05-17
162K
Social-Democrat Party (PPD/PSD)
https://www.facebook.com/ppdpsd
2010-05-17
151K
Liberal Initiative (IL)
https://www.facebook.com/iniciativaliberal
2018-08-22
75K
Popular Party (CDS-PP)
https://www.facebook.com/CDSPP
2014-12-11
36K
Enough (C)
https://www.facebook.com/PartidoChegaOficial
2018-10-23
87K
Source: Authors’ elaboration
Table 3. The Portuguese political parties’ type, Facebook page posts, comments, shares, and reactions
from 2019.
Party
PCP
CDU
PEV
BE
L
PS
PAN
PPD/
PSD
IL
CDS-
PP
C
Type
Old/
Min/
Left
Old/
Min/
Left
Old/
Min/
Left
Old/
Min/
Left
New/
Fri/
Left
Old/
Maj/
Left
New/
Fri/
Left
Old/
Maj/
Right
New/
Fri/
Right
Old/
Min/
Right
New/
Fri/
Right
Posts
590
757
591
591
608
636
582
650
599
662
664
Comments
3,382
1,519
292
4,991
10,315
26,189
17,883
15,777
19,473
5,555
34,073
Shares
60,199
35,886
4,214
18,724
28,100
25,851
47,958
21,725
84,626
18,201
108,935
Likes
71,323
55,277
11,086
31,693
71,468
132,444
179,940
98,912
355,986
58,900
340,464
Love
5,624
3,541
484
3,799
11,172
12,814
34,339
7,280
17,767
3,371
23,881
Haha
601
109
22
1,183
4,502
1,662
4,321
843
12,605
1,895
8,771
Wow
97
31
18
386
625
206
1,657
227
1,675
218
1,991
36 OBS* Journal, 2022, 17(2)
Sad
1,382
200
99
2,313
949
534
10,316
765
4,621
942
5,425
Angry
1,425
456
31
2,889
1,037
1,291
10,867
735
9,941
666
18,715
Total
144K
98K
16K
65K
128K
201K
307K
146K
506K
89K
542K
Source: Authors’ elaboration
As one can observe from the daily number of posts in the collected data over time shown in Figure 2, two
peaks in the posting activities during 2019 occurred for the periods of two major elections in Portugal (i.e.
European and Legislative Elections). Figure 3 compares the total post counts between different parties for
the considered 2.5-month periods around the two elections, i.e., before, during, and after their electoral
campaigns.
Figure 2. Number of posts per day for all party pages during 2019.
Source: Authors’ elaboration
Figure 3. Total number of posts in the month before, during, and after the campaign of the European and
Legislative Elections over party pages (2.5 months in total).
Source: Authors’ elaboration
Data analysis
The previous analysis of the aggregate data set with posts by all parties during 2019 reveals the overall
behavior, the peaks in posting activities during the two major election periods, and the differences in the
total number of posts for these elections. For a more detailed information on the posting dynamics and
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 37
social media engagement over time, in this section we elaborate the analysis performed to test the three
hypotheses described in the Introduction, which involve comparison between the following pairs of
complementary groups:
(1) Older vs. Newer Parties;
(2) Major vs. Minor Parties (including fringe parties);
(3) Right vs. Left Parties.
For each group, posts and corresponding interactions were taken into account, with the latter considering
reactions, shares and comments separately. Weekly averages were used instead of daily counts, in order to
smooth out sporadic spikes. To be able to compare the aforementioned party groups, we calculated the
differences between opposing groups in all these weekly averages. More precisely, the differences were
calculated by subtracting the number obtained for the second group (in the listed order) from the
corresponding number obtained for the first one. For example, for the first complementary pair, the weekly
average number of posts by newer parties is subtracted from the weekly average number of posts by the
older parties.
Our study takes two perspectives; the first one, being a typical approach in literature, focuses on the two
major elections. While the official campaigns take place only during the two weeks preceding the event
(excluding the election day itself and the “reflection day” on the eve), we additionally consider a month
before the campaign, usually but unofficially called “pre-campaign”, and a month after the election, to verify
if any changes in the behavior occur after the results are announced. The second one considers the
aggregate data from the whole span of the year 2019. All of the null-hypotheses in this work are set up to
test if the pages have equal activity levels on average (i.e. by considering the number of posts per week),
and the weekly average levels of engagement per post (i.e. by jointly considering the number of reactions,
comments, and shares). We apply ANOVA tests (Ott and Longnecker, 2015) to the data sets for the two
groups in each hypothesis, in order to observe the differences between these groups’ means, with the
standard choice of 5% test significance being adopted.
Supply side
We assume that the most reliable indicator of a political party’s social media activity is the frequency of
posts on its Facebook page. Our findings from the joint consideration of the data for the whole span of the
year 2019 suggest that older parties are more active than newer ones (contradicting H1.a.), minor and
fringe parties are more active than the major ones (confirming H2.a.), and the left parties are more active
than the right-leaning ones (contradicting H3.a.). This can be observed from Figure 4, showing the difference
in the weekly average number of posts between the two groups, for each of the three considered pairs.
Moreover, by focusing on the two major elections periods, the findings show that there is a statistically
significant difference between the older and newer parties post activities, except in the month after the
Legislative Elections. This can be seen from the summary of hypothesis testing results, summarized in Table
4, with the scientific E-notation being used in p-values because of their generally low values.
38 OBS* Journal, 2022, 17(2)
Figure 4. Differences in weekly average numbers of posts for the considered groups of parties.
Source: Authors’ elaboration
Table 4. P-Values for the Null Hypotheses on the party pages’ activities overall span of 2019, and for the
periods of the European and Legislative Elections; a month before the election campaigns, during the
election campaigns (2-weeks), and a month after the election campaigns. (R: Reject, NR: Not Reject)
Overall
span of
2019
European Elections
Legislative Elections
Hypotheses
Variables
A month
before the
election
campaign
During the election
campaign
A month after
the election
campaign
A month before
the election
campaign
During the
election
campaign
A month after
the election
campaign
H1a
Posts
count
8.97e-19
(R)
7.50e-11 (R)
4.63e-03 (R)
1.76e-04 (R)
2.00e-04 (R)
9.81e-03 (R)
6.42e-01
(NR)
H2a
Posts
count
4.78e-50
(R)
2.26e-16 (R)
5.16e-03 (R)
2.14e-08 (R)
3.78e-13 (R)
5.78e-03 (R)
4.91e-07 (R)
H3a
Posts
count
7.39e-15
(R)
1.09e-04 (R)
4.67e-02 (R)
1.93e-04 (R)
3.48e-06 (R)
3.99e-06 (R)
1.09e-03 (R)
Source: Authors’ elaboration
Unsurprisingly, the weekly average number of posts during the month before and during the election
campaigns is higher than after the elections, for both the European and Legislative Elections (Figure 5). The
most notable outlier is CDU’s activity during the election campaigns for both elections, due to its excessive
effort to show their place in the voting ballots. From what we could find out looking at the post activity of
this party over a more extended time, this behavior is exclusive to the campaign periods. Similarly, C’s high
level of activity is evident for the Legislative Elections (Figure 5b). As already shown above (Figure 3), the
total number of posts was higher during the Legislative Elections than in the European Elections, though not
for PPD/PSD, IL, and CSD-PP (right-leaning parties), and also barely higher in the European Elections for
left-leaning BE and PS. Except for the older right-wing party CDS-PP during the Legislative Elections, a
consistently higher activity is observed during the campaign weeks, followed by reduced activity during the
post-election month, with the average numbers of posts falling below the values seen during the month
preceding the campaigns.
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 39
Figure 5. Weekly average number of posts in the month before, during, and after the campaign of the (a)
European and (b) Legislative Elections over party pages.
Source: Authors’ elaboration
Demand side
While the previous subsection focuses on the activities of parties in terms of posting content on their
Facebook pages, this one considers the interaction with the posted content in terms of reactions (a general
term that also includes Likes
6
), comments, and shares from users. Due to the existence of several different
types of user engagement, this analysis requires an additional degree of complexity compared to the
supply
side
.
To measure user engagement on Facebook, we used the number of comments, shares, likes, and reactions
(given that Facebook introduced additional reactions to the Like button in February 2016 – Love, Haha,
Wow, Sad, and Angry). “The like button was originally introduced to replace short affective statements like
Awesome and Congrats!” (Gerlitz and Helmond, 2013). According to these authors, the intuitive assumption
behind the Like button is that the number of likes implies exposure, attention, and some sort of affirmation,
ratification, or endorsement of what is posted. While ambiguity can never be fully eradicated, given that any
reaction may be used in an ironic fashion (Gerlitz and Helmond, 2013), these new reactions present
nevertheless a more straightforward and nuanced division, particularly between those associated with
positive sentiments (Like and Love) and the remaining ones (Haha, Wow, Sad and Angry
7
). We thus
conceived two additional metrics; namely,
positive reactions
(sum of Likes and Loves), and
other reactions
(sum of Hahas, Wows, Sads, and Angrys).
Although reactions represent the most common form of engagement, especially positive ones such as Likes,
which are always above 80% of all reactions (Figure 6), the other types of engagement are still essential
for the understanding of users’ interactions with the party pages. Therefore, we modified Bonson and
6
This is acknowledged not only by Facebooks API, which in the field “reactions.summary.total count” sums all reactions including “like.summary.total
count” but also in Facebook’s FAQs, e.g. in https://en.facebookbrand.com/facebookapp/assets/reactions/: “The collection of Reactions includes Like,
Love, Care [unavailable before March 2020], Haha, Wow, Sad and Angry.”
7
Eberl et al. (2020) have a different approach, classifying Haha and Wow also as positive reactions, and Sad and Angry as negative ones. We find
this distinction problematic, especially in the cases of Haha (which has a very high potential of sarcasm – laughing
at
the post, dismissing it as
ridiculous, rather than
with
the post, considering it amusing) and Sad (which may have been used to display compassion before the introduction of
Care in March 2020). For this reason, only Like and Love were counted as
positive
(Sandoval-Almazan and Valle-Cruz, 2020; Gerodimos and
Justinussen, 2015), while the remaining ones – which may be analyzed more closely in future studies – were grouped as
other
reactions. We did
however check all of them separately to look for relevant information.
40 OBS* Journal, 2022, 17(2)
Ratkais’ metrics by considering the weekly averages of
positive
reactions,
other
reactions, comments, and
shares.
Figure 6. Percentage of weekly average of a specific reaction over total reactions.
Source: Authors’ elaboration
Considering the reactions of the complementary groups of parties, the graphics show that newer parties
(H1.b) tend to receive more reactions overall than older ones (Figure 7). However, considering either the
party size (H2.b) or its position in the spectrum (H3.b), both
positive
and
other
reactions render mixed
results. For the pair major-minor parties the difference in the average number of all reactions is not
significant after election campaigns, for both elections. Moreover, no significant difference between major
and minor parties can be observed during the European Elections campaign. However, before each
campaign, major parties generally receive more
positive
reactions than the minor ones (Figure 7). For the
other
reactions (Figure 9), right-wing parties drew more of this type of reactions than the left-leaning ones,
with significant statistical results during the overall span of 2019. These results also show that the right-
leaning parties receive more reactions of both kinds before, during and after the Legislative Elections,
whereas there is no significant difference between left and right-leaning parties during the European
Elections, except for the
positive
reactions during the month before the election campaign.
Figure 7. Differences in weekly average numbers of
positive
reactions between the considered groups of
parties.
Source: Authors’ elaboration
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 41
The results for the hypotheses tests for the
demand side
are summarized in Table 5, again with the scientific
E-notation being used in p-values because of their generally low values.
The most interesting behavior around the European Elections is observed for PAN (newer, fringe, and left-
leaning) and IL (newer, fringe, and right-leaning), which received considerably higher numbers of
positive
reactions than the other parties. On the other hand, IL and C (both newer, fringe, and right-leaning) received
the highest positive reactions after the Legislative Elections (Figure 8).
Table 5. P-Values for the Hypotheses on the party pages’ user engagement overall span of 2019, and for
the periods of the European and Legislative Elections; a month before the election campaigns, during the
election campaigns (2-weeks), and a month after the election campaigns. (R: Reject, NR: Not Reject).
Overall span of
2019
European
Elections
Legislative
Elections
Hypotheses
Variables
Before
During
After
Before
During
After
H1b
Reactions
9.59e-174
(R)
1.73e-11
(R)
2.22e-06
(R)
6.22e-15
(R)
8.78e-23
(R)
6.58e-11
(R)
5.94e-20
(R)
Positive
reactions
7.94e-163
(R)
3.77e-11
(R)
2.40e-06
(R)
2.15e-14
(R)
5.18e-22
(R)
2.27e-10
(R)
8.42e-20
(R)
Other reactions
5.09e-67
(R)
6.48e-05
(R)
2.70e-03
(R)
6.98e-05
(R)
2.90e-10
(R)
2.51e-04
(R)
4.56e-06
(R)
Comments
1.05e-37
(R)
4.59e-01
(NR)
7.62e-02
(NR)
6.82e-03
(R)
3.71e-01
(NR)
2.67e-03
(R)
1.17e-08
(R)
Shares
1.18e-42
(R)
1.01e-01
(NR)
4.45e-01
(NR)
2.70e-05
(R)
1.88e-04
(R)
5.05e-03
(R)
2.48e-03
(R)
All engagement
7.05e-140 (R)
1.24e-06 (R)
2.03e-05 (R)
3.39e-13 (R)
1.19e-14 (R)
1.18e-08 (R)
1.33e-14 (R)
H2b
Reactions
2.87e-03
(R)
3.15e-03
(R)
7.10e-01
(NR)
3.67e-01
(NR)
8.04e-08
(R)
1.21e-04
(R)
7.01e-01
(NR)
Positive
reactions
3.01e-02
(R)
2.39e-03
(R)
7.07e-01
(NR)
3.62e-01
(NR)
1.09e-09
(R)
3.92e-05
(R)
8.52e-01
(NR)
Other reactions
2.02e-09
(R)
7.12e-01
(NR)
9.97e-01
(NR)
8.49e-01
(NR)
3.01e-01
(NR)
4.98e-01
(NR)
2.97e-01
(NR)
Comments
3.03e-15
(R)
3.61e-09
(R)
6.67e-03
(R)
6.48e-03
(R)
7.75e-38
(R)
7.11e-08
(R)
4.01e-01
(NR)
Shares
1.48e-08
(R)
5.59e-01
(NR)
1.22e-01
(NR)
2.37e-01
(NR)
7.06e-01
(NR)
8.37e-01
(NR)
6.13e-01
(NR)
All engagement
2.33e-03
(R)
1.19e-03
(R)
7.37e-01
(NR)
4.75e-01
(NR)
9.92e-07
(R)
1.22e-03
(R)
7.39e-01
(NR)
H3b
Reactions
3.64e-73
(R)
3.24e-03
(R)
9.17e-01
(NR)
8.62e-01
(NR)
4.84e-11
(R)
3.98e-06
(R)
7.86e-26
(R)
Positive
reactions
1.71e-73
(R)
2.64e-03
(R)
9.41e-01
(NR)
8.33e-01
(NR)
6.55e-12
(R)
1.17e-05
(R)
1.83e-26
(R)
Other
reactions
3.94e-15
(R)
9.89e-01
(NR)
1.12e-01
(NR)
7.43e-01
(NR)
4.30e-02
(R)
7.75e-04
(R)
5.25e-06
(R)
Comments
1.10e-19
(R)
7.94e-01
(NR)
6.58e-01
(NR)
7.22e-01
(NR)
1.14e-02
(R)
4.08e-03
(R)
4.46e-10
(R)
Shares
8.90e-16
(R)
1.62e-01
(NR)
1.53e-02
(R)
2.38e-01
(NR)
8.20e-02
(NR)
2.23e-02
(R)
1.26e-05
(R)
All
engagement
4.07e-58
(R)
8.36e-02
(NR)
6.94e-01
(NR)
7.02e-01
(NR)
3.48e-07
(R)
2.16e-05
(R)
1.01e-19
(R)
Source: Authors’ elaboration
42 OBS* Journal, 2022, 17(2)
Figure 8. Weekly average number of
positive
reactions in the month before, during, and after the
campaign of the (a) European and (b) Legislative Elections over party pages.
Source: Authors’ elaboration
For
other
reactions, the most identifiable difference is observed for the party group pair based on the
foundation time, for both electoral periods (Figure 9). Also, we observe significant differences in this type
of engagement for the party group pair based on the political position, but only during the Legislative
Elections periods. In other words, the right-leaning parties obtained more
other
reactions than the left-
leaning ones (Figure 9); with IL and PAN having the highest number of these reactions after the European
Elections, and C and IL assuming the top positions for the Legislative ones (Figure 10), as already observed
with
positive
reactions.
Figure 9. Differences in weekly average numbers of
other
reactions between the considered groups of
parties.
Source: Authors’ elaboration
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 43
Figure 10. Weekly average number of
other
reactions in the month before, during, and after the campaign
of the (a) European and (b) Legislative Elections over party pages.
Source: Authors’ elaboration
Regarding the average number of
comments
per post for each party, for the overall span of the year 2019,
the newer and right-leaning parties (H1.b. and H3.b) received more comments than their older and left-
leaning counterparts (Figure 11), respectively. However, for the month before and during the two-week
European Elections campaign, there is no significant difference between the older and newer parties, as
between the pair Left-Right, additionally non-significant before that campaign. The pair older-newer parties
also shows comparable levels of comments during the month before the Legislative Elections, while the
same can be observed for the pair major-minor parties during the month after.
Figure 11. Differences in weekly average numbers of comments between the considered groups of parties.
Source: Authors’ elaboration
PS received the most comments among all parties, before both election campaigns as well as during the
Legislative Elections campaign (Figure 12). PAN got the highest number of comments during the European
Elections campaign, while IL and PAN had the most comments in the month after. The most interesting
behavior around the Legislative Elections is found for C, the far right party, whose Facebook page was the
most commented one in the period after the Legislative Elections, followed by IL and PPD/PSD (Figure 12b).
44 OBS* Journal, 2022, 17(2)
Figure 12. Weekly average number of comments in the month before, during, and after the campaign of
the (a) European and (b) Legislative Elections over party pages.
Source: Authors’ elaboration
Finally, we consider the number of shares of parties’ posts on their Facebook pages; the differences in the
weekly average number of shares between the party groups is shown in Figure 13. The pair major-minor
parties shows no significant differences during both elections, while the differences between right and left-
leaning parties are significant during the European Elections campaign, and also during and after the
Legislative Elections. A similar difference in shares between older and newer parties is observed in the period
after the European Elections campaign and all periods of Legislative Elections.
Figure 13. Differences in weekly average numbers of shares between the considered groups of parties.
Source: Authors’ elaboration
For the specific periods around the elections considered in this work, IL’s posts were the most shared ones
before and after both campaigns, whereas PAN received the most shares during the European Elections
campaigns (Figure 14a). IL and C also show visible differences in the number of shares in the month after
the Legislative Elections (Figure 14b).
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 45
Figure 14. Weekly average number of shares in the month before, during, and after the campaign of the
(a) European and (b) Legislative Elections over party pages.
Source: Authors’ elaboration
By summing up all types of engagement and considering the weekly average of all reactions together, the
graphics show that the newer parties posts inspire more engagement than the older ones (H1.b), and that
right-leaning parties draw more engagement than the left-leaning ones (H3.b), especially around the
Legislative Elections period (Figure 15). During the European Elections campaign and the month after, only
the pair older-newer parties shows significant differences. For the Legislative Elections period, more precisely
in the month after the campaign (Figure 16), major and minor parties show similar overall user engagement,
contrary to the assumed hypothesis (H2.b).
Figure 15. Differences in weekly average numbers of all interactions (any type of engagement) for the
considered groups of parties.
Source: Authors’ elaboration
When it comes to the overall interaction with the parties’ posts, Figure 16 shows again that IL and PAN
achieved the highest levels of engagement during the European Elections campaign and the month after.
IL’s posts are also seen as the most engaging ones in the same periods around the Legislative Elections,
with C attracting similar levels in the month after the elections. Also, for the overall user engagement, the
Legislative Elections period received more attention compared to the European Elections period.
46 OBS* Journal, 2022, 17(2)
Figure 16. Weekly average number of all interactions (any type of engagement) in the month before, during,
and after the campaign of the (a) European and (b) Legislative Elections over party pages.
Source: Authors’ elaboration
Discussion and conclusion
Our research findings suggest that the opposition between major and minor parties (being the fringe ones
included in the latter), covered by hypotheses H2.a and H2.b, is the one that in general yields less statistically
significant results, thus leaving the “normalization vs. equalization” debate undecided. For the overall span
of 2019, although the level of posting activities (
supply side
) of older and left-leaning parties (H1.a and
H3.a) is slightly higher than that observed for the newer and right-leaning ones, the user engagement
(
demand side
) is greater for the newer and right-leaning parties, respectively (H1.b and H3.b).
The common
supply side
hypothesis related to the party’s foundation time (H1.a.) (Lilleker et al., 2015;
Magin et al., 2017) fails when tested against the data set collected for this study. Moreover, the hypothesis
related to the party’s political position (H3.a.) cannot be accepted when looking at the variations in the
frequency of posts and different periods around the two considered elections. In both cases the results are
significant, but contradicting the assumptions of the existing literature.
However, even in cases in which the p-values in the ANOVA tests were consistent with the hypotheses,
these may be either attributed to short-term or even more ephemeral factors, or to emerging long-term
trends that only future research can acknowledge. This is particularly noticeable in the following illustrations
from the
demand side
. In spite of the three newer and fringe parties (IL, PAN, and C) as the most active
and engaging content owners, PS achieves a comparable degree of
comments
from their followers,
especially before the official campaign periods. As such, none of the hypotheses (H1.b., H2.b., and H3.b)
can be consistently validated, even if there are occasionally statistically significant differences regarding this
engagement metric. This particular exception to the reports in the literature, i.e. for a major, older, and
center-left party to have a good performance regarding
comments
, could be attributed to the fact that this
party was at the time in the government and won both elections, and as such attracted more public attention.
Only a different methodology of analysis may allow an answer for this increase in attention. A similar case
can be argued by considering the number of
shares
. Four parties are observed with a higher average number
of shares, with PCP appearing alongside the usual suspects IL, PAN, and C. Being a left-wing and most of
all an older party, it is somewhat surprising to find it as a matching contender in shares with these newer
ones.
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 47
By considering the rise of right-wing populism reported by some authors (Hameleers et al., 2017; Engesser
et al., 2017; Stier et al., 2017), our results show that there is a significant difference between the right and
left leaning parties’ user engagement (H3.b) in the periods around the Legislative Elections – this difference,
with an advantage for parties on the Right, can be clearly seen in Figure 15. The party that allegedly adopts
a more distinctive populist style, C, does not show a higher degree of activity, although a considerable
growth in its activity nearing the Legislative Elections and even after (as the party earned a seat in the
parliament), can be observed, along with occasional spikes that call for further research, e.g. based on
content analysis, for a more assertive explanation. For another newer right-wing party, IL, to which one
could arguably attach the epithet of populism, the evidence seems to outright invalidate the
supply-side
of
the hypothesis: its Facebook posting activity is similar to that observed from the other parties.
Along with these findings, some structural constraints and limitations derived from contextual methodologies
must also be acknowledged. First of all, the automated data retrieval tools, while providing an extremely
convenient way of obtaining large amounts of information in standardized and interoperable formats, only
allow collection of data authorized by the APIs, dependent on laws such as those relating to the protection
of private data, e.g. General Data Protection Regulation. This means that any study of this kind is practically
unable to obtain other information such as the followers’ demographics, or which was the number of
followers at the time of the posts. Other hypotheses extending those presented in this study, for example
whether the user engagement comes mostly from younger followers, or if there is some kind of generational
gap between the followers of older and newer parties, are thereby beyond the reach of this method. Another
structural limitation arises from the fact that, in spite of the general knowledge that Facebook algorithms
introduce biases regarding who gets to see which posts in their feed, the collected data does not provide a
way to know how many people have actually seen the post, if the post appeared only in the feed of the
followers of the corresponding page, if it had a wider audience because it was sponsored, etc. The political
parties may be able to access some of that information via the page analytics provided by Facebook and
adapt their campaigns accordingly, but that is not available for researchers.
Other limitations are specific to this study, and may be addressed in further research. This study considers
only the number of posts and engagement metrics, without analyzing the content or media in these posts
or comments. However, content-aware analysis of the interactions between parties and Facebook users
would be beneficial for further analyses, namely the post types (text, photo, video, and link), as each of
those may convey, or at least potentiate, specific communication styles adopted by parties and enable
different levels or types of user engagement. As an illustration, we have observed that photos and videos
constitute the bulk of the collected posts; however, the analysis on the parties’ usage of different media and
the user engagement levels associated with different types of posted content is beyond the scope of this
work.
A few important corollaries emerged from the obtained results. One was the finding that significant
differences between considered party groups occurred mostly on the
demand side
and during the electoral
periods, and more assertively in Legislative than in the European elections. Even though parties show
generally similar behaviors in the posting frequency (e.g., an increase in the activity during the campaigns,
followed by a sharp decrease afterwards), significant differences were observed in the user engagement for
the newer and right-leaning parties as they received more attention from their followers, especially after the
Legislative Elections period.
48 OBS* Journal, 2022, 17(2)
Some of the findings were intriguing enough to inspire or at least reshape further research plans. That is
particularly pertinent for a more in-depth look at the interactions between the post activities and user
engagement by the means of emotion and content analysis. Albeit less common than the
positive
reactions
(Like and Love), the
other
reactions may constitute a suitable and original research path, especially
considering the fact that these seem to be more frequent with the content posted by newer parties, however
in a seemingly conflicting way. Are the newer (and also fringe) parties such as PAN or IL more prone to
antithetical forms of engagement, as they are simultaneously attracting reactions of both laughter and
anger? Can the comments’ semantics reveal some kind of battleground between supporters and detractors?
The same questions can be asked about the posts that attract more comments or those that generate more
shares. Do some political themes and issues entice the audience more than others, and which ones? Do
they reveal something about the rise of right-wing populism, fake news and hate speech on social network
platforms? For future research, we will aim at detecting humor, irony, and provocation in posts and
investigating contradictory reactions, such as Love, Sad, and Angry.
Acknowledgements/Funding information
This work was supported by a grant from FCT - Fundação para a Ciência e a Tecnologia (Portugal) for the
research project titled “Political Interest Networks in Facebook Portugal” (PDTC/COM-CSS/28269/2017).
Declaration of interest statement
The authors have no conflict of interest to declare.
Bibliographical references
Bennett, W. L. (2012). The Personalization of Politics: Political Identity, Social Media, and Changing Patterns
of Participation.
The Annals of the American Academy of Political and Social Science
, 644(1):20–39.
doi: 10.1177/0002716212451428.
Boulianne, S. and Larsson, A. O. (2021). Engagement with Candidate Posts on Twitter, Instagram, and
Facebook during the 2019 Election.
New Media & Society
, 25(1):119–140. doi:
10.1177/14614448211009504.
Cardenal, A. S. (2013). Why Mobilize Support Online? The Paradox of Party Behavior Online.
Party Politics
,
19(1):83–103. doi: 10.1177/1354068810395059.
Comissão Nacional de Eleições (n.d.). https://www.cne.pt/.
Eberl, J.-M., Tolochko, P., Jost, P., Heidenreich, T., and Boomgaarden, H. G. (2020). What’s in a Post? How
Sentiment and Issue Salience Affect Users Emotional Reactions on Facebook.
Journal of Information
Technology & Politics
, 17(1):48–65. doi: 10.1080/19331681.2019.1710318.
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 49
Effing, R., Jos, V. H., and Huibers, T. (2011). Social Media and Political Participation: Are Facebook, Twitter
and YouTube Democratizing our Political Systems? In
International Conference on Electronic
Participation
, 25–35, Springer. doi: 10.1007/978-3-642-23333-3_3.
Engesser, S., Fawzi, N., and Larsson, A. O. (2017). Populist Online Communication: Introduction to the
Special Issue.
Information, Communication & Society
, 20(9):1279–1292. doi:
10.1080/1369118X.2017.1328525.
Enli, G. S. and Skogerbø, E. (2013). Personalized Campaigns in Party-Centred Politics: Twitter and Facebook
as Arenas for Political Communication.
Information, Communication & Society
, 16(5):757–774. doi:
10.1080/1369118X.2013.782330
Fuchs, C. (2017).
Social Media: A Critical Introduction
, Sage. doi: 10.1080/1369118X.2013.782330.
Gerlitz, C. and Helmond, A. (2013). The Like Economy: Social Buttons and the Data Intensive Web.
New
Media & Society
, 15(8):1348–1365. doi: 10.1177/1461444812472322.
Gerodimos, R. and Justinussen, J. (2015). Obama’s 2012 Facebook Campaign: Political Communication in
the Age of the Like Button.
Journal of Information Technology & Politics
, 12(2):113–132. doi:
10.1080/19331681.2014.982266
Hameleers, M., Bos, L., and de Vreese, C. H. (2017). The Appeal of Media Populism: The Media Preferences
of Citizens with Populist Attitudes.
Mass Communication and Society
, 20(4):481–504. doi:
10.1080/15205436.2017.1291817.
Heiss, R., Schmuck, D., and Matthes, J. (2019). What Drives Interaction in Political Actors Facebook Posts?
Profile and Content Predictors of User Engagement and Political Actors Reactions.
Information,
Communication & Society
, 22(10):1497–1513. doi: 10.1080/1369118X.2018.1445273.
Janssen, D. and Kies, R. (2005). Online Forums and Deliberative Democracy.
Acta Politica
, 40(3):317–335.
doi: 10.1057/palgrave.ap.5500115.
Jensen, J. L. and Schwartz, S. A. (2022). Introduction: A Decade of Social Media Elections.
Social
Media+Society
, 8(1):1–4. doi: 10.1177/20563051211063461.
Jünger, J. and Keyling, T. (2019). Facepager. An Application for Automated Data Retrieval on the Web.
Source code and releases available at https://github.com/strohne/Facepager/
Kalsnes, B. (2016). The Social Media Paradox Explained: Comparing Political Parties Facebook Strategy
versus Practice.
Social Media+Society
, 2(2):1–11. doi: 10.1177/2056305116644616.
Kapoor, K. K., Tamilmani, K., Rana, N. P., Patil, P., Dwivedi, Y. K., and Nerur, S. (2018). Advances in Social
Media Research: Past, Present and Future
. Information Systems Frontiers
, 20(3):531–558. doi:
10.1007/s10796-017-9810-y.
Klinger, U. (2013). Mastering the Art of Social Media: Swiss Parties, The 2011 National Election and Digital
Challenges.
Information, Communication & Society
, 16(5):717–736. doi:
10.1080/1369118X.2013.782329.
Larsson, A. O. (2016). Online, All the Time? A Quantitative Assessment of the Permanent Campaign on
Facebook.
New Media & Society
, 18(2):274–292. doi: 10.1177/1461444814538798.
Larsson, A. O. and Kalsnes, B. (2014). ‘Of course we are on Facebook’: Use and non-use of social media
among Swedish and Norwegian politicians.
European Journal of Communication
, 29(6):653–667. doi:
10.1177/0267323114531383.
50 OBS* Journal, 2022, 17(2)
Leston-Bandeira, C. (2009). Dissent in a Party-Based Parliament: The Portuguese Case.
Party Politics
,
15(26):695–713. doi: 10.1177/1354068809334556.
Lilleker, D. G., Tenscher, J., and Štětka, V. (2015). Towards Hypermedia Campaigning? Perceptions of New
Media’s Importance for Campaigning by Party Strategists in Comparative Perspective.
Information,
Communication & Society
, 18(7):747–765. doi: 10.1080/1369118X.2014.993679.
Macnamara, J. (2011). Pre and Post-Election 2010 Online: What Happened to the Conversation?
Communication, Politics & Culture
, 44(2): 18–36.
https://search.informit.org/doi/10.3316/informit.627292905802447.
Magin, M., Podschuweit, N., Haßler, J., and Russmann, U. (2017). Campaigning in the Fourth Age of Political
Communication. A Multi-Method Study on the Use of Facebook by German and Austrian Parties in the
2013 National Election Campaigns. Information,
Communication & Society
, 20(11):1698–1719. doi:
10.1080/1369118X.2016.1254269.
Norris, P. (2003). Preaching to the converted? Pluralism, Participation and Party Websites.
Party Politics
,
9(1):21–45. doi: 10.1177/135406880391003.
Ott, R. L. and Longnecker, M. T. (2015).
An Introduction to Statistical Methods and Data Analysis
. Nelson
Education.
Palacios Cerezales, D. (2017). Civil Resistance and Democracy in the Portuguese Revolution.
Journal of
Contemporary History
, 52(3):688–709. doi: 10.1177/0022009416641496.
Sandoval-Almazan, R. and Valle-Cruz, D. (2020). Sentiment Analysis of Facebook Users Reacting to Political
Campaign Posts.
Digital Government: Research and Practice
, 1(2):1–13. doi: 10.1145/3382735.
Serra-Silva, S., Dias Carvalho, D., and Fazendeiro, J. (2018). Party-Citizen Online Challenges: Portuguese
Parties’ Facebook Usage and Audience Engagement. In
Changing Societies: Legacies and Challenges.
Vol. 2. Citizenship in Crisis
, 185–214. doi: 10.31447/ics9789726715047.08.
Skovsgaard, M. and Van Dalen, A. (2013). Dodging the Gatekeepers? Social Media in the Campaign Mix
during the 2011 Danish Elections.
Information, Communication & Society
, 16(5):737–756. doi:
10.1080/1369118X.2013.783876.
Sobaci, M. Z. and Hatipoğlu, ˙I. (2017). Measuring the Engagement Level of Political Parties with Public on
Facebook: The Case of Turkey. In
2017 Conference for EDemocracy and Open Government (CeDEM)
,
209–216, IEEE. doi: 10.1109/CeDEM.2017.15.
Stetka, V., Surowiec, P., and Mazák, J. (2019). Facebook as an Instrument of Election Campaigning and
Voters’ Engagement: Comparing Czechia and Poland.
European Journal of Communication
, 34(2):121–
141. doi: 10.1177/0267323118810884.
Stieglitz, S., Brockmann, T., and Dang-Xuan, L. (2012). Usage of Social Media for Political Communication.
In
Proceedings: Pacific Asia Conference on Information Systems
, 22.
https://aisel.aisnet.org/pacis2012/22.
Stieglitz, S. and Dang-Xuan, L. (2013). Emotions and Information Diffusion in Social Media: Sentiment of
Microblogs and Sharing Behavior.
Journal of Management Information Systems
, 29(4):217–248. doi:
10.2753/MIS0742-1222290408.
Stier, S., Posch, L., Bleier, A., and Strohmaier, M. (2017). When Populists Become Popular: Comparing
Facebook Use by the Right-Wing Movement Pegida and German Political Parties.
Information,
Communication & Society
, 20(9):1365–1388. doi: 10.1080/1369118X.2017.1328519.
OBS* Journal, 2023, 17(2) G.Bacaksizlar Turbic, J. Martins Rosa 51
Stromer-Galley, J. (2003). Diversity of Political Conversation on the Internet: Users’ Perspectives.
Journal of
Computer-Mediated Communication
, 8(3):JCMC836. doi: 10.1111/j.1083-6101.2003.tb00215.x.
Theocharis, Y., Barberá, P., Fazekas, Z., Popa, S. A., and Parnet, O. (2016). A Bad Workman Blames his
Tweets: The Consequences of Citizens’ Uncivil Twitter Use when Interacting with Party Candidates.
Journal of Communication
, 66(6):1007–1031. doi: 10.1111/jcom.12259.
Vergeer, M. and Hermans, L. (2013). Campaigning on Twitter: Microblogging and Online Social Networking
as Campaign Tools in the 2010 General Elections in the Netherlands.
Journal of Computer-Mediated
Communication
, 18(4):399–419. doi: 10.1111/jcc4.12023.
