Sage Journals: Discover world-class research

Abstract

Scholars connect the rise of populist parties and the growing importance of social media platforms for political communication as the emerging social media logic seems well-suited to populist communication. The following research note presents a novel dataset that connects Facebook accounts to party-level populism data from the POPPA database. Over 600,000 posts from 226 parties between 2017 and 2019 were analysed. The results show that in most European countries, populist parties are more active than non-populist parties on Facebook. However, high variation in the frequency of posts emphasises that country-specific aspects play an important role. Generally, Italian parties are much more active than those from other countries, whereas those in Northern and Western Europe are the least active. The most active party in Europe, the Italian right-wing populist Lega, showed the most extreme level of activity: its daily activity is around 20 times higher than the median. Furthermore, this analysis highlights why researchers should carefully check Facebook data for implausible inactivity and how connecting different data resources can help overcome potential biases resulting from missing data. Future studies analysing any party communication on Facebook will benefit from the insights and the list of party accounts featured herein.

Keywords

Populism social media Europe political parties

Introduction

Many observers connect the rise of populist parties to the increasing relevance of social media in political communication (De Vreese et al., 2018; Kriesi, 2014: 366–367), as data-driven social media campaigns play a crucial role (Roemmele and Gibson, 2020). This development favours populists, as the emerging social media logic seems well-suited to populist communication (Engesser et al., 2017). Populists use social media platforms to bypass the established media, which they feel mistreats them (Krämer, 2017: 1303). This is in congruence with populist voters, who prefer news from social media over that produced by mass media outlets (Pew Research Center, 2018). Among the various platforms, many assume that Facebook, the most popular platform with almost three Billion monthly active users (DataReportal, 2021), offers ideal opportunity structures for populists. Many young users recently left the platform (Pew Research Center, 2022), presumably in protest against the platform algorithm’s fostering of polarisation (Barrett et al., 2021). However, few comparative empirical studies have analysed whether populist parties are actually more active on Facebook. The following research note offers insight into the post frequency of 226 European parties by covering over 600,000 posts between 2017 and 2019, connected with party-level datasets. It also problematises Facebook data quality and offers possible solutions. The list of party accounts published along with this research note (Thomeczek, 2023) facilitates future studies on Facebook communication.

Social media communication: an effective tool for populist parties

Research suggests that the shift from mass media to social media communication has created a favourable communication environment for populists. Comparing the inherent logic of social and mass media exemplifies this (Klinger and Svensson, 2015). Historically, populists have propagated direct communication with “the people”, which fits well with social media’s logic of distribution (Engesser et al., 2017: 1113). Furthermore, given the logic of usage, social media enables populists to reach many voters who have turned their back on traditional outlets (Pew Research Center, 2018). Those users pass populist messages on like chain letters, to other users with similar preferences (Jacobs and Spierings, 2018: 1686; Klinger and Svensson, 2015: 1248–1249). In line with the logic of production, populist parties also use social media to circumvent the evaluation of journalistic gatekeepers (Engesser et al., 2017: 1110, 1113; Jacobs and Spierings, 2018: 1685). Although no party can control political mass media communication, given that the outlet chooses the message’s framing (De Jonge, 2019), it is primarily populists who distrust the mainstream media (Krämer, 2017: 1303). In fact, research on the relationship between the media and populist parties shows that in many countries, populists receive more critical coverage than other parties (De Jonge, 2019; Ernst, et al., 2019a, Esser, et al., 2019b: 183), even in tabloid media (Rooduijn et al., 2014). This prompts populists to seek out alternative avenues of communication. Manucci (2017) differentiates between indirect and direct political communication against the backdrop of the party-voter relationship. Typical interactions between mass media and parties are indirect communication, while social media interaction is a form of direct communication. Accordingly, indirect communication is characterised by dependency on intermediaries, as the outlet decides whether the party’s message will be distributed. Ultimately, potential voters will only see the framed (i.e., mediated) message. In contrast, direct communication allows parties to send out unmediated messages to their supporters. Here, no intermediaries evaluate, process or change the message that parties seek to send out. Populists, who would otherwise be evaluated more critically when using indirect communication channels, can avoid these journalistic gatekeepers by focusing on direct communication strategies (Jacobs and Spierings, 2018: 1685).

Multiple studies have found empirical evidence that populist actors put more emphasis on political social media communication than non-populist actors. Ceccobelli (2018) has shown that populist politicians use social media regularly, while the activity of non-populist politicians is strongly focused on campaigns. This suggests that populists dedicate more attention to their social media communication. For populists, social media is an important communication channel for spreading populist content (Bracciale and Martella, 2017; Engesser et al., 2017). Furthermore, studies have suggested that populist content is more closely associated with Facebook than Twitter (Ernst, et al., 2019b; Jacobs and Spierings, 2018), where populist actors get, on average, more likes and spark more attention (Ceccobelli et al., 2020). Facebook’s text-centred posts are an ideal alternative to mass media communication and provide populists with an unmediated platform to voice their demands in a way that character-limited Tweets do not. Furthermore, Facebook, as the most widespread social media platform, offers the opportunity to communicate with a comparatively broad audience that more closely represents “the people”. In contrast, the audience of other networks is quite specific. For example, Instagram clearly has a target group under 34 (DataReportal, 2021). However, thus far, there is little comparative research on the Facebook activity of populist parties, a research gap that the following analysis seeks to tackle.

Defining and measuring populism

The following analysis relies on the populism index from the Populism and Political Parties Expert Survey (POPPA, Meijers and Zaslove, 2020). The dataset is based on the so-called ideational approach to populism and covers different variables on the subdimensions of populism. The ideational approach, which has become widespread in empirical analyses, defines it as a “unique set of ideas” (Hawkins and Kaltwasser, 2018: 3) centred on the idea that politics is a Manichaean struggle between the “good people” and the “bad elite” (Mudde and Rovira Kaltwasser, 2013). The independent variable for the following analysis is POPPA’s populism score (range: 0–10), which is derived from a factor analysis of the variables measuring the elements of this definition (people-centrism, anti-elitism, people homogeneity, general will, Manichaean worldview). POPPA data was collected in 2018 for 249 parties in 28 European countries. The continuous populism score enables an analysis of the correlation between party-level populism and activity, addressing the question: Are populist parties more active on Facebook than non-populist parties?

Facebook data

I collected Facebook data in 2022 using Crowdtangle, a platform designed to facilitate academic access to Facebook data (Garmur et al., 2019). The year range of 2017–2019 was selected as it includes the year before and after POPPA was conducted based on the assumption that party-level populism did not differ drastically between those years. The Facebook activity for all parties was linked to the POPPA dataset (see Thomeczek, 2023 and Table A4, Appendix). However, there are several caveats in terms of data collection. Firstly, in three cases, the official party websites referenced only the accounts of their leaders (Lega, PVV, DLF), although the parties also had their own accounts. As the goal is to compare party activity, leader accounts were disregarded (but see Figure A5, Appendix for a comparison of the three parties with the activity for their leaders). Secondly, no data could be collected for some parties primarily because of the absence or deletion of accounts (see Appendix, Table A4). Thirdly, access was sometimes restricted to certain geographical areas (geo-blocking).¹ The German SPD is a prominent example of a geo-blocked account. A manual check of account activity over time revealed “activity gaps” that could indicate such restrictions. Therefore, the data was simultaneously collected via the Facebook API access by Germany-based Facepager (Jünger and Keyling, 2019) to minimise the problems arising from missing data. As Crowdtangle’s data is not affected by the Facebook API limit (Ho, 2020), it served as a baseline (574,026 posts). In the second stage, the missing posts collected using Facepager were added to the Crowdtangle dataset (53,913). The final dataset included 627,711 posts from 226 accounts.² The monthly activity pattern of all parties can be found in the Appendix (Figures A1-A4).

Analysis

I calculated the average number of posts sent out per day by each account. The number of days was corrected for accounts created after 2017 and for dissolved parties (see Table A4, Appendix). Table 1 shows the distribution of the two relevant variables. The median of the populism score (3.8) indicates that most parties are located below the middle point of the scale. The differences could not be more drastic regarding the distribution of Facebook activity. There are two accounts for which no posts could be obtained for a specific year (see next section). In contrast, the most active party, Italian right-wing populist Lega, sent out, on average, 34.7 posts/day. The first and third quartile of the activity variable, where 50% of the values are located, ranges from 1.03 to 2.95 posts.

Table 1.

Distribution of populism score and number of posts (aggregated by year).

Variable	Min	Max	1^st Quartile	Mean	Median	3^rd Quartile
Populism Score	0	10	2.5	4.5	3.8	6.7
Average Posts per Day	0	34.7	1.03	2.3	1.7	2.95

Figure 1 gives an overview of a party’s average daily activity (aggregated for the full three-year period) and its populism score. The correlation (

r = 0.22

) is significant

(p < 0.01)

Figure 1.

Post Frequency of Parties in Europe (2017-2019).

The left plot of Figure 1 shows how Lega’s activity level stands out. Although Italian parties were generally among the most active in Europe, in 2018/2019, the Lega account posted almost three times more content than the second-placed party, Italian M5S. The plot on the right shows the logged activity to improve the readability for the non-outlier cases. There were also well-known populist parties from other countries that were highly active on Facebook, such as French Rassemblement National (former Front National, FN), Czech SPD and Spanish Podemos. Nevertheless, there were also counter-examples. The Dutch PVV and German AfD were two relatively inactive populist parties, and there were also examples of particularly active non-populist parties (e.g. Italian PD, Czech ODS and Bulgarian BSP). Table A1 (Appendix) shows that activity is also significantly positively correlated with all subdimensions of populism and the variable measuring party personalisation and significantly negatively correlated with intra-party democracy. Overall, Figure 1 suggests that country-specific differences should be accounted for when analysing the party activity. Therefore, the following sections discuss the results by country, grouped into the EuroVoc regions.³ I will focus on the aggregation by year to avoid active and inactive phases cancelling each other out. Table 2 summarises the correlations and indicates a significant correlation between populism and the number of posts in around half of the countries examined.

Table 2.

Correlation between post activity and populism by Country.

Country	Pearsons’s Correlation Coefficient	Signifiance
Austria	0.31
Belgium-Flanders	0.48	**
Belgium-Wallonia	0.69	***
Bulgaria	−0.31
Croatia	0.39	**
Cyprus	−0.19
Czech Republic	0.25
Denmark	−0.18
Estonia	0.32
Finland	0.61	***
France	0.52	***
Germany	0.039
Greece	0.23
Hungary	−0.36
Ireland	0.67	***
Italy	0.62	***
Lithuania	−0.14
Netherlands	0.4	**
Poland	0.28
Portugal	−0.51	*
Romania	−0.48	**
Slovakia	0.44	**
Slovenia	0.75	***
Spain	0.1
Sweden	0.69	***
Switzerland	−0.24
United Kingdom	0.15

Note: *p< .1; **p< .05; ***p< .01.

Figure 2 illustrates the relationship between populism and the number of daily posts in Western Europe. The Y-axes of the following figures are scaled differently based on the country-specific minimum/maximum activities. Overall, Western Europe is the region with the strongest correlation between populism and Facebook activity (significant in five countries, Table 2). This is most consistent in the Netherlands, Belgium and France. The most active parties in Western Europe were located in the UK, France and Austria, where populist parties posted five to six times per day. In contrast, parties in Germany and Belgium were about half as active. In some countries, populist parties were less active than non-populist parties in parliamentary election years (Austria, 2017; Germany, 2017; UK, 2017). However, there is counter-evidence from Switzerland: the right-wing populist SVP was more active than the other parties in the year of the Swiss elections (2019). The British UKIP showed a high level of activity in 2019. The monthly activity pattern indicates that the party focused on the European campaign (Appendix, Figure A1). In contrast, in Germany, the AfD’s low level of activity suggests that it was almost entirely inactive during the EU election campaign.

Figure 2.

Frequency of Posts in Western Europe.

However, this conclusion seems implausible and could indicate that posts were deleted, possibly related to the internal party struggles that led to the departure of Chairman Meuthen. In the Netherlands, the PVV’s account was relatively inactive. Party communication was mainly done through the personal account of PVV’s leader and only member Wilders (Appendix, Figure A5). In contrast, the right-wing populist party FvD and the left-wing populist party SP were among the most active Dutch parties. In Belgium, where populist parties were very active on Facebook, it seems unlikely that the Flemish Social Democrats did not post any content in 2017, which could indicate gated or deleted posts.

The trend for Central and Eastern Europe is displayed in Figure 3. Generally, post activity was higher than in Western Europe. In five countries, there is a positive correlation between the frequency of posts and party-level populism that is significant for Croatia, Slovakia and Slovenia (Table 2). In Czechia, the two most populist parties showed divergent trends. In Hungary, Romania and Bulgaria, higher populism scores were associated with less activity (significant in Romania). However, in Bulgaria, this was mainly driven by the high activity of the BSP party. In all three countries, the most populist parties governed at least some of the time between 2017 and 2019, which may explain their low activity level, as they had access to government accounts. This trend is also supported by Poland’s governing populist party PiS. On the contrary, governing SDS was Slovenia’s most active party on Facebook. The Polish right-wing populist Kukiz’15 became less active over time, especially after the European elections in 2019 (Appendix, Figure A2). In Slovakia, the positive correlation between populism and Facebook activity seems to decrease over time, but this figure is skewed by the far-right populist LSNS, which only created its Facebook page in late 2019. Overall, some parties in Central and Eastern Europe showed remarkable phases of inactivity (Appendix, Figure A2). In most cases, this was related to newly created pages or the dissolution of parties and, in other cases, to activities strongly focused on election campaigns.

Figure 3.

Frequency of Posts in Central/Eastern Europe.

In Southern Europe, populist parties were more active than other parties in Italy and Greece (Figure 4), but the correlation is only significant in Italy (Table 2). The two most populist Italian parties, Lega and M5S, were also the most active in Europe, whereas the two parties classified as (slightly) less populist, the FdI and FI, were only moderately more active than the non-populist parties. The correlation is negative in Cyprus, but the variation of post frequency in Cyprus was high, while the range of populism scores was small. The situation is similar in Portugal (significant negative correlation), although the most populist party, the CDU, showed highly irregular activities before 2019, which could indicate gating (Appendix, Figure A3). Furthermore, the left-wing populist Bloco de Esquerda is not included in the analysis, as the party only created its account in 2020. In Greece, populist parties were generally more active, although the populist ANEL and non-populist Potami deviated from that trend. A closer examination of the activity patterns reveals that those Greek parties drastically reduced their Facebook activity around the national elections in 2019. ANEL did not compete in the elections, and Potami was dissolved in late 2019. In Spain, the left-wing populist Podemos was the most active party, while the regional populist parties were relatively inactive. This may be related to their regional focus. Moreover, the trend was reversed in 2019 when Spain had two general elections. Here, the two largest parties, the PP and PSOE, which probably had more resources, were unusually active, while the populist parties were less active.

Figure 4.

Frequency of Posts in Southern Europe.

Finally, Figure 5 shows the activity in Northern Europe, which was the lowest of all four regions overall. On average, the most active Nordic parties sent out around three posts per day. Populist parties were more active in Estonia (not significant), Finland and Sweden (both with significant correlations). The relationship is strongest in Sweden, where the Swedish Democrats (SD) were highly active, and in Finland, although here, the most active were the non-populist Christian Democrats (KD). In Denmark, the right-wing populist DF was highly active on Facebook only in 2019, the year of the European and Danish general elections, while right-wing populist newcomer Nye Borgerlige (D) was quite inactive. Nevertheless, both populist parties posted more content before the Danish elections in June 2019 (Appendix, Figure A4). In Lithuania, the three populist parties show very different levels of Facebook activity. While the TT (dissolved in 2020) was almost inactive and the LVLS was highly active, the LCP’s number of posts was not different from the other parties. However, the data for TT is likely incomplete as there is no recorded activity for 2018. Overall, the Lithuanian parties were the least active European parties on Facebook.

Figure 5.

Frequency of Posts in Northern Europe.

As the findings from Figures 2 –5 and Ceccobelli (2018) suggest that populism is not the only factor explaining Facebook activity, multivariate models that control for additional variables can be found in Table A3 (Appendix). Furthermore, I check the robustness of the results using an alternative, binary classification of populist parties based on the PopuList (Rooduijn et al., 2020). The results show that, regardless of the control factors introduced or how populism is measured, the higher average activity of populist parties is confirmed. The models also show that being in government has a negative effect on the number of posts, while a higher seat share positively impacts activity. Furthermore, radicalism positively affects the activity level, which also aligns with the visual evidence that both far-left and far-right populist parties are more active.

Discussion

This research note has tracked whether populist parties are more active on Facebook, something which is often assumed but rarely comparatively tested. To this end, this study has analysed more than 600,000 Facebook posts of 226 European parties.

Overall, in most countries, party-level populism positively correlates with Facebook activity. This relationship is clearest in Western Europe. These results are robust after controlling for factors related to ideology, government status, election campaigns and parliamentary strength. However, the results also highlight several issues and caveats for future studies. Firstly, Facebook activity varied strongly between countries. Generally, party accounts from Western and Northern Europe were the least active, while those from Southern and Eastern Europe showed higher activity. Italy, the country with the most active parties on Facebook, deviated strongly from the European average. Lega, the most active European party, was 20 times more active than the European median party. Secondly, there are several additional factors that should be accounted for in future studies. Election campaigns can affect activity level dynamics, although in various directions. In some countries, the non-populist parties became more active during campaigns (e.g., in Austria); in others, the populist parties became more active before elections (e.g., in Switzerland). Other factors to consider are seat share, government status and ideological radicalism. Thirdly, data should be evaluated critically and cross-checked. There are some apparent “activity gaps”. Most of these gaps in the dataset occurred when parties were dissolved or new accounts were created. However, in other cases, content has likely been restricted or deleted. Combining different data resources helps to tackle this problem but cannot fully solve it. Once deleted, posts cannot be obtained anymore. Although large-n analyses are promising, researchers must perform manual checks and seek to understand deviating patterns. The activities of the German parties in 2019 are a good example. The patterns suggest that the AfD was not as active on Facebook as the other parties during the EU election campaign, but the almost total inactivity during that time is highly implausible. Furthermore, relying only on Crowdtangle data would mean that the SPD activity cannot be accessed.

Overall, many of the most well-known populist parties in Europe are among the most active parties on Facebook in their countries. This is empirical evidence of the theorised connection between populism and social media in general and Facebook in particular. It is quite astonishing as these populist parties differ in many other characteristics, such as ideology, government status and age. Nevertheless, other factors could explain why these parties have turned to social media and future research should focus on whether, for example, the media landscape could play a role. Populist parties may have fewer incentives to use social media if they have more direct mass media access, such as Hungarian Fidesz. Furthermore, Facebook activity is only one facet of social media performance. Research would benefit from a broader approach that includes more social media platforms and different aspects of social media communication. Against the backdrop of the social media logic, it may also be interesting to investigate whether some parties have adapted to the users’ needs more effectively. Elements of interest related to this question include post length, emojis and media usage, which may contribute to simplification and polarisation. Finally, future studies should be more critical of the data obtained and be sure to perform manual checks, as missing data can systematically bias the results.

Supplemental Material

Supplemental Material - Political communication on Facebook: Do populist parties send out more posts?

Supplemental Material for Political communication on Facebook: Do populist parties send out more posts? by Jan Philipp Thomeczek in Party Politics

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received financial support for the research from the Volkswagen Foundation (Az. 96 999).

ORCID iD

Jan Philipp Thomeczek

Supplemental Material

Supplemental material for this article is available online.

Notes

Author biography

Jan Philipp Thomeczek is a research associate working at the University of Münster. In his research, he focuses on various aspects of populism, which include populist parties, populist communication, especially on social media, and populist voters. He is currently working on the PRECEDE project, which is funded by the Volkswagen Foundation.

References

Barrett

Hendrix

Sims

(2021) Polarization Report. New York, USA: New York University. Available at: https://bhr.stern.nyu.edu/polarization-report-page (accessed 09 November 2022).

Bracciale

Martella

(2017) Define the populist political communication style: the case of Italian political leaders on Twitter. Information, Communication and Society 20(9): 1310–1329.

Ceccobelli

(2018) Not every day is election day: a comparative analysis of eighteen election campaigns on Facebook. Journal of Information Technology and Politics 15(2): 122–141.

Ceccobelli

Quaranta

Valeriani

(2020) Citizens’ engagement with popularization and with populist actors on Facebook: a study on 52 leaders in 18 Western democracies. European Journal of Communication 35(5): 435–452.

DataReportal (2021) Social media users. Available at: https://datareportal.com/social-media-users (accessed 09 November 2022).

de Jonge

(2019) The populist radical right and the media in the benelux: friend or foe? The International Journal of Press/Politics 24(2): 189–209.

de Vreese

Esser

Aalberg

, et al. (2018) Populism as an expression of political communication content and style: a new perspective. The International Journal of Press/Politics 23(4): 423–438.

Engesser

Ernst

Esser

, et al. (2017) Populism and social media: How politicians spread a fragmented ideology. Information, Communication and Society 20(8): 1109–1126.

Ernst

Esser

Blassnig

, et al. (2019a) Favorable opportunity structures for populist communication: comparing different types of politicians and issues in social media, television and the press. The International Journal of Press/Politics 24(2): 165–188.

10.

Ernst

Blassnig

Engesser

, et al. (2019b) Populists prefer social media over talk shows: an analysis of populist messages and stylistic elements across six countries. Social Media + Society 5(1): 1–14.

11.

Garmur

King

Mukerjee

et al. (2019) CrowdTangle Platform and API. Harvard Dataverse. doi: 10.7910/DVN/SCCQYD.

12.

Hawkins

Kaltwasser

(2018) Introduction. In: The Ideational Approach to Populism. Concept, Theory, and Analysis. London, UK: Routledge: pp. 1–24.

13.

JCT

(2020) How biased is the sample? Reverse engineering the ranking algorithm of Facebook’s Graph application programming interface. Big Data and Society 7(1): 1–15.

14.

Jacobs

Spierings

(2018) A populist paradise? Examining populists’ Twitter adoption and use. Information, Communication and Society 22(12): 1681–1696.

15.

Jünger

Keyling

(2019) Facepager. An application for automated data retrieval on the web. Available at: https://github.com/strohne/Facepager.

16.

Klinger

Svensson

(2015) The emergence of network media logic in political communication: a theoretical approach. New Media and Society 17(8): 1241–1257.

17.

Krämer

(2017) Populist online practices: the function of the Internet in right-wing populism. Information, Communication and Society 20(9): 1293–1309.

18.

Kriesi

(2014) The Populist Challenge. West European Politics 37(2): 361–378.

19.

Manucci

(2017) Populism and the Media. In: Rovira Kaltwasser

Taggart

Ochoa Espejo

, et al. (eds) The Oxford Handbook of Populism. Oxford: Oxford University Press, pp. 467–488.

20.

Meijers

Zaslove

(2020) Measuring populism in political parties: appraisal of a new approach. Comparative Political Studies 54(2): 372–407.

21.

Mudde

Rovira Kaltwasser

(2013) Populism. In: Freeden

Stears

Sargent

(eds) The Oxford Handbook of Political Ideologies. Oxford: Oxford University Press, pp. 493–512.

22.

Pew Research Center (2018) News Media in Western Europe: Populist Views Divide Public Opinion. Washington, DC: Pew Research Center. Available at: https://www.pewresearch.org/journalism/2018/05/14/in-western-europe-public-attitudes-toward-news-media-more-divided-by-populist-views-than-left-right-ideology/ (accessed 09 November 2022).

23.

Pew Research Center (2022) Teens, social media and technology 2022. Washington, DC: Pew Research Center. Available at: https://www.pewresearch.org/internet/2022/08/10/teens-social-media-and-technology-2022/ (accessed 09 November 2022).

24.

Roemmele

Gibson

(2020) Scientific and subversive: the two faces of the fourth era of political campaigning. New Media and Society 22(4): 595–610.

25.

Rooduijn

de Lange

van der Brug

(2014) A populist zeitgeist? programmatic contagion by populist parties in Western Europe. Party Politics 20(4): 563–575.

26.

Rooduijn

Van Kessel

Froio

, et al. (2020) The PopuList. An Overview of Populist, Far Right, Far Left and Eurosceptic Parties in Europe. Version 2.0. https://popu-list.org/.

27.

Thomeczek

(2023) Facebook Accounts of Political Parties included in the POPPA (2018) Dataset. Open Science Foundation (OSF). doi: 10.17605/OSF.IO/8BWM2.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.47 MB