Unlike most adults, which use the internet as researching and knowledge enhancing tools, young people's consumption of the web revolves around social media. According to CBS, almost 100% of 18 to 25-year-olds use social media on a daily basis (CBS 2019). Indeed a consequent amount of their social life is experienced online, which in turn creates a set of standards and norms to abide by. Despite being a place for like-minded individuals to connect, this also means that social media platforms are prone to forming clusters of hateful and discriminatory behaviour. Across internet users, almost 50% of them have recorded observing racist behaviour very often online (German Landesanstalt für Medien NRW 2018). Hate online has always existed due to the anonymity provided by online platforms. The distance and anonymity between the subject and the receiver of such hateful comments allow for more aggressive behaviour online than in real life. Despite the widespread nature of hate speech, social media platform providers struggle to contain this type of behaviour and to counteract hate speech online.
To begin to understand and counteract discrimination and hate speech online, the Institute Movisie, the School for Applied Sciences Inholland and Diversity Media have been funded by Adessium and VSB Fonds to develop and run a three-year project called #datmeenjeniet (#youmustbekiddin). This project seeks to train, support and follow micro and more established Dutch influencers to try and change their followers from bystanders to upstanders, thus stand up themselves against the hate speech they observe online. As part of the 2020 Amsterdam Digital Methods Initiative’s (DMI) Winter School, the researchers of this report collected data in the name of the #datmeenieniet project. There were two clear aims for this project in order to help the #datmeenjeniet initiative: Mapping out the Dutch influencer ecosystem and understanding the relationship between influencers and anti-discrimination discourse/positivity online. It was found interesting to conduct this research across three social media platforms: Facebook, Instagram and Youtube.
Defining: influence and influencers
Influence is a “process whereby people agree or disagree about appropriate behaviour, form, maintain or change social norms and social conditions that give rise to, and the effects of, such norms” (Turner, 1991, p45). In other words, it represents the alteration in the feeling, behaviour or attitude of an individual or a group of individuals due to a myriad of interacting factors (Rashotte, 2011). According to King (1975), the following factors predispose an individual to be influenced in a certain way: sociological, psychological, cultural, and situational. All these factors create a filter through which an individual will be able to be influenced.
In the project #Youmustbekiddin influencers are seen as micro-celebrities who are active on social media, like Facebook, Instagram and/or YouTube. According to the public and based on their online content, stories and posts, influencers are seen as interesting voices. The scenes or communities they are active in can vary, from fashion and lifestyle to health and sport. They can also differentiate in level of education, religion, ethnicity, gender, age and their amount of followers. Some have millions of followers, others a “few” thousand. Examples of big Dutch influencers are Nikkietutorials, Monica Geuze, Gio and Anna Nooshin.
The influencers who are asked to join the #Youmustbekiddin project are social influencers and are called upstanders. They use their social media channels to speak up for societal topics and issues, like hate speech and discrimination. They can also be active in the communities mentioned above, but social influencers specifically stand up for a specific cause to make a change (Movisie, 2019).
What it is online? Hate Speech Online - The state of it
Hateful and discriminatory language against different groups appear to have become more and more common online. It seems that the social norm that online discrimination is not acceptable is less strong in comparison to real life (Steinfeldt et al., 2010). The amount of complaints about and notifications of online discrimination are increasing (Zitek & Hebl, 2007).
Although there is not much known about how to strategically counter online discrimination, research (Felten & Taouanza, 2017) shows which factors help online hate speech and discrimination to originate as well as to maintain it, namely:
the lack of direct and non-verbal responses to an online comment;
the offender doesn’t see that the victim gets hurt;
the lack of authority;
the lack of people who intervene;
and the feeling of anonymity.
Research shows (Livingstone & Haddon, 2009) that young people themselves see this as a problem as well. European teenagers find that after sharing personal information and seeing unwished porn, the third most online risk is to see hateful or harmful content. They see this as a greater risk than online bullying or receiving unwished sexual comments.Why is it significant to map out the influencer Ecosystem?
What does an overview of Dutch influencers look like, organised by size (in number of followers and traffic) and fields of interest?
Where and how are they involved and active in discussing societal issues (specifically on issues and topics like discrimination, sexism, racism of any kind) and non-political issues?
Where do these two sets of issues touch on each other or overlap?
In order to provide a general overview of the influencer landscape in the Netherlands, two Datasets were provided by the research team. The Dataset “Upstanders“ includes personal contacts with a known reputation for engaging with social issues, whereas the Dataset “Google Research“ is based on basic desk research by the research team. Both lists were based on Instagram influencers. In order to conduct a “cross-platform” - approach to influence in the Netherlands additional Data on YouTube and Facebook was gathered. For one a research towards the corresponding Instagram Influencers on these platforms was conducted. Additionally, for reasons of validity a “Statista“ Dataset, comprising the 10 „Most-followed influencers on Instagram based in the Netherlands as of September 2019“ according to statista.com was included and the 10 most followed Youtubers from the Netherlands according to socialbaker.com. Accordingly, research on Instagram, Facebook and YouTube was conducted. In regards to the amount of data collected 3 Sub-projects were formed:
Exploring the Influencer discourse surrounding discrimination/positivity on Facebook
Mapping out the Dutch Influencer ecosystem on Youtube
Evaluating the relationship between their followers on Instagram
The focus of the Facebook research lied on the interaction of users with content that is associated with specific influencers. The project supplied us with a list based on their Google research, which contained 62 Dutch influencers. The project leaders considered them as major players in the Dutch influencer landscape, which reach a big audience of followers. We decided to add a list of Statista, an online portal for statistics, stating the top ten Dutch influencers with the most Instagram followers (Statista). The combination of both lists enabled us to paint a more distinct picture of the influencer landscape in the Netherlands.
We used BuzzSumo, content research and monitoring online tool, to depict with which kind of content people interact on Facebook. By querying the name of an influencer [“name of specific influencer”] as a keyword (Rogers 27), BuzzSumo provides the posts on Facebook which include the queried name. The results are ranked according to how much the users are interacting with each post. The interaction score of a post is based on the accumulated amount of likes, comments, and shares, a post receives. In our analysis, we decided to focus on the first ten results, since we were interested in what kind of content users engage with the most. Focusing on a smaller sample enabled us to study the posts in more depth, following a qualitative approach. We noted the thumbnails, post titles, and URLs of the posts in a sheet. In this step of the research project, we kept the project list and the Statista list separately. We queried the names on the 15th and 16th of January 2020. Therefore, this research only includes posts until this point in time.
In the following step, we analyzed the extracted content and market positive content with yellow. According to the project, positive content means speaking out against hate speech and discrimination; hence content that is empowering; that helps people, or that supports a social cause. After distinguishing the positive content, we analyzed the influencers which had at least one positive classified post in their top ten results. We analyzed the content and assigned keywords to it. We also paid attention to the source distance: a method “which seeks to measure distance from the top of the web for a given name or sub-issue, in a larger issue space. It is the web equivalent of studying the top of the news” (Rogers 12).
In the last step, we classified the fields in which the influencer with positive content is mainly active. Based on the positive content, we developed categories for topics, which are already debated in the various fields and connected to high interaction. In this board, we merged both lists.
As this report solely focuses on Youtube, the research conducted on Facebook and Instagram is not included in this particular report but can be found in the corresponding Wiki Report. In the particular case of Youtube, it was our main goal to Map ecosystems surrounding actors within our lists of influencers in order to understand the connectedness and placement of actors within the networks surrounding our lists of influencers and to discover other relevant actors and other potentially important “influencers” within the networks, particularly focusing on Dutch influencers.
In order to conduct research on Youtube, a collection of Youtube Account User IDs for all corresponding Instagram influencers in all 4 Data Sets had to be conducted by querying the username and Profile URL. User IDs as shown in the URL can be edited by the Account owner manually e.g.:
UCmKm7HJdOfkWLyml-fzKlVg vs. AFROJACKTV
Therefore the actual ID numbers were extracted through the YouTube ID Identifier called commentspicker.com in those instances. Afterwards, the IDs were introduced to the YouTube Data tool Channel Network. When the relevant search queries were confirmed, channel IDs were manually introduced as seeds to the Tool, which “crawls a network of channels connected via the "featured channels" (and via subscriptions) tab from a list of seeds. Featured channels are retrieved via channels/list#brandingSettings and subscriptions via subscriptions/list.”
The Crawl depth was set to 1, it specified how far from the seeds the script should go. The tool’s output was a Gephi edge table file featuring relationships between channels which allowed us to observe and analyse clusters, as well as to identify prominent channels who were well connected within the network. Many mappings were created (see Appendix):
Figure 2. YouTube channel network surrounding “UPSTANDERS” list
(upstanders highlighted in purple)
Figure 3. YouTube channel network surrounding “UPSTANDERS” list
Figure 4. YouTube channel network surrounding “UPSTANDERS” already involved in the project
Figure 5. YouTube channel network surrounding the project’s “Google Research” list of influencers
Figure 6. Communities within the YouTube channel network surrounding the Google research list of influencers
Figure 7. Full YouTube channel network including all influencer lists, with Dutch and influencers involved in the project, highlighted.
When working with Gephi, the Force Atlas and Force Atlas 2 were used spreading the networks apart in order to have a closer look into the forces between the channels. In order to make the data readable and easier to discuss and analyse, the node’s sizes were changed according to the page rank of the corresponding influencer. This is an algorithm within the network visualisation tool Gephi. According to GitHub it is "an iterative algorithm that measures the importance of each node within the network. The metric assigns each node a probability that is the probability of being at that page after many clicks." In simple terms, page rank is a measure of the significance of a channel within the network, based on how well connected the channel is.
The node colour was set according to the modularity score, which measures how well a network clusters into modular communities. According to GitHub: “A high modularity score indicates sophisticated internal structure. This structure often called a community structure, describes how the network is compartmentalized into sub-networks. These sub-networks (or communities) have been shown to have significant real-world meaning.”
We started with the 10 most followed Instagram influencers. They were then introduced into the Phantombuster Instagram follower collector tool which provides a list of all profiles that follow the corresponding Instagram influencer. As some of our influencers had more than 5 Mio. followers the size of the Dataset would not be operational. For that reason we limited the number of followers to 100.000 per account, so that in the end we had 1.000.000 followers total for all 10 accounts. For reasons of workability the tool was used on every team members’ laptop (9). Some members had to create Instagram accounts for these purposes. Using the Phantombuster was only feasible within the Google Chrome or Mozilla Firefox browser. The number of followers collected per profile was limited to 100.000 and the number of profiles processed per launch was set to only 1. Launches of the tool were set to occur 4 times per working hour. Phantombuster collected around 9.000 followers per launch.
After scraping all followers from Instagram, the data were cleaned from bots and combined all lists of followers into a single spreadsheet with "source" and "target" accounts (source is follower and target is influencer). OpenRefine was used to locate follower accounts who follow more than one influencer. Those who follow just one account were removed. This spreadsheet was imported into Gephi. In Gephi the algorithm Force atlas 2 was applied and modularity for community detection was calculated.
Figure 1: Top 10 influencer of the Statista list
Figure 1 demonstrates the findings retrieved from BuzzSumo relating to the list of top influencers in the Netherlands according to Statista. The order of influencers in the visualization is in line with their amount of followers - Negin Mirsalehi being the one with the most and Nicolette van Dam the one with the least in this list. What immediately becomes apparent from this visualization, is that there is only a scarce number of positive posts - seven posts in total. These are the posts that are marked yellow. Above that, it stands out that the top four Dutch influencers did not depict any positive content on Facebook at all. The first influencer in this list who did engage with positive content on Facebook is Dee, a Dutch singer, and YouTuber. Following up, there is Dutch rapper Boef with two positive posts. These posts were related to helping others in society who are having a hard time, for instance by providing presents for children living in an asylum center for ‘Sinterklaas’ - a Dutch traditional winter holiday. Football player Lieke Martens also engaged in two positive posts, mostly relating to empowering content surrounding the idea that female football players should be equally treated as male players. And lastly, TV host and actress Nicolette van Dam engaged in one positive post that focused on raising awareness for World Children’s Day.
Furthermore, it is important to make three notes about the visualization itself. The squares without thumbnails mean that the thumbnails were missing on BuzzSumo. Therefore we could not include them in our visualization. And both for Lieke Martens and Gio we could not gain ten results of the query, therefore we only used what was available.
Figure 2: Influencer with positive content of the Google research list
Figure 2 illustrates the results retrieved by BuzzSumo through querying the influencer’s names from the Google research list. This time the order is alphabetical according to the Google search list. What becomes apparent from this visualization is that a lot more positive content was found in this list of influencers. We found a total of 61 posts that could be marked as positive - again indicated by the yellow colors. The average of the source distance in this list is five, meaning that on average the positive content would be in fifth place when regarding the ranking of the top 10 most engaged with posts. What stands out from figure 2 is that some influencers in this list are engaged with a lot of positive content on Facebook. To gain a bit more context, the cases containing five or more pieces of positive content will now be further discussed to gain a broader understanding of how these Dutch influencers are engaged in empowering posts on Instagram.
Starting with the influencer with the most positive content, namely Ruba Zai. Ruba Zai is a vlogger, blogger, and Instagrammer who shares her daily life with her followers. The queries results have shown that the positive content had a specific focus on empowerment surrounding hijabs and how to wear them fashionably in the more Western culture. Besides that, she also touches upon topics such as mothership, female empowerment, diversity, and inclusion.
Following up with eight positive items of content is Nikkie de Jager. She is a Dutch influencer who is mainly known for her beauty tutorials on YouTube. The big amount of positive content can be ascribed to a recent video she shared on her YouTube channel in which she revealed to be a transgender woman. This caused a lot of responses online, which were almost all positive. This can be seen as empowerment for the transgender community.
Another influencer that stands out with six yellow posts of positive content is Elio Heres. He is a Dutch gay influencer who shares content on Instagram and YouTube. The positive posts mostly have an emphasis on gay pride, being a gay ambassador in market campaigns and breaking the tabu in showing fashion tips for the gay community. Rachel van Sas (huismuts) also has six yellow posts of positive content. She is a Dutch lifestyle blogger and vlogger who shares content on her blog, YouTube and Instagram. The positive posts here mostly surrounded the topics of mothership, vulnerability, and intuition. The last example, with five items of positive content, is Dutch rapper Lexxxus. Here the positive content mainly came about as a result of the rapper supporting a 15-year-old boy who was being abused by his friends. Above that, Lexxxus tends to spread positivity and kindness under his followers by showing that he is supporting others. Again, it is important to note here that the empty squares in figure 2 mean that the thumbnails were missing on BuzzSumo. Furthermore, for some of the influencers in the list, BuzzSumo did not show a total of ten results. For these influencers - OnneDi, Fadim Kurt, Sisi Bolatini - we have used only the results BuzzSumo did retrieve.
Figure 3: Topics connected with specific fields in which influencers are active
Figure 3 demonstrates the keywords we have connected to each item of positive content we have found on BuzzSumo. After finding the items of positive content - highlighted in yellow in the visualization board - we started to further examine the contexts of the posts, deeply reading each article and mapping the social causes and main terms. Since BuzzSumo provided us with the top ten most engaged with posts for the search query, we were able to notice some of the trendy social topics that people seem to be aware of. To further understand in which ways the social issues are being discussed on Facebook via influencers, we created different categories: vloggers, beauty, fashion, sports, music, and arts. For instance, as shown in figure 3 it becomes apparent that sports are very much related to topics surrounding respect, fair play and feminism. Some topics are not solely connected to one of the categories but seem important in others as well. This goes for the topics of sustainability, feminism, and diversity. They seem significant topics as they all are debated in three different categories. But also the topics against stereotypes, no hate speech, and self-love are debates appearing in two different categories. To gain a better understanding of these debates, we will be depicting it further through specific case studies in the following section.
We have analyzed in total 72 influencers and just 20 of them were associated with positive content, which points to a lack of involvement from influencers in social causes until this moment. Most of the influencers, who had positive content in their sample, are connected to more than one positive post. Thus, this can be evidence that once an influencer starts to get involved in social causes, they become a pillar of their content. The social issue is no longer sporadic content on social networks. It becomes not only part of their identity and personal mission, but also a form of personal branding. Said that it is understood that influencers can create a “positive persona” on social networks. For example, the content linked to Ruba Zai engages specifically with connecting the hijab, an important symbol of the Muslim religion, to the fashion standards of the western culture. Ten out of ten results were marked as positive. This demonstrates that her social media presence has transformed into a social mission. It also indicates the high interest of users on topics addressing the needs of underrepresented communities, suggesting that influencer can have a great impact on topics that currently lack empowerment.
Based on the classifications of the influencer fields, it was possible to notice different approaches to depict the same social cause. For instance, the positive content of the sports segment is mainly related to diversity and feminism. These two topics are important to soccer player Lieke Martens, who connects them to gender equality as well. In one of the analyzed articles, Lieke says: “When I was a child, I wanted to be like Ronaldinho. Nowadays, I met girls that want to be like me” (NOS). Also, Elio Heres, interacting in the fashion segment, identifies himself with the pro-diversity discourse. He is an ambassador of the gay community, who talks about the importance of breaking gender stereotypes and how to include this in an individual's fashion style. So, the first point that should be highlighted is that social issues can be defended from various segments, adding their specific angle to the topic.
Along the same lines of thought, it is also possible to notice that one influencer can be associated with multiple social issues. This statement is illustrated by the case of Nikkie de Jager. The beauty influencer came out as transgender recently. In this context, she speaks not only for the empowerment of transgenders, a community that suffers under the lack of acceptance in society. She also connects this topic to self-love and diversity. Eight out of ten posts associated with her are classified as positive. All of her top five posts are engaging with positivity. This elucidates the importance of authenticity for influencers. It is not only about evolving with social causes, but also about communication through an honest lens.
In summary, positive content can become a strategy leading to a change in the profession of an influencer. These new dialogues that some influencers have been creating are now evolving debates about social causes. So, authenticity and empowerment become social currencies in the online environment of Facebook. This can represent a win-win situation, once the influencer can increase empathy and develop authentic relationships with their community. Moreover, more humanized media content could minimize the association between social media use and anxiety in emerging adults, since the contents do not usually reflect the reality, causing unattainable expectations through comparisons (Vannucci 165). In this line of thinking, Fadin Kurt, an influencer, made a video about unrealistic lives on Instagram, pointing out the shallowness of the influencer business. Also, Influencer Rachel van Sas illustrates empowerment through authenticity. She disillusions maternity, by pointing out her vulnerability, her failings, and learnings. Thus, challenging the stereotypes of perfect families.
Diversity, feminism, and sustainability are topics that were found in three distinct segments. The social issues against stereotypes, no hate speech, and self-love appeared each in two different market segments. It is essential to point out that the content of greater engagement reflects not only the content production of the influencers but also the interest of their followers.
Considering the source distance of the positive content visualized in the maps, we can see that positivity is neither necessarily on the top nor the last ranks. Positive content appears on different positions. Therefore, there is no guarantee that positive content is automatically related to high engagement. In this context, authenticity and the topic itself play an important role, as pointed out earlier.
However, once a piece of positive content is created, it can be replicated various times through different sources and generate the same amount of engagement each time. This example can be perceived through the advertisement of Ruba Zai about how to wear a hijab stylishly. The same post was replicated eight times, but each post showed a similar level of engagement. This shows that the audience is willing to engage with the same content if it is addressing their needs or of value to them.
Our main achievement has been the production of various maps of different ecosystems of influencers on Youtube, in line with the project’s aims. However, during the research period a focus on variations of seven main networks became apparent (see Appendix). These maps reveal that Instagram influence is not necessarily mirrored on Youtube as not all big Instagram influencers had a Youtube account or an aligned amount of followers. Furthermore, the Upstanders do not show up within networks of the Top 10 Dutch YouTube influencers (Fig.1), which indicates that Upstanders do not yet appear within the recognised and popular Dutch influencer networks. This means they are not interconnected enough via featured channels and subscriptions in order to insert themselves in a more influential way, in the sense of the original research project. It also became apparent that other significant influencers according to their page rank showed up in the mappings, who are not yet included in the original target Upstanders list. In the sense of clusters, a first helpful overview was provided by a mapping based on the Google Research list of influencers (Fig.5).
As this project particularly aims at the Dutch influencer landscape, influencers were filtered according to their country-category as shown within the data included in the Gephi file. Only influencers with their categories set to the Netherlands were included, whereas influencers within the “other country” or “no country” categories were excluded. This provided some helpful clusters of key Dutch influencers, and an overview of what kinds of topics Dutch YouTubers operate within (Fig. 6).
Key clusters found are:
Science and Tech
Vloggers (no specific topic)
However, not all maps exhibit such thematic clustering within the networks. Due to a limited amount of time and the overwhelming amount of data collected the research team encountered several limitations, which will be shortly discussed in the following chapter.
As aforementioned, the goal of the research project was to understand discrimination and hate speech online in the Netherlands. During the one-week Data Sprint 2020 at the UvA our small team of students tried to assist this project by mapping key influencers on the three biggest platforms Youtube, Instagram and Facebook. The overall aim was a cross-platform analysis. The first obstacle was that additional data had to be collected in order to secure representativity, which further decreased the already limited amount of time to analyze the data collected. Furthermore, the enormous amounts of data collected led to a division into 3 expert groups respectively focussing on one particular platform. Moreover, the large data sets caused several breakdowns of research programs as well as the respective hardware (eg Laptops). In the particular case of the Youtube research Gephi crashed numerous times leading to delays.
Nevertheless, the research team managed to produce several mappings and conduct the first investigation into concerning clusters. During this first examination, it became apparent that the investigation of the interconnectedness of channels via subscription and featured channel networks allow for a first overview of an influencer landscape. It does, however, require a closer look into IN - and OUT - Going connections. As some Youtubers such as Rui Jun Luong follow more than 700 channels on Youtube and therefore appear more connected than others. But she herself only has approx. 5000 subscribers and does not appear in other users featured channels at all.. Gephi shows this as a strong interconnectedness it does, however, not mean that she is a key player in the Dutch influencer landscape. This circumstance needs to be taken into account when further investigating the mappings.
It should be noted that due to the limited amount of time all visualizations require further analysis in order to extract valuable insights. Additionally, no content analysis was undertaken, the mappings only hint towards key influencers and thematic clusters within which they act. They do not reflect on opinions or behaviour towards hate speech and discrimination this would need close reading/viewings of comments and videos. Finally, all three platforms need comparison and further research so that one may fully understand the overall Dutch Influencer and Upstanders landscape.
The result for the Instagram analysis were two graphs, each showing the influencers and their followers according to the collected data.
The first chart showed the top 10 influencers of the Netherlands and from each of the 100,000 followers those who followed at least two of these influencers. Half of these influencers achieved their popularity not only through social media. They became famous through music, sports or TV. For the influencers Gephi formed five categories: green for daily life vlogger (dee and gio), pink for rapper (lilkleine and boef), blue for tv-celebrities (nicolettevandam1 and monicageuze), okker for football (liekemartens) and light orange for influencers of the "ideal" life about beauty, fitness and richness (negin_mirsalehi, tavicastro, eswaratti). The followers are assigned as individual points to the influencers and categories. The closer a follower's point is to the influencer's point, the stronger the connection. The colour of the follower points indicates the category to which they belong. The pink and green follower clouds at the top right and left show that both rappers and lifestyle vloggers have a large and strong follower community. They also have the largest number of shared followers among the top 10. These influencers seem to serve topics that a large number of followers are interested in. The tv-celebrities also have a large, cohesive fan community (blue cloud in the middle left). The influencers of the "ideal" life have a less strong connection to the other influencers because they are further away from the center. Furthermore, there are only a few followers who follow one of these channels and at the same time another one from the top 10. Although their accounts serve similar themes, they do not have a large, strong common follower community.
The second graphic shows the "upstanders" and their shared followers. Here too, categories were formed according to their affiliation with Gephi, which are visible in the different colors. However, there are not as clear thematic references as in the Dutch Top 10 Influencers. Only in the bottom middle there is a purple cluster with Arabic cultural background. Overall, they have not yet built up their follower community as effectively as the top influencers. But some of the upstanders are already building a similarly strong follower community, especially the influencers of the green category on the right. For the project #youmustbekiddin' this means that compared to the top influencers it will be more difficult to reach a large and strong community of interests with a single influencer. Their communication strategy should therefore include several " upstanders ", but especially the green and pink category influencers. Because here they can be sure that the message is carried into the follower community from multiple sources.
In conclusion, in order to help the project #datmeenjeniet and as part of the Winter School 2020 three sub-projects were conducted on YouTube, Instagram and Facebook in order to map out the Dutch Influencer Ecosystem and understand its relationship to hate speech online.
CBS. (2019, October). Internet; toegang, gebruik en faciliteiten. opendata.cbs.nl/statline/#/CBS/nl/dataset/83429NED/table?ts=1573200013592.
FELTEN, H. & Taouanza, I. (2017). Online discriminatie aanpakken: wat werkt? Movisie.
KING, S.W. (1975). Communication and Social Influence. London: Addison Wesley.
LIVINGSTONE, S., HADDON, L. (2009) EU Kids Online: Final report. LSE, London: EU Kids Online. (EC Safer Internet Plus Programme Deliverable D6.5)
MOVISIE (2019). #datmeenjeniet - Samen met jongeren discriminatie op sociale media tegengaan. Basisinformatie financieringsaanvraag.
RASHOTTE, L. (2011). Social Influence. Oxford: Blackwell Publishing.
SOBANDE, F. (2017). Watching me watching you: Black women in Britain on YouTube. European Journal of Cultural Studies, 20(6), 655-671.
STEINFELT, J.A. et al (2010). Racism in the electronic age: Role of online forums in expressing racial attitudes about American Indians. Journal of Cultural diversity & ethnic minority psychology 2010 July: 16(3): 362-71.
TURNER, J. (1991). Social Influence. Buckingham: Open Press University.
ZITEK, E. M. & Hebl, M. R. (2007). The role of social norm clarity in the influenced expression of prejudice over time. Journal of Experimental Social Psychology v43 n6 (200711): 867-876.
|jpg||Areagraph epoche.jpg||manage||202 K||21 Oct 2019 - 13:30||EmilieDeKeulenaar|
|jpg||Areagraph_03_Tavola disegno 1.jpg||manage||302 K||21 Oct 2019 - 13:36||EmilieDeKeulenaar|
|jpg||Atlantis_WikiTimeline_Tavola disegno 1.jpg||manage||86 K||21 Oct 2019 - 13:28||EmilieDeKeulenaar|
|jpg||Crusade_WikiTimeline-02.jpg||manage||70 K||21 Oct 2019 - 13:27||EmilieDeKeulenaar|
|png||Screenshot 2019-07-22 at 15.22.51.png||manage||429 K||21 Oct 2019 - 13:20||EmilieDeKeulenaar|
|png||Screenshot 2019-07-22 at 16.42.17.png||manage||527 K||21 Oct 2019 - 13:37||EmilieDeKeulenaar|
|png||Screenshot 2019-07-23 at 12.25.46.png||manage||60 K||21 Oct 2019 - 13:24||EmilieDeKeulenaar|
|png||Screenshot 2019-07-23 at 16.10.01.png||manage||327 K||21 Oct 2019 - 13:31||EmilieDeKeulenaar|
|jpg||WW2_WikiTimeline-03.jpg||manage||66 K||21 Oct 2019 - 13:28||EmilieDeKeulenaar|
|png||cluster 2.png||manage||1 MB||21 Oct 2019 - 13:44||EmilieDeKeulenaar|
|png||image-wall-e3b55f6d8e296e95f13bd18fc943dd55.png||manage||934 K||21 Oct 2019 - 13:33||EmilieDeKeulenaar|
|png||pasted image 0.png||manage||1 MB||21 Oct 2019 - 13:23||EmilieDeKeulenaar|
|png||pasted image 2.png||manage||1 MB||21 Oct 2019 - 13:32||EmilieDeKeulenaar|
|png||unnamed-2.png||manage||12 K||21 Oct 2019 - 13:34||EmilieDeKeulenaar|
|png||unnamed-3.png||manage||11 K||21 Oct 2019 - 13:34||EmilieDeKeulenaar|
|png||unnamed-4.png||manage||54 K||21 Oct 2019 - 13:37||EmilieDeKeulenaar|