Platformization

Published

January 30, 2026

AI summaries

Note that all the summary texts on this page are currently generated automatically using a Large Language Model (Google’s Gemini), based on the results tables and other information from the data. Therefore, please do not cite the verbal summary, but refer to the figures and tables for accurate information.

Video and text features

Video format

We analyzed the share of native vertical videos posted by news outlets on various platforms, observing that 76% of all short video posts were in the native vertical format. While the overall adoption was high, there were notable differences across platforms, outlet types, and countries. On platforms, TikTok had the highest share with 79% of its videos being native vertical, followed by YouTube Shorts at 75%, and Instagram slightly lower at 71%. Significant differences emerged by outlet type: public broadcasters led with 88% native vertical videos, and print outlets and digital native outlets closely followed around 80-81%. In contrast, private broadcasters showed a substantially lower adoption, with only 61% of their videos being native vertical. Furthermore, the share of native vertical videos varied widely by country, from a low of 18% in South Korea to a high of 96% in Switzerland.

Overall average

estimate p.value conf.low conf.high
0.76 0 0.71 0.8

Contrasts

term contrast estimate p.value conf.low conf.high
outlet_type Print - Digital -0.02 0.64 -0.08 0.05
outlet_type Private broadcaster - Digital -0.17 0.00 -0.24 -0.09
outlet_type Public broadcaster - Digital -0.04 0.39 -0.12 0.05
platform Instagram - TikTok -0.05 0.00 -0.09 -0.02
platform YouTube Shorts - TikTok 0.01 0.41 -0.02 0.05

Predictions by platform

platform estimate p.value conf.low conf.high
TikTok 0.79 0 0.75 0.83
Instagram 0.71 0 0.66 0.76
YouTube Shorts 0.75 0 0.70 0.80

Predictions by outlet type

outlet_type estimate p.value conf.low conf.high
Digital 0.79 0 0.71 0.87
Print 0.81 0 0.76 0.87
Private broadcaster 0.61 0 0.55 0.67
Public broadcaster 0.88 0 0.82 0.93

Predictions by country

country estimate p.value conf.low conf.high region
Argentina 0.62 0 0.55 0.68 Latin America
Australia 0.91 0 0.89 0.94 Asia-Pacific
Austria 0.85 0 0.80 0.89 Western Europe
Belgium 0.94 0 0.92 0.96 Western Europe
Brazil 0.57 0 0.50 0.64 Latin America
Bulgaria 0.87 0 0.84 0.90 Eastern Europe
Canada 0.86 0 0.83 0.90 North America
Chile 0.73 0 0.67 0.80 Latin America
Colombia 0.64 0 0.57 0.72 Latin America
Croatia 0.78 0 0.73 0.82 Southern Europe
Czech Republic 0.82 0 0.78 0.87 Eastern Europe
Denmark 0.83 0 0.78 0.87 Northern Europe
Finland 0.89 0 0.86 0.93 Northern Europe
France 0.92 0 0.89 0.95 Western Europe
Germany 0.93 0 0.91 0.95 Western Europe
Greece 0.56 0 0.50 0.63 Southern Europe
Hungary 0.65 0 0.60 0.71 Eastern Europe
India 0.82 0 0.78 0.86 Asia-Pacific
Indonesia 0.71 0 0.64 0.78 Asia-Pacific
Ireland 0.89 0 0.86 0.92 Northern Europe
Italy 0.81 0 0.76 0.86 Southern Europe
Japan 0.38 0 0.33 0.44 Asia-Pacific
Kenya 0.72 0 0.68 0.76 Africa
Malaysia 0.68 0 0.62 0.73 Asia-Pacific
Mexico 0.85 0 0.81 0.89 Latin America
Morocco 0.52 0 0.44 0.59 Africa
Netherlands 0.92 0 0.89 0.94 Western Europe
Nigeria 0.72 0 0.69 0.76 Africa
Norway 0.92 0 0.89 0.94 Northern Europe
Peru 0.45 0 0.38 0.53 Latin America
Philippines 0.43 0 0.38 0.49 Asia-Pacific
Poland 0.76 0 0.71 0.82 Eastern Europe
Portugal 0.71 0 0.64 0.78 Southern Europe
Romania 0.51 0 0.45 0.57 Eastern Europe
Singapore 0.90 0 0.87 0.92 Asia-Pacific
Slovakia 0.85 0 0.82 0.89 Eastern Europe
South Africa 0.79 0 0.74 0.83 Africa
South Korea 0.18 0 0.13 0.22 Asia-Pacific
Spain 0.84 0 0.80 0.89 Southern Europe
Sweden 0.95 0 0.93 0.97 Northern Europe
Switzerland 0.96 0 0.95 0.98 Western Europe
Taiwan 0.43 0 0.38 0.47 Asia-Pacific
Thailand 0.57 0 0.50 0.64 Asia-Pacific
Turkey 0.49 0 0.42 0.56 Southern Europe
United Kingdom 0.86 0 0.82 0.90 Northern Europe
United States 0.87 0 0.84 0.91 North America

Emoji use

We observed that, on average, approximately 25% of short video news posts included emojis. This usage varied significantly across platforms: posts on Instagram featured emojis most frequently, appearing in 41% of videos, a substantially higher rate than on TikTok (18%) and YouTube Shorts (6%). We found only small and non-significant differences in emoji use among different outlet types, such as digital-only news organizations, print outlets, private broadcasters, and public broadcasters. Emoji adoption also varied widely by country, ranging from a low of 2% in Japan and Malaysia to a high of 63% in the Czech Republic and Spain.

Overall average

estimate p.value conf.low conf.high
0.25 0 0.2 0.29

Contrasts

term contrast estimate p.value conf.low conf.high
outlet_type Print - Digital -0.05 0.09 -0.11 0.01
outlet_type Private broadcaster - Digital -0.01 0.86 -0.07 0.06
outlet_type Public broadcaster - Digital 0.06 0.14 -0.02 0.14
platform TikTok - Instagram -0.18 0.00 -0.20 -0.16
platform YouTube Shorts - Instagram -0.29 0.00 -0.32 -0.26

Predictions by platform

platform estimate p.value conf.low conf.high
Instagram 0.41 0 0.35 0.47
TikTok 0.18 0 0.14 0.23
YouTube Shorts 0.06 0 0.04 0.08

Predictions by outlet type

outlet_type estimate p.value conf.low conf.high
Digital 0.31 0 0.24 0.38
Print 0.22 0 0.17 0.27
Private broadcaster 0.24 0 0.19 0.29
Public broadcaster 0.24 0 0.18 0.30

Predictions by country

country estimate p.value conf.low conf.high region
Argentina 0.58 0 0.49 0.67 Latin America
Australia 0.07 0 0.05 0.09 Asia-Pacific
Austria 0.30 0 0.25 0.36 Western Europe
Belgium 0.31 0 0.24 0.38 Western Europe
Brazil 0.32 0 0.24 0.40 Latin America
Bulgaria 0.39 0 0.33 0.46 Eastern Europe
Canada 0.03 0 0.02 0.04 North America
Chile 0.54 0 0.46 0.63 Latin America
Colombia 0.34 0 0.29 0.40 Latin America
Croatia 0.40 0 0.33 0.48 Southern Europe
Czech Republic 0.63 0 0.57 0.70 Eastern Europe
Denmark 0.28 0 0.24 0.31 Northern Europe
Finland 0.34 0 0.26 0.43 Northern Europe
France 0.22 0 0.16 0.28 Western Europe
Germany 0.06 0 0.04 0.08 Western Europe
Greece 0.26 0 0.20 0.31 Southern Europe
Hungary 0.45 0 0.40 0.49 Eastern Europe
India 0.11 0 0.07 0.14 Asia-Pacific
Indonesia 0.12 0 0.08 0.16 Asia-Pacific
Ireland 0.08 0 0.05 0.11 Northern Europe
Italy 0.12 0 0.08 0.15 Southern Europe
Japan 0.02 0 0.01 0.02 Asia-Pacific
Kenya 0.10 0 0.06 0.14 Africa
Malaysia 0.02 0 0.01 0.03 Asia-Pacific
Mexico 0.18 0 0.13 0.23 Latin America
Morocco 0.29 0 0.25 0.34 Africa
Netherlands 0.10 0 0.07 0.14 Western Europe
Nigeria 0.15 0 0.12 0.19 Africa
Norway 0.39 0 0.32 0.46 Northern Europe
Peru 0.50 0 0.43 0.57 Latin America
Philippines 0.16 0 0.12 0.20 Asia-Pacific
Poland 0.23 0 0.17 0.29 Eastern Europe
Portugal 0.25 0 0.20 0.30 Southern Europe
Romania 0.23 0 0.17 0.30 Eastern Europe
Singapore 0.13 0 0.08 0.17 Asia-Pacific
Slovakia 0.60 0 0.51 0.68 Eastern Europe
South Africa 0.09 0 0.06 0.12 Africa
South Korea 0.03 0 0.02 0.04 Asia-Pacific
Spain 0.63 0 0.55 0.70 Southern Europe
Sweden 0.10 0 0.06 0.13 Northern Europe
Switzerland 0.56 0 0.47 0.65 Western Europe
Taiwan 0.21 0 0.15 0.27 Asia-Pacific
Thailand 0.16 0 0.12 0.20 Asia-Pacific
Turkey 0.11 0 0.07 0.15 Southern Europe
United Kingdom 0.06 0 0.04 0.08 Northern Europe
United States 0.03 0 0.02 0.04 North America

Hashtag use

Overall, 46% of short video posts by news outlets included hashtags. Hashtag use varied significantly across platforms, with posts on Instagram including hashtags 71% of the time, substantially more than on TikTok or YouTube Shorts, where only 31% of posts featured them. Differences by outlet type were less pronounced; public broadcasters used hashtags in 53% of their posts, slightly more than digital or private broadcasters (both around 48%), and more than print outlets, which included hashtags in 38% of posts. We observed considerable variation in hashtag use across countries, ranging from a low of 8% in Sweden to highs of 82% in South Korea and 78% in Japan.

Overall average

estimate p.value conf.low conf.high
0.46 0 0.42 0.5

Contrasts

term contrast estimate p.value conf.low conf.high
outlet_type Print - Digital -0.05 0.16 -0.12 0.02
outlet_type Private broadcaster - Digital 0.02 0.48 -0.04 0.09
outlet_type Public broadcaster - Digital 0.04 0.36 -0.05 0.13
platform TikTok - Instagram -0.42 0.00 -0.44 -0.40
platform YouTube Shorts - Instagram -0.42 0.00 -0.44 -0.39

Predictions by platform

platform estimate p.value conf.low conf.high
Instagram 0.71 0 0.67 0.75
TikTok 0.31 0 0.26 0.35
YouTube Shorts 0.31 0 0.27 0.36

Predictions by outlet type

outlet_type estimate p.value conf.low conf.high
Digital 0.48 0 0.41 0.55
Print 0.38 0 0.33 0.43
Private broadcaster 0.48 0 0.43 0.53
Public broadcaster 0.53 0 0.45 0.62

Predictions by country

country estimate p.value conf.low conf.high region
Argentina 0.44 0 0.39 0.49 Latin America
Australia 0.44 0 0.39 0.49 Asia-Pacific
Austria 0.42 0 0.37 0.46 Western Europe
Belgium 0.35 0 0.30 0.39 Western Europe
Brazil 0.55 0 0.50 0.60 Latin America
Bulgaria 0.55 0 0.50 0.60 Eastern Europe
Canada 0.28 0 0.24 0.32 North America
Chile 0.44 0 0.40 0.48 Latin America
Colombia 0.53 0 0.48 0.58 Latin America
Croatia 0.68 0 0.63 0.72 Southern Europe
Czech Republic 0.42 0 0.38 0.46 Eastern Europe
Denmark 0.18 0 0.16 0.21 Northern Europe
Finland 0.28 0 0.23 0.32 Northern Europe
France 0.17 0 0.15 0.20 Western Europe
Germany 0.60 0 0.56 0.64 Western Europe
Greece 0.61 0 0.57 0.64 Southern Europe
Hungary 0.50 0 0.46 0.55 Eastern Europe
India 0.55 0 0.50 0.59 Asia-Pacific
Indonesia 0.43 0 0.38 0.48 Asia-Pacific
Ireland 0.37 0 0.32 0.41 Northern Europe
Italy 0.32 0 0.28 0.35 Southern Europe
Japan 0.78 0 0.75 0.81 Asia-Pacific
Kenya 0.61 0 0.56 0.66 Africa
Malaysia 0.31 0 0.27 0.35 Asia-Pacific
Mexico 0.57 0 0.51 0.62 Latin America
Morocco 0.48 0 0.44 0.51 Africa
Netherlands 0.53 0 0.49 0.58 Western Europe
Nigeria 0.59 0 0.56 0.62 Africa
Norway 0.52 0 0.48 0.56 Northern Europe
Peru 0.55 0 0.50 0.60 Latin America
Philippines 0.44 0 0.40 0.48 Asia-Pacific
Poland 0.40 0 0.35 0.45 Eastern Europe
Portugal 0.44 0 0.39 0.48 Southern Europe
Romania 0.35 0 0.30 0.39 Eastern Europe
Singapore 0.45 0 0.41 0.50 Asia-Pacific
Slovakia 0.44 0 0.39 0.49 Eastern Europe
South Africa 0.59 0 0.54 0.64 Africa
South Korea 0.82 0 0.79 0.85 Asia-Pacific
Spain 0.42 0 0.38 0.46 Southern Europe
Sweden 0.08 0 0.06 0.10 Northern Europe
Switzerland 0.48 0 0.43 0.53 Western Europe
Taiwan 0.65 0 0.60 0.69 Asia-Pacific
Thailand 0.50 0 0.45 0.54 Asia-Pacific
Turkey 0.55 0 0.50 0.59 Southern Europe
United Kingdom 0.34 0 0.29 0.38 Northern Europe
United States 0.22 0 0.18 0.26 North America

Audience interactions

Addressing the audience

Across all platforms and news outlets, approximately 5% of short video posts from news outlets worldwide directly addressed the audience. This average masked some differences by platform, as posts on Instagram addressed the audience slightly more at 6%, followed by TikTok at 5%. YouTube Shorts, however, showed a significantly lower rate, with only 3% of its posts directly engaging the audience, a notable 2 percentage points less than TikTok. When examining outlet type, the differences were small and not statistically significant, with print outlets having the highest proportion at 7% and private broadcasters the lowest at 4%. We observed some variation across countries, where Brazil stood out with the highest proportion of posts addressing the audience at 8%, while several countries, including the Netherlands, reported the lowest at 3%.

Overall average

estimate p.value conf.low conf.high
0.05 0 0.03 0.06

Contrasts

term contrast estimate p.value conf.low conf.high
outlet_type Print - Digital 0.01 0.44 -0.02 0.05
outlet_type Private broadcaster - Digital -0.02 0.22 -0.05 0.01
outlet_type Public broadcaster - Digital 0.01 0.76 -0.03 0.04
platform Instagram - TikTok 0.01 0.19 -0.01 0.03
platform YouTube Shorts - TikTok -0.02 0.01 -0.04 -0.01

Predictions by platform

platform estimate p.value conf.low conf.high
TikTok 0.05 0 0.03 0.07
Instagram 0.06 0 0.04 0.09
YouTube Shorts 0.03 0 0.01 0.04

Predictions by outlet type

outlet_type estimate p.value conf.low conf.high
Digital 0.06 0 0.03 0.09
Print 0.07 0 0.04 0.09
Private broadcaster 0.04 0 0.02 0.05
Public broadcaster 0.05 0 0.02 0.07

Predictions by country

country estimate p.value conf.low conf.high region
Argentina 0.05 0 0.03 0.07 Latin America
Australia 0.04 0 0.03 0.06 Asia-Pacific
Austria 0.06 0 0.04 0.08 Western Europe
Belgium 0.04 0 0.03 0.06 Western Europe
Brazil 0.08 0 0.06 0.10 Latin America
Bulgaria 0.04 0 0.03 0.06 Eastern Europe
Canada 0.05 0 0.03 0.07 North America
Chile 0.04 0 0.03 0.05 Latin America
Colombia 0.05 0 0.03 0.07 Latin America
Croatia 0.05 0 0.03 0.06 Southern Europe
Czech Republic 0.04 0 0.03 0.05 Eastern Europe
Denmark 0.07 0 0.05 0.09 Northern Europe
Finland 0.05 0 0.03 0.06 Northern Europe
France 0.04 0 0.03 0.06 Western Europe
Germany 0.07 0 0.05 0.09 Western Europe
Greece 0.04 0 0.03 0.05 Southern Europe
Hungary 0.04 0 0.02 0.05 Eastern Europe
India 0.05 0 0.03 0.07 Asia-Pacific
Indonesia 0.06 0 0.04 0.08 Asia-Pacific
Ireland 0.04 0 0.03 0.06 Northern Europe
Italy 0.06 0 0.04 0.08 Southern Europe
Japan 0.04 0 0.02 0.05 Asia-Pacific
Kenya 0.05 0 0.03 0.06 Africa
Malaysia 0.04 0 0.03 0.05 Asia-Pacific
Mexico 0.05 0 0.03 0.06 Latin America
Morocco 0.05 0 0.03 0.06 Africa
Netherlands 0.03 0 0.02 0.05 Western Europe
Nigeria 0.05 0 0.03 0.06 Africa
Norway 0.07 0 0.05 0.09 Northern Europe
Peru 0.05 0 0.03 0.06 Latin America
Philippines 0.04 0 0.03 0.06 Asia-Pacific
Poland 0.05 0 0.03 0.07 Eastern Europe
Portugal 0.04 0 0.03 0.06 Southern Europe
Romania 0.05 0 0.03 0.06 Eastern Europe
Singapore 0.05 0 0.03 0.07 Asia-Pacific
Slovakia 0.06 0 0.04 0.08 Eastern Europe
South Africa 0.04 0 0.03 0.06 Africa
South Korea 0.04 0 0.03 0.05 Asia-Pacific
Spain 0.06 0 0.04 0.07 Southern Europe
Sweden 0.04 0 0.03 0.06 Northern Europe
Switzerland 0.07 0 0.05 0.10 Western Europe
Taiwan 0.04 0 0.02 0.05 Asia-Pacific
Thailand 0.07 0 0.04 0.09 Asia-Pacific
Turkey 0.04 0 0.03 0.06 Southern Europe
United Kingdom 0.04 0 0.03 0.06 Northern Europe
United States 0.05 0 0.03 0.07 North America

Interaction calls

Overall, we observed that 24% of short video posts from news outlets across different platforms worldwide called for audience interactions. A closer look at specific platforms revealed substantial and statistically significant differences: posts on Instagram had the highest call for interaction at 50%, significantly more than posts on TikTok (16%), while YouTube Shorts posts were the least likely to include such calls, at just 7%. We found significant variation among outlet types as well; digital-native outlets encouraged interaction in 37% of their posts, a notably higher proportion compared to private broadcasters (18%) and public broadcasters (25%). Print outlets, with 27%, fell between these groups. Furthermore, the prevalence of interaction calls varied considerably by country, ranging from a low of 10% in South Korea to a high of 46% in Slovakia.

Overall average

estimate p.value conf.low conf.high
0.24 0 0.21 0.27

Contrasts

term contrast estimate p.value conf.low conf.high
outlet_type Print - Digital -0.09 0.06 -0.18 0.00
outlet_type Private broadcaster - Digital -0.19 0.00 -0.27 -0.10
outlet_type Public broadcaster - Digital -0.16 0.00 -0.26 -0.05
platform Instagram - TikTok 0.38 0.00 0.33 0.42
platform YouTube Shorts - TikTok -0.08 0.00 -0.11 -0.05

Predictions by platform

platform estimate p.value conf.low conf.high
TikTok 0.16 0 0.11 0.20
Instagram 0.50 0 0.45 0.54
YouTube Shorts 0.07 0 0.04 0.09

Predictions by outlet type

outlet_type estimate p.value conf.low conf.high
Digital 0.37 0 0.30 0.44
Print 0.27 0 0.22 0.32
Private broadcaster 0.18 0 0.14 0.22
Public broadcaster 0.25 0 0.17 0.34

Predictions by country

country estimate p.value conf.low conf.high region
Argentina 0.25 0 0.23 0.28 Latin America
Australia 0.27 0 0.23 0.30 Asia-Pacific
Austria 0.33 0 0.29 0.36 Western Europe
Belgium 0.22 0 0.19 0.26 Western Europe
Brazil 0.27 0 0.24 0.30 Latin America
Bulgaria 0.33 0 0.30 0.37 Eastern Europe
Canada 0.22 0 0.19 0.25 North America
Chile 0.23 0 0.20 0.26 Latin America
Colombia 0.29 0 0.26 0.33 Latin America
Croatia 0.26 0 0.22 0.29 Southern Europe
Czech Republic 0.28 0 0.25 0.31 Eastern Europe
Denmark 0.23 0 0.20 0.26 Northern Europe
Finland 0.30 0 0.27 0.34 Northern Europe
France 0.13 0 0.11 0.15 Western Europe
Germany 0.24 0 0.21 0.27 Western Europe
Greece 0.24 0 0.21 0.27 Southern Europe
Hungary 0.24 0 0.21 0.27 Eastern Europe
India 0.15 0 0.12 0.17 Asia-Pacific
Indonesia 0.28 0 0.25 0.32 Asia-Pacific
Ireland 0.24 0 0.20 0.27 Northern Europe
Italy 0.15 0 0.13 0.18 Southern Europe
Japan 0.14 0 0.12 0.17 Asia-Pacific
Kenya 0.25 0 0.22 0.29 Africa
Malaysia 0.26 0 0.22 0.29 Asia-Pacific
Mexico 0.24 0 0.21 0.27 Latin America
Morocco 0.15 0 0.13 0.18 Africa
Netherlands 0.15 0 0.13 0.17 Western Europe
Nigeria 0.29 0 0.25 0.32 Africa
Norway 0.35 0 0.31 0.39 Northern Europe
Peru 0.27 0 0.24 0.30 Latin America
Philippines 0.19 0 0.17 0.22 Asia-Pacific
Poland 0.23 0 0.20 0.26 Eastern Europe
Portugal 0.30 0 0.27 0.34 Southern Europe
Romania 0.33 0 0.29 0.37 Eastern Europe
Singapore 0.23 0 0.20 0.27 Asia-Pacific
Slovakia 0.46 0 0.42 0.50 Eastern Europe
South Africa 0.28 0 0.24 0.32 Africa
South Korea 0.10 0 0.09 0.12 Asia-Pacific
Spain 0.26 0 0.23 0.29 Southern Europe
Sweden 0.25 0 0.22 0.28 Northern Europe
Switzerland 0.33 0 0.29 0.36 Western Europe
Taiwan 0.23 0 0.20 0.26 Asia-Pacific
Thailand 0.21 0 0.18 0.24 Asia-Pacific
Turkey 0.15 0 0.13 0.18 Southern Europe
United Kingdom 0.20 0 0.18 0.23 Northern Europe
United States 0.23 0 0.20 0.27 North America

Measures

Video format was coded using zero-shot classification with multimodal LLMs using video frames. Emoji use was measured using the emo_detect() function from the {emoji} package (Hvitfeldt 2024). Audience interactions were coded using zero-shot classification with multimodal LLMs using post text, transcript and video frames.

Your task is to analyze the provided short video content based on the provided video frames, text and transcript. Do not infer beyond what you can directly observe in the text or images. Please evaluate the provided video content against each of the following 4 characteristics independently:

direct_audience_address: Code “True” if the narrator speaks directly to the viewer or the text addresses the audience using direct pronouns or imperatives (e.g., “Have you heard?”, “You need to see this,” or a person pointing at the camera). Code “False” if the content is purely observational or third-person reporting.

community_mentions: Code “True” if the video, caption, or transcript mentions specific audience members, usernames, or community feedback. This includes “Replying to (user?),” screenshots of comments, or phrases like “You guys have been asking” or “In the comments, people said.” Code “False” otherwise.

call_to_like: Code “True” if there is a verbal or visual prompt to “like,” “heart,” “save,” “bookmark,” or “upvote” (e.g., “Double tap if you agree,” “Save this for later”). Code “False” otherwise.

call_to_comment_discuss: Code “True” if there is a verbal or visual prompt to write a comment, answer a specific question, or engage in a discussion (e.g., “Tell us your thoughts,” “What do you think?”). Code “False” otherwise.

call_to_share: Code “True” if there is a verbal or visual prompt to share the video, send it to others, or repost it to a story/feed. Code “False” otherwise.

call_to_follow_subscribe: Code “True” if there is a verbal or visual prompt to follow the account, subscribe to the channel, or visit the profile/page for more content (e.g., “Hit the plus sign,” “Follow for more”). Code “False” otherwise.

call_to_website_app: Code “True” if there is a verbal or visual prompt to visit the outlet’s own digital products, such as a website for a full article or an app (e.g., “Link in bio,” “Read the full story on our site”). Code “False” otherwise.

call_to_traditional_media: Code “True” if there is a verbal or visual prompt to consume the outlet’s traditional media, such as watching a TV broadcast, listening to a radio show, or reading a print edition (e.g., “Watch the full report tonight on [Channel Name]”). Code “False” otherwise.

call_to_social_platform: Code “True” if there is a verbal or visual prompt to visit the outlet on a specifically mentioned social media platform (e.g., “Check our YouTube,” “Live updates on our Twitter”). Code “False” otherwise.

Reliability

Models used: gemini-2.5-flash-lite, Qwen3 235B VL.

Variable Posts Models Agreement Krippendorffs_Alpha
NA 186 2 0.88 0.40
NA 186 2 0.98 0.39
NA 186 2 1.00 1.00
NA 186 2 0.97 0.77
NA 186 2 1.00 1.00
NA 186 2 1.00 1.00
NA 186 2 0.99 0.96
NA 186 2 0.99 0.80
NA 186 2 0.99 0.50

TODO

References

Hvitfeldt, Emil. 2024. “Emoji: Data and Function to Work with Emojis.” https://doi.org/10.32614/CRAN.package.emoji.