Data from the UC1 Experiments on the perception of deepfake videos and their effects on attitude

The data generated during UC1experiments that examined whether deepfakes can influence attitudes toward climate change and immigration, two highly polarized and politically relevant issues, and whether these effects depend on video quality, perceived trustworthiness (as measured with our previously developed Perceived Deepfake Trustworthiness Questionnaire). The dataset contributes to a more comprehensive understanding of how artifically generated content affects cognition, emotion, behavioral intentions, and attitudes. It also addresses both the potential risks and possible constructive uses of deepfakes, highlighthing the conditions under which the influence of artifically generated content is most pronounced, and identify individual differences that shape people’s vulnerability or resilience to manipulated media. This knowledge is critical for developing evidence-based approaches to mitigate the risks associated with malicious deepfakes while exploring their potential for positive applications. Relevant References: Plohl, N., Mlakar, I., Aquilino, L., Bisconti, P., & Smrke, U. (2024). Development and Validation of the perceived deepfake trustworthiness questionnaire (PDTQ) in three languages. International Journal of Human–Computer Interaction, 41(11), 6786 –6803. https://doi.org/10.1080/10447318.2024.2384821 Plohl, N., Mlakar, I., Aquilino, L., Brienza, M., Bisconti, P., & Smrke, U. (2025a). The moderating role of perceived trustworthiness in explaining the attitudinal effects of political deepfakes. Social Science Research Network. https://doi.org/10.2139/ssrn.5351533 Plohl, N., Mlakar, I., Aquilino, L., Brienza, M., Bisconti, P., & Smrke, U. (2025b). How deepfake quality, media literacy, and personal attitudes shape detection, liking, and social media sharing of political deepfakes. PsyArXiv. https://doi.org/10.31234/osf.io/knvby_v1本数据集为UC1实验过程中生成的实验数据，该实验旨在探究深度伪造（deepfakes）是否会影响人们对气候变化与移民这两个高度极化且具有政治相关性议题的态度，同时验证此类影响是否取决于视频质量与感知可信度——其中感知可信度通过我们此前开发的感知深度伪造可信度问卷（Perceived Deepfake Trustworthiness Questionnaire）进行测量。 本数据集有助于更全面地理解人工智能生成内容如何作用于个体的认知、情绪、行为意向与态度。同时，该数据集既探讨了深度伪造的潜在风险，也分析了其可能的建设性应用场景，明确了人工智能生成内容影响效应最为显著的情境条件，并识别出能够影响个体对操控性媒体的易感性与抗御韧性的个体差异。 此类研究成果对于制定循证策略以缓解恶意深度伪造带来的风险、同时探索其正向应用潜力而言至关重要。 相关参考文献： Plohl, N., Mlakar, I., Aquilino, L., Bisconti, P., & Smrke, U. (2024). 三种语言版本下感知深度伪造可信度问卷（PDTQ）的编制与验证. 《国际人机交互期刊》, 41(11), 6786–6803. https://doi.org/10.1080/10447318.2024.2384821 Plohl, N., Mlakar, I., Aquilino, L., Brienza, M., Bisconti, P., & Smrke, U. (2025a). 感知可信度在解释政治类深度伪造态度效应中的调节作用. 社会科学研究网. https://doi.org/10.2139/ssrn.5351533 Plohl, N., Mlakar, I., Aquilino, L., Brienza, M., Bisconti, P., & Smrke, U. (2025b). 深度伪造质量、媒体素养与个人态度如何影响政治类深度伪造的检测、喜爱度与社交媒体分享行为. PsyArXiv. https://doi.org/10.31234/osf.io/knvby_v1