Konferencja „Miki Wesołowska. In memoriam”

Dominik Puchała, University of Warsaw
Karolina Hansen, University of Warsaw, Leibniz Institute for the German Language

Nonbinary people use various linguistic forms to express their identity. Therefore, so far, psychological research on perceptions of this minority has focused specifically on a particular language. However, in everyday contact with nonbinary individuals, people pay attention not only to their speech but also to their voice or appearance. To fill this gap, we will present the results of an online experiment (N = 242) that aimed to investigate perceptions of nonbinary people due to their language and external characteristics. To do this, we used pretested photos of faces and recordings of short speeches. Our study showed that only nonbinary people with a masculine voice were perceived worse and were less accepted than binary people. This effect did not occur for those with a feminine voice. Perceptions were mediated by the level of understanding of nonbinary people’s speech. The results of the study will be discussed in relation to their possible practical applications in developing strategies to combat discrimination against nonbinary people, particularly those with the male gender assigned at birth.

Katarzyna Jaśko, Jagiellonian University

The goal of my presentation is to examine selected psychological mechanisms involved in political activism across different social causes. Specifically, I will explore the motivational dynamics of political engagement, the role of social norms, and the perceptions of specific political actions. I will compare how these mechanisms operate in the context of LGBTQ+ activism with other types of activism, such as climate activism, women’s rights activism, and activism for refugees and migrants’ rights.

Lucía Estevan-Reina (Jagiellonian University)
Anna Potoczek (Jagiellonian University)
Dominika Gurbisz (Jagiellonian University)
Laura Schäfer (Fern University)
Oliver Christ (Fern University)
Marcin Bukowski (Jagiellonian University)

Promoting social norms based on equality and respect can be a powerful tool to address the polarization threatening our societies. This study aimed to develop an experimental norm manipulation to promote tolerance towards homosexual rights in Poland. Two manipulations were pre-tested using fictitious Facebook discussions. Across several studies (Pilot Study 1: n = 98; Study 1: n = 381; Pilot Study 2: n = 85; Study 2: n = 377), the manipulations consistently affected participants’ perceptions of descriptive and prescriptive norms in the Facebook posts and in Polish society. The moderation analyses revealed that the effect of our manipulation on tolerance appeared only among participants who endorsed fewer positive attitudes towards homosexuals (Pilot Study 1) or towards same-sex marriage (Study 2). This aligns with the conceptualization of tolerance as acceptance despite disapproval. We discuss the implications, limitations, and future directions of research on the promotion of inclusivity norms and tolerance.

Michał Bilewicz, University of Warsaw
Kamila Zochniak, University of Warsaw
Oliwia Lewicka, University of Warsaw
Zuzanna Wybrańska, University of Warsaw

Research on the consequences of derogatory language revealed that hate speech victims experience lower levels of quality of life, higher incidence of depression and posttraumatic stress disorder (PTSD), as well as more frequent suicide attempts. Most of this research used correlational studies in which participants reported their experiences of being exposed to hate speech. Such studies do not allow examination of the exact causal effects of hate speech exposure on well-being and are prone to be affected by numerous confounding variables. This talk presents two experimental studies, performed in Italy and Poland, showing that exposure to homophobic hate speech reduces well-being among LGBT+ people, particularly among those highly identified with the LGBT+ community. Study 1 showed that exposure to homophobic words increases negative mood among highly identifying LGBT+ people, but it does not influence the negative affect of loosely identifying participants. Study 2 showed that exposure to homophobic statements reduces general life satisfaction among highly identifying LGBT+ people, as opposed to low-identifiers.

Konstancja Orzechowska, University of Warsaw

First part: research Multiple studies show that there is a link between gender dysphoria and autistic traits and between transgender identity and autistic traits. A number of theories explaining the relationship between the greater prevalence of transgender identity and greater severity of gender dysphoria and autism spectrum exist. Based on current research and literature, for the purpose of the PhD project, a theoretical model in which the ability to mentalize and resistance to social norms mediate the relationship between autistic traits and gender dysphoria is proposed. A CAWI study was conducted on a research panel of 1156 adults, where autistic traits turned out to significantly predict gender dysphoria and the indirect effect through mentalizing abilities was stronger compared to that through resistance to social norms. The results also suggested that gender dysphoria and autistic traits were correlated with both depression and anxiety symptoms. Exploring such a topic is important because people with transgender identity and people experiencing gender dysphoria, as well as those with autism spectrum, are at a higher risk of depression and anxiety disorders as compared to the general population. This is due to belonging to one of the marginalized populations (those with gender dysphoria/transgender identity and those with autism spectrum). Furthermore, when an individual belongs to both marginalized communities at the same time, the risk increases even more.

Second part: mental health guidelines for the co-occurrence of autism spectrum and gender dysphoria in adolescents The World Professional Association for Transgender Health published clinical guidelines that support transgender and gender-diverse people, but they do not include concrete guidelines for working with neurodivergent individuals. However, a study was conducted by Strang et al. (2016), developing initial clinical guidelines for co-occurring autism spectrum and GD or incongruence in adolescents. These guidelines include several areas: Assessment, Treatment, Social Areas, Medical Treatment, Medical Safety, Risk of Victimization/Safety, Young Adulthood, School and Employment, and Romantic Relationships. Expanding existing research is very important due to higher risk of mental health problems in individuals with gender dysphoria and autism spectrum. Increasing knowledge may potentially explain the relationship between transgender identity and autism spectrum which may be beneficial for planning psychological support.

Małgorzata Mróz

The experience of grief is universal. Almost every person experiences the loss of a loved one at some point in their life. Depending on the research, from 7 to 20% of people experience complicated grief. LGBT+ people are at much greater risk of complicated grief and post-traumatic stress disorder related to grief. This is due to both a higher percentage of traumatic losses, i.e. losses resulting from suicide or acts of violence, as well as much lower social acceptance for public mourning and low availability of professional support. In my speech, I will discuss the specificity of queer grief in the Polish legal and social context, key risk factors for complicated grief and possible solutions to the problem of insufficient support for LGBT+ people in grief.

Miki Wesołowska, University of Warsaw
Karolina Hansen, University of Warsaw, Leibniz Institute for German Language
Paulina Górska, University of Warsaw

The usage of Natural Language Processing (NLP) in social psychology has seen unprecedented development in recent years. One area that benefited from this advancement is collective action research. Through the NLP studies of online debates, researchers managed to demonstrate that protest behaviours have their linguistic markers and may be quite precisely predicted based on the content of social media posts. What does not seem to be fully exploited, however, is the NLP’s potential to test and extend collective action theories as well as overcome the limitations of conventional collective action research (e.g., low ecological validity). In our talk, we will present the results of a study analyzing the language used in 376,864 Facebook posts published during the 2020-2021 wave of protests against the abortion ban. By relating textual data from social media to data concerning actual protest events, we will try to establish the directionality of the association between language and protest behaviour.

Paulina Górska, University of Warsaw
Tomasz Oleksy, University of Warsaw
Anna Wnuk, University of Warsaw
Maciej Górski, University of Warsaw
Miki Wesołowska, University of Warsaw
Danna Galván Hernández, University of Granada

Collective action research suffers from a range of methodological issues. Two of those are the limited possibility of conducting ecologically valid experiments and the scarcity of longitudinal data collected during actual, particularly violent, protest events. We propose that transferring collective action research to virtual reality (VR) addresses both of these issues. In this talk, we will discuss the advantages and disadvantages of using VR to study protest behaviour. Next, we will present the results of a pilot study (N = 148) in which study participants attended a collective action event (i.e., the Equality March in Opole) through the means of a 360-degree video. As suggested by the analyses, attending a virtual protest may be particularly empowering for individuals showing low levels of life satisfaction before the event.

Paulina Górska, University of Warsaw
Maciej Górski, University of Warsaw
Maria Mirucka, University of Warsaw
Miki Wesołowska, University of Warsaw

This study (N = 477) explored whether changes in group-based emotions over time can explain the gradual decline in third-party members’ collective actions supporting a disadvantaged outgroup. We categorized group-based emotions into three types: those directed at the bystander ingroup (e.g., pride and guilt), those aimed at the disadvantaged outgroup (e.g., empathy and anxiety), and those associated with an advantaged outgroup (e.g., anger and fear). The study focused on the collective actions of Poles (a not directly engaged third group) in support of Ukrainian refugees (a disadvantaged outgroup) fleeing from territories attacked by Russia (an advantaged outgroup) in February 2022. Data were collected in four waves between March 2022 and February 2023. The findings indicated that all group-based emotions, except guilt, diminished over time. Poles’ willingness to support Ukrainian refugees also declined. Importantly, the decrease in collective action was attributed to reduced empathy towards the outgroup and diminished pride toward the ingroup, rather than changes in other group-based emotions. These findings imply that the involvement of third-party members in supporting a disadvantaged outgroup is influenced by both outgroup- and ingroup-focused motivations.

Claire Wonjeong Jo, University of California, Davis
Miki Wesołowska, University of Warsaw
Magdalena Wojcieszak, University of California, Davis

This research investigates the performance of Large Language Models (LLMs), specifically OpenAI’s GPT-4-Vision, and human evaluators in identifying ‘harmful’ multimodal content in a dataset of 20,000 YouTube videos. We develop a comprehensive taxonomy for harmful YouTube content, encompassing six categories: information harms, addictive harms, hate and harassment harms, clickbait harms, sexual harms, and physical harms. The study compares the classification outputs from GPT-4-Vision, crowdworkers (MTurkers), and domain experts, highlighting their unique strengths and limitations in the subjective and context-dependent nature of harmful content. We expect to offer two significant contributions. First, by providing a systematically scoped harm taxonomy, it can assist future industry or academic efforts in identifying harmful content with specific categories, eliminating the need to rely on arbitrary or inconsistent definitions. Second, we propose an alternative to human annotators that is more cost-effective and efficient, thereby protecting human evaluators from exposure to harmful content.