A validity and reliability study of the artificial intelligence attitude scale (AIAS-4) and its relationship with social media addiction and eating behaviors in Turkish adults

In recent years, there has been a rapid increase in the use of the internet and social media. Billions of people worldwide use social media and spend an average of 2.2 h a day on these platforms. At the same time, artificial intelligence (AI) applications have become widespread in many fields, such as health, education, and finance. While AI has the potential to monitor eating behaviors and provide personalized health support, excessive use of social media and AI can lead to negative effects. These include addiction and reduced quality of life. It is important to examine the attitude toward AI and its relationship with social media addiction, eating behavior, and life satisfaction. Research on the connection between AI attitudes and eating habits is lacking, which emphasizes the necessity of validating AIAS-4 in Turkish in order to ensure its efficacy in this context. The first stage of the study aimed to adapt Grassini’s (2023) Artificial Intelligence Attitude Scale (AIAS-4) into Turkish and assess its validity and reliability. In the second stage, it was aimed to examine the relationship between artificial intelligence attitude and social media addiction, eating behavior, and life satisfaction. This study cross-sectional and methodological study was conducted in two stages in Türkiye. 172 adult individuals underwent a validity and reliability study in the first stage (43% of them were men and 57% were women), which involved adapting the AIAS-4 into Turkish. In the second stage, the relationships between artificial intelligence attitude, social media addiction, eating behavior, and life satisfaction of 510 individuals were evaluated with an average age of 24.88 ± 7.05 years (30.8% male, 69.2% female). Using the snowball sampling technique, the survey was carried out on adults by reaching out to staff and their families from both universities (Gazi University and Tokat Gaziosmanpaşa University) as well as students and their relatives. A face-to-face survey approach (delivered by an interviewer) was used for the study. In this study, the Social Media Addiction Scale-Adult Form(SMAS-AF) was used to assess social media addiction, the Scale of Effects of Social Media on Eating Behavior (SESMEB) was used to measure the impact of social media on eating behavior, the Contentment with Life Assessment Scale was used to evaluate life satisfaction, and the Eating Disorder Examination Questionnaire (EDE-Q total) was used to assess eating disorder symptoms. Pearson Correlation and Spearman Correlation according to normality and Linear regression analysis were used to analyse variables. AIAS-4 was a valid and reliable instrument in this study conducted in Türkiye (Cronbach’s alpha = 0.90 and McDonald’s omega = 0.89). Individuals spend an average of 3.7 ± 1.99 h per day on social media. All participants used WhatsApp, while 89.8% used Instagram. A negative correlation was found between AIAS and EDE-Q total, (r=-0.119 p < 0.05). BMI correlated positively with EDE-Q total (r = 0.391, p < 0.01). Higher AIAS scores were associated with increased time spent on social media (r = 0.129, p < 0.001). Conversely, higher AIAS scores were associated with lower EDE-Q total scores (r= -0.119, p < 0.001). SESMEB correlated positively with EDE-Q total (r = 0.169; p < 0.001). The model showed that BMI (β = 0.311; p < 0.001), AIAS (β =-0.157, p = 0.005), SMAS-AF (β = 0.036; p = 0.002) and SESMEB (β = 0.022; p = 0.038) affected EDE-Q total (p < 0.001 R2 = 0.198). This study revealed that the Artificial Intelligence Attitude Scale (AIAS) is valid and reliable for Turkish adults. The results show that BMI, social media addiction have positive, and AI attitude has negative impact on eating behaviors. These findings emphasize the importance of multidisciplinary approaches and awareness programs in the prevention and management of eating disorders.