The Effect of Choice Overload in the Carousel Design of OTT Services

Abstract

Choice overload for OTT users can be affected by the way items are presented in the user interface (UI). The difficulty of presentation format can be taxing a burden on an individual’s working memory capacity so, understanding how working memory efficiently processes with the presentation format when consumers use OTT platforms is an important factor to consider in the OTT market. Nowadays, the most popular presentation format to display contents is a swipeable carousel-based interface. Carousel is referred to as a single row in which each collection is displayed. Through this, users slide sideways or click arrows to navigate contents while searching for their desired content. In this case, the number of items that can be seen at once without turning the page is called the visible length (VL). However, the presentation format remains an underexplored UI design criteria in the field of OTT service research, even though it can significantly impact the user experience. This study aims to investigate the impact of visible length in carousel design on user’s choice overload phenomenon by using indicators of working memory and behavioural outcomes. Thirty-two healthy students (16 females, mean age 21.13 ± 2.45 years old) were recruited for the experiment, based on a sample size calculation using G*Power to ensure sufficient statistical power. In the experiment, a carousel design was used to present movie posters, and the conditions were set by adjusting the visible length (VL) to 4, 6, 8, and 12, having total of 24 set size for each condition. Each condition was conducted 30 times, resulting in a total of 120 trials. For measuring working memory, the electroencephalography (EEG) data were collected during the experiment. Frontal theta power and occipital alpha power were analysed using the Power Spectral Density (PSD) as the main neural indicators. Also, behavioural outcomes were measured through choice accuracy (ACC), reaction time (RT), and subjective evaluations as well. After the experiment, interviews were conducted to assess the difficulty of the task. Behavioural data analysis revealed that RT was shortest in VL6, indicating VL6 < VL4 (p < .01), VL6 < VL8 (p < .05), VL6 < VL12 (p < .001). On the other hand, no significant differences in ACC and subjective states were observed across the VL conditions. EEG analysis showed no significant differences in frontal theta power across the VL conditions; however, occipital alpha power showed 5 significant differences across the conditions with the lower values in VL4 compared to VL8 (p < .05) and VL12 (p < .05). In the interviews for task difficulty assessment, participants reported the VL4 condition was the most difficult to decide on. This was because the number of movies visible at once was the shortest, requiring them to turn the pages more frequently to explore the movies. Additionally, it was reported that they had to recall movies more on the previous pages when making a choice compared to other VL conditions, and as a result, they felt mentally fatigued. On the other hand, the longest visible length, which is VL12, was reported as the easiest to decide on. Taken together, the results suggests that VL4 required working memory more and led participants mentally fatigued and a subsequent decline in attention. Contrary to the idea that a higher number of options leads to choice overload, this study found that choice overload in VL4. Therefore, it is essential to set an appropriate visible length that allows for efficient content navigation to encourage easier content viewing for OTT users.