Introduction
Advances in science and technology, combined with globalization, have drastically changed the way humans conduct economic activities, giving consumers a wide range of options (Lee & Lee, 2004; Lurie, 2004) even when making simple decisions such as what to eat for breakfast. While it may not be consequential to make the wrong choice for a daily decision, it is a different story when it comes to lifetime decisions such as selecting a healthcare plan or making a retirement investment, where one has to make the best choice from dozens or even hundreds of options (Gourville & Soman, 2005). It is thus essential that individuals, as well as corporations, governments, and state agencies, make informed decisions among the numerous options available to them, as the cost of making the wrong decision or delaying it can be quite significant (Sagi & Friedland, 2007).
Theories in economics, marketing, and psychology suggest that having more choices increases the sense of freedom and self-determination (Warren & Lupinek, 2019), and is more attractive to merchants as it can satisfy a wider range of consumer preferences (Buturak & Evren, 2017). However, large choice sets can lead to an increase in the cost of decision-making, such as the fear of mistakenly passing up the ideal choice, making trade-offs, and being hesitant in decision-making (Diehl & Poynor, 2010; Lee et al., 2021; Reed et al., 2011). This can result in decreased satisfaction with choices, alteration of the initial choice, or even postponement of the choice, all of which are unfavorable outcomes (Diehl & Poynor, 2010; Wan et al., 2009). Evidence of this phenomenon (i.e., choice overload) (D’Angelo & Toma, 2017; Iyengar & Lepper, 2000).is seen in the financial investment field, where fewer types of pension plans are offered and more employees are willing to join pension plans (Soroya et al., 2021).
Research has attempted to identify the optimal number of options for consumers, and it has been found that a choice set between 8-15 options is preferred (Chernev et al., 2015; Sharma & Nair, 2017). Large choice sets may initially be appealing, however, if cognitive load is present, the motivation to make a selection is reduced (Basili & Vannucci, 2015). Studies conducted in the areas of charitable giving and travel souvenirs indicated that choice overload did not occur (Lindkvist & Luke, 2022; Sthapit, 2018). A meta-analysis revealed that the effects of choice overload are moderated by boundary conditions (Scheibehenne et al., 2009; Spassova & Isen, 2013) such as the difficulty of the decision task (e.g. time pressure; (Basso et al., 2019)), the complexity of the choice set (e.g. option attributes and dominant options; (Townsend & Kahn, 2014; Wan et al., 2009)), preference uncertainty (e.g. expertise; (Hadar & Sood, 2014)), individual differences (e.g. trait anxiety, (Hu et al., 2023)), and the decision goal (e.g. decision focus; (Wang & Shukla, 2013)).
The cognitive load theory (Sweller, 1988) argued that the decision-making process required a substantial amount of cognitive resources. When the amount of cognitive resources required to make a decision exceed the individual’s available resources, they experience cognitive overload. Compared to a small choice set, individuals must process more information when presented with a larger choice set, which can exceed their cognitive capacity (McShane & Bockenholt, 2018). This can lead to decreased satisfaction, increased negative experiences, and even delays or abandonment of choices (Gerasimou, 2018; Hills et al., 2013). Previous researchers have primarily evaluated cognitive load through self-report methods, such as subjective evaluations of task difficulty and negative emotions (Gerasimou & Papi, 2018; Li, 2017; Saltsman et al., 2019; Song et al., 2019; Sthapit et al., 2017; Turri & Watson, 2022). Current explanatory models of choice overload have been largely derived from behavioral experiments or self-reports, with only one study combining eye-tracking and functional magnetic resonance imaging to investigate choice overload. This study found that the cost of choosing from a large choice set can be reflected in the activity of task-relevant areas of visual and sensorimotor processing, saccade frequency, and average saccade amplitudes (Reutskaja et al., 2018). Additionally, some researchers have used electroencephalogram (EEG) to explore the effects of information overload on decision-making (Kim et al., 2018). It has been found that information overload may impair the decision-making process, as evidenced by a decrease in P2 and P3 amplitudes, as well as an increase in LPC. However, it is important to note that choice overload is distinct from information overload, and it is possible for consumers to experience choice overload without experiencing information overload (Kuan et al., 2014; Wan et al., 2009).
This study utilizes EEG to investigate the neural foundations of choice overload and to evaluate the cognitive load theory. As per the Cognitive Load Theory, larger choice sets necessitate more attentional resources, resulting in the choice overload effect. However, it is still unclear when and how choice overload affects attentional engagement. Based on prior research, the early component P1 reflects attention-based early visual perceptual processing and allocation of attention to stimuli (Grand et al., 2004; Mangun, 1995; Mangun & Buck, 1998). The magnitude of P2 is related to the allocation of attentional resources and the ability to sustain perceptual processing (Ferreira-Santos et al., 2012; Huang & Luo, 2006; Mercado et al., 2006; Schupp et al., 2003). P3 is a positive component related to the allocation of attentional resources allocation (Folstein et al., 2008; Polich, 1987), and is also associated with task difficulty and decision confidence (Nieuwenhuis et al., 2005; Qin & Han, 2009). LPC is a widely used physiological measure that reflects emotional responses and the allocation of attentional resources during cognitive processes (Alomari et al., 2015; Cuthbert et al., 2000; Ferrari et al., 2011; Hajcak & Foti, 2020). In summary, P1 and P2 were selected as markers of attention during the initial processing stage, while P3 was used to measure attention during the intermediate processing stage. Additionally, LPC was employed as an indicator of attention during the final processing stage.
We additionally conducted multivariate pattern analysis (MVPA) to capture the differences in processing patterns between different choice set sizes. This method has become increasingly popular in cognitive neuroscience due to its increased sensitivity compared to traditional univariate analysis (Carlson et al., 2019; Meng et al., 2023). By combining MVPA analysis with event-related potential (ERP) analysis, we can gain a more comprehensive understanding of the neural mechanisms of choice overload and develop a more accurate model for it.