Attentional bias is commonly observed in individuals with addiction, often manifested in heightened sensitivity to cues related to substance use (Field et al., 2014). This heightened sensitivity contributes significantly to a pronounced inclination toward engaging in drug-seeking behaviors (Berridge and Robinson, 2016). The Incentive Sensitization Theory (IST) provides a comprehensive framework elucidating how attentional mechanisms play a pivotal role in sustaining and triggering relapses in addictive behavior (Robinson and Berridge, 2000).

According to IST, distinct neural mechanisms underlie the motivational (“wanting”) and hedonic (“liking”) effects of substance use. The two systems exhibit distinct reactions to repeated substance administration, with tolerance affecting the “liking” (resulting in reduced responsiveness) and sensitization influencing the “wanting” system (leading to increased responsiveness). Consequently, individuals with substance use disorders may experience an increased desire for the substance, even in the absence of an expectation of pleasure from its consumption (Robinson and Berridge, 2001). IST posits that the motivational and hedonic systems operate independently of conscious cognition, and the conscious perception of liking and wanting emerges through the interaction of brain systems distinct from those associated with “wanting” and “liking” (Robinson and Berridge, 2003). This separation is evident from the study of Winkielman et al. (2005), in which the impact of subliminally presented happy versus angry faces on motivation and affect was investigated, revealing that basic affective reactions operated unconsciously and interacted with incentive motivation. Crucially, IST asserts that in typical situations, the conscious experience of liking and wanting arises from the interplay between the distinct brain systems associated with conscious awareness and unconscious affect (Berridge and Winkielman, 2003), thus it does not rule out the possibility of conscious experience—at least in part—after the expression of “wanting” and “liking” (Anselme and Robinson, 2016).

The psychological process underlying “wanting” involves assigning salience to stimuli and their representations. This process is thought to modify both neural and psychological representations of ordinary stimuli, rendering them highly salient and attention-grabbing (Robinson and Berridge, 2000). Numerous studies have provided evidence of attentional sensitization, as demonstrated in various scenarios. For instance, studies have shown that individuals using marijuana (Field, 2005), alcohol (Townshend and Duka, 2001), tobacco (Field et al., 2009), cocaine (Liu et al., 2011), or engaging in internet use (Nikolaidou et al., 2019) exhibit facilitated information processing for visual stimuli associated with their respective addictions. These studies consistently reveal that addicts, when compared to control subjects, exhibit quicker detection of changes, faster response times, and reduced ability to inhibit distracting visual stimuli when the stimuli are related to their specific addiction.

Given that “wanting” and “liking” are believed to be influenced by subconscious mechanisms, it is crucial to explore the feasibility of assessing them through self-report methods. However, this presents a significant challenge, as obtaining a direct measure of both explicit and implicit aspects of “wanting” and “liking” proves to be difficult, due to limitations in the original research paradigm used in animal studies for investigating these processes in human subjects (see Pool et al., 2016). It is important to emphasize that while the IST suggests that the motivational and hedonic systems operate independently of conscious cognition, it does not rule out the possibility of conscious experience after their expression. Instead, cognitive self-control often contends with “wanting” implying that it may be influenced, at least in part, by conscious cognitive processes (Anselme and Robinson, 2016).

Despite the methodological concerns, a few survey methods were developed, such as the Sensitivity to Reinforcement of Addictive and other Primary Rewards (STRAPR, Goldstein et al., 2010), the Desires for Alcohol (DAQ) (Love et al., 1998) and Speed Questionnaire (DSQ, James et al., 2004; Willner et al., 2005), and the Imaginative Wanting and Liking Questionnaire (File et al., 2022). While these survey methods have yielded findings that are partially consistent with the predictions of IST, their relationship to implicit measures is not clear. The question of the accessibility of “wanting” and liking to the conscious mind holds a dual significance. Firstly, it is of theoretical importance as it sheds light on the degree to which these psychological processes can be consciously experienced. This understanding is crucial for developing therapeutic and intervention approaches aimed at addressing addictive behaviors or other related conditions. Secondly, from a research methodology standpoint, exploring explicit measures (self-reports, conscious evaluations) that are related to behavioral measures justifies the use of large-scale survey approaches, which allows the comprehensive investigation of the associations between wanting and liking and other psychological properties.

The aim of the current study was to investigate the relationship between smokers’ attentional bias toward smoking-related stimuli and self-reported imbalance of the motivational and hedonic systems (wanting minus liking of smoking: WmL), measured with the Imaginative Wanting and Liking Questionnaire. To this end, an online experiment was designed deploying a flicker paradigm (see Rensink et al., 1997). The flicker paradigm is a frequently employed technique in studies investigating change blindness, which refers to the perceptual phenomenon where an observer fails to notice a change in a visual stimulus (Rensink et al., 1997). The flicker paradigm involves a task where individuals are asked to identify a single difference between two otherwise identical visual scenes displayed on a screen. The scenes are presented repeatedly, with a mask in between, until the change is detected (Jones et al., 2006). Interestingly, it takes a surprisingly long time to notice the change, and this process is strongly influenced by attention (Turatto et al., 2003). One reason for directing attention to the changing element is that viewers have a personal interest in the object that carries the change (Rensink et al., 1997), which makes it suitable to study attentional bias toward specific objects, such as those related to smoking, as in the current study. Three conditions were employed, with neutral and smoking-related changes presented for current-, former, and non-smoker participants. Based on File et al. (2022, 2023), a positive correlation was hypothesized between nicotine dependence and WmL (H1) and a positive correlation between WmL and the performance (number of alternations needed for change detection) of smoking-related change detection compared to neutral changes (H2). Also, a positive correlation was hypothesized between nicotine dependence and performance of smoking-related change detection compared to neutral changes (H3).

Based on previous studies (e.g., Jones et al., 2003, 2006), a replication hypothesis was formed, by which a significant difference between smokers, non-smokers, and former smokers in change detection performance was expected in smoking-related changes relative to neutral changes (smokers > former smokers > non-smokers) (H4). Lastly, a significant difference between smokers, non-smokers, and former smokers in change detection performance was expected in case of neutral changes when a smoking related object was present (smokers < former smokers < non-smokers) (H5).

Out of the 422 participants, 214 were non-smokers (mean age: 26.68, SD = 10.34, 70.1% female), 123 current smoker (mean age: 28.70, SD = 10.70, 67.5% females, average Fagerström score: 3.65, SD = 2.4), and 85 former smoker (mean age: 30.72, SD = 12.71, 72.9% females, average Fagerström score: 2.99, SD = 2.52). For an overview of the variables used in the analyses, see Table 1.

There was a weak positive correlation between nicotine dependence and WmL [H1, r_s(121) = 0.28, p = 0.002]. Also, there was a moderate positive correlation between WmL and change detection performance of smoking related changes [H2, r_s(121) = 0.33, p < 0.001]. Contrary to our expectations, the correlation between nicotine dependence and change detection performance of smoking related changes was not significant [H3, r_s(121) = 0.07, p = 0.456]. For an overview, see Table 2.

Since the smoking related attentional bias variable within the Smoker group did not follow a normal distribution, Kruskal-Wallis test was used to compare groups non-smoker, smoker and former smoker (H4). The analysis revealed a significant effect of smoking related attentional bias χ²(2, N = 422) = 6.37, p = 0.041 (Figure 2). The effect size, Omega-squared was small (ω² = 0.00431). Post-hoc tests using Dunn’s test indicated that smokers needed fewer stimulus presentations to detect smoking related changes (relative to neutral changes) (p = 0.012) than non-smokers. There were no significant differences observed between former smokers and both smokers (p = 0.205) and non-smokers (p = 0.0.406). There was no significant difference between groups in the Latency of change detection of neutral changes where smoking related object was present relative to neutral changes where no smoking related object was present (NC minus NC-S) [χ²(2, N = 422) = 0.93, p = 0.628] (H5).

The Incentive Sensitization Theory of addictions (Robinson and Berridge, 1993) offers a comprehensive neuropsychological framework to explain attentional sensitization in addictions, which is supposed to be the result of largely implicit cognitive mechanisms. However, the extent to which these underlying mechanisms are consciously accessible for reports remains limited in our current knowledge and is a subject of ongoing debate. For the best of the authors’ knowledge, no study has investigated the link between the imbalance of self-reported wanting and liking and attentional bias for addiction related stimuli. Thus, the main objective of this study was to investigate the association between self-reported wanting and liking measured with the Imaginative Wanting and Liking Questionnaire of smokers and their performance in detecting changes in smoking-related stimuli on a relatively large sample.

In line with previous research (e.g., File et al., 2022, 2023), the current study found a positive relationship between nicotine dependence and the self-reported imbalance of wanting and liking. Specifically, higher levels of nicotine dependence were associated with a greater discrepancy between wanting and liking, indicating a stronger desire for nicotine relative to the pleasurable experience derived from it.

The primary discovery of the study highlighted a positive correlation between wanting-minus-liking (WmL) and change detection performance in smoking-related stimuli among smokers, aligning with the predictions of IST (Robinson and Berridge, 1993). This correlation between the self-report measure of IST and attentional bias lends support to the credibility of self-report assessments, indicating their ability to capture not only conscious evaluations but also their connection to relevant cognitive processes. Considering that the original research paradigm is not suitable to investigate wanting and liking in humans (Tibboel et al., 2011), it becomes imperative to undertake a thorough exploration of diverse approaches. It is unquestionable that physiological and neurobiological measures are better suited for assessing IST, as evident in the methods commonly employed in the realm of human IST studies (Pool et al., 2016). However, the current results indicate that self-report methods are potentially useful too, which opens avenue for survey studies, allowing researchers to explore a more comprehensive understanding of psychological correlations of IST. In line with that, WmL demonstrated heightened sensitivity to attentional sensitization compared to the addiction score itself, suggesting that its utilization can potentially contribute to a more comprehensive understanding of addictive behaviors. As Field and Cox (2008) summarized, it is not clear that the strength of attentional bias is dependent on the severity of nicotine dependence. Previous studies reported positive correlation (e.g., Zack et al., 2001), negative correlation (Mogg et al., 2005), or no relationship (Munafò et al., 2003). The current results showed no correlation between attentional bias strength and nicotine dependence, but there was a positive correlation between attentional bias strength and WmL. A possible explanation for the inconsistent results might be that nicotine addiction is usually measured with the Fagerström Test for Nicotine Dependence (Heatherton et al., 1991), which shows floor effects when measuring populations of light smokers (Etter et al., 1999). Another potential explanation is that the attention-capturing capability of substance-related stimuli is not directly linked to dependence severity (as in Ryan, 2002 or Jones et al., 2006), but is more directly associated with the imbalance between wanting and liking. In line with that, Hobson et al. (2013) reported an association between change detection performance with levels of craving, but not with alcohol consumption. However, as both of these explanations remain speculative, additional research is required to explain the inconsistent results from previous studies.

Previous research has pointed out that attentional bias may play a role in sustaining smoking habits and contributing to relapse (Masiero et al., 2019). However, the current findings suggest that the observed attentional bias likely emerges early in the progression of problematic use, given the relatively low level of nicotine dependence in the sample, as indicated by an average Fagerström score of 3.65 (SD = 2.4), indicating low to moderate dependence. This observation is consistent with Jones and Bruce’s (2005) proposal of a graded continuity of attentional bias across the consumption continuum. This implies that altered attentional processes may not only be confined to maintaining smoking and contributing to relapse but also extend to the development of addiction by emphasizing smoking-related environmental cues that trigger the behavior.

While the majority of studies have concentrated on the association between the degree of addiction and the temporal aspects of attentional bias, such as the time needed to detect changes (Jones and Bruce, 2005) or the speed differences between congruent and incongruent conditions in Go/no-Go or Stroop tasks (Zack et al., 2001; Munafò et al., 2003), the present findings suggest that exploring alterations in the set of cues capable of triggering attentional bias could be a promising avenue for research. In the current study, only the most salient, closely tobacco-related stimuli (pack of cigarette and cigarette on ashtray) were linked to WmL, although the sensitized attentional system is assumed to be not restricted to direct substance cues, other associated cues are able to activate it (Robinson and Berridge, 1993). The restricted generalization observed in the current study may be attributed to the relatively low level of nicotine dependence of the sample. To draw more robust conclusions, future studies should consider a more diverse set of stimuli and include participants with higher levels of nicotine dependence. In the present study, we successfully replicated a widely observed group difference (e.g., Jones et al., 2003, 2006); smokers exhibited faster detection of smoking-related changes compared to non-smokers, indicating an attentional bias toward smoking-related stimuli. The practical significance of the observed bias is heightened by the fact that the study was conducted outside the controlled confines of a laboratory setting, taking place within a web browser under uncontrolled conditions. The current results only partially supported the notion that the altered sensitivity of the attentional system persists beyond the cessation of actual smoking, as proposed by Berridge et al. (2009). Former smokers displayed an intermediate level of attentional bias, falling between the levels observed in smokers and non-smokers. Notably, the attentional bias in former smokers did not significantly differ from either group.

Overall, the findings of the current study support the predictions of the Incentive Sensitization Theory (IST) and contribute to the growing literature on that attentional mechanisms play a crucial role in addictive behaviors. The results suggest that self-reported wanting and liking may be linked to the implicit cognitive mechanisms described in IST, which suggests that using self-report measures in the field of IST should not be discarded at this early phase of research. Instead, developing and testing such measures can potentially provide valuable information complementing experimental data. Moreover, the possibility that the dynamics of wanting and liking are at least partially accessible to conscious awareness opens up possibilities for treatment and interventions. For instance, attention bias modification induces enduring automatic and sustained avoidance responses to cues associated with smoking, making it a potential strategy to help smokers quit (Lopes et al., 2014). In line with that, File et al. (2023) reported that self-reported wanting showed a positive correlation with intention to quit smoking, indicating that by directing attention to the motivational and hedonic aspects of smoking, it may be feasible to enhance motivation to quit.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by United Ethical Review Committee for Research and Psychology. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

DF: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. BP: Conceptualization, Writing – original draft, Writing – review & editing. PK: Writing – original draft, Writing – review & editing. LK: Writing – original draft, Writing – review & editing. ZG: Writing – original draft, Writing – review & editing. ZD: Writing – original draft, Writing – review & editing. IC: Conceptualization, Funding acquisition, Supervision, Writing – original draft, Writing – review & editing.