Difficulty: Intermediate | Prerequisites: Part 1 of these study notes (Theory and Method).
Sternberg et al. (2021) tested whether deliberately thinking about neutral content (attentional distraction) while viewing tempting Facebook images could reduce the desire to use social media. It did: distraction lowered self-reported desire, reduced attentional engagement (LPP), and reduced the amount of social-media content held in working memory (CDA). This part covers the statistical results, what they mean, and how to prepare for exam questions on this study.
All three dependent measures confirmed the predictions. Each was analysed with a paired-samples t-test comparing the temptation and attentional-distraction conditions.
Distraction condition: M = 2.02, SD = 0.63
Temptation condition: M = 2.72, SD = 0.78
t(28) = 8.19, p < .001, 95% CI for the mean difference = [0.52, 0.87]
Cohen's d = 1.5 (a very large effect)
100% of participants (29/29) showed lower desire in the distraction condition
This is the behavioural headline: distraction reliably and substantially reduced the self-reported urge to use Facebook.
The LPP was measured across the full window (300 to 2,500 ms) and an early window (300 to 1,000 ms).
Full LPP window (300 to 2,500 ms):
Distraction condition: M = -1.72 µV, SD = 2.66
Temptation condition: M = -1.06 µV, SD = 2.89
t(28) = 2.54, p = .02, 95% CI = [0.13, 1.18], Cohen's d = 0.4
72% of participants (21/29) showed the pattern
Early LPP window (300 to 1,000 ms):
Distraction condition: M = -2.43 µV, SD = 2.20
Temptation condition: M = -1.61 µV, SD = 2.79
t(28) = 3.47, p = .002, Cohen's d = 0.32
The early LPP finding is important because it suggests that attentional disengagement from social-media stimuli begins early in processing, not just after extended viewing.
Distraction condition: M = -0.33 µV, SD = 1.03
Temptation condition: M = -1.03 µV, SD = 0.75
t(29) = 3.55, p = .001, 95% CI = [0.30, 1.11], Cohen's d = 0.6
79.3% of participants (23/29) showed the pattern
The CDA result is the mechanistic centrepiece. Less negative CDA in the distraction condition means less social-media content was being actively maintained in working memory. This supports the theoretical account that attentional distraction works by restricting the WM representation of temptations.
Note the direction: CDA is a negative component, so "less negative" (closer to zero) = less WM representation. Graphs in the paper plot CDA with the y-axis reversed (negative is up) to make this visually intuitive.
The study's three main contributions:
First direct causal evidence that social-media temptations can be self-regulated. Prior work treated social media as a tool for regulating other emotions, not as the target of regulation itself.
Neural mechanism identified. CDA modulation shows that distraction works by restricting the WM representation of tempting content. This supports Hofmann and Van Dillen's (2012, 2018) theoretical accounts that emphasise WM as the arena where self-regulation succeeds or fails.
LPP modulation confirms reduced attentional engagement, consistent with prior distraction studies on food cravings, smoking cues, and sexual desire.
The authors also highlight a psychological advantage of self-regulation over automated digital tools (app blockers, screen-time limits). Successful self-regulation produces feelings of agency and internal control (Tangney et al., 2004), which can reinforce future self-control. Digital tools, by contrast, attribute success to an external force, which may not build the same internal capacity.
The study extends findings from Sternberg et al. (2018), which showed that low specific (but not general) ability to control WM representation of Facebook stimuli was associated with maladaptive Facebook use. Together, these studies build the case that the ability to regulate social-media content specifically, rather than general self-regulation ability, is what matters for problematic usage.
Long-term costs of distraction. Attentional distraction prevents the temptation from being fully processed and represented in WM. This blocks habituation, meaning the temptation may remain just as potent the next time it arises. Future work should combine distraction (strong short-term efficacy) with cognitive reappraisal (which allows meaning-making and offers longer-term benefits).
Two distraction subtypes were combined. Participants could think about geometric shapes or daily routine activities. Both involve disengaging attention from the stimulus, but they may have different downstream effects. The study does not tease apart which type of neutral thought is more effective.
Facebook-only stimuli. The study used Facebook images exclusively. Results may or may not generalise to Instagram, TikTok, Snapchat, or non-social-media temptations like food. Given that distraction has worked across multiple temptation domains in prior research, the authors expect it would generalise, but this remains untested for social-media platforms other than Facebook.
Individual differences not examined. Facebook stimuli were probably not equally tempting for all 29 participants. Future studies should look at how personal temptation level, interference with other goals, and problematic usage patterns moderate the regulation effect.
Not preregistered. The design and analysis plans were not preregistered, though the data are openly available on OSF (https://osf.io/f9sh3).
This study provides the first direct causal evidence that social-media temptations (not just emotions regulated via social media) can be self-regulated. Expect this distinction to be tested.
Know the difference between CDA and LPP: CDA = WM representation, LPP = attentional allocation. Both were reduced by distraction, but they index different cognitive processes at different neural sites.
The largest effect size was for self-reported desire (d = 1.5), not for the neural measures (d = 0.4 for LPP, d = 0.6 for CDA). This is a common exam question: which measure showed the strongest effect?
The within-participants design is important: the same participants, viewing the same stimuli, produced different outcomes depending on the instruction. This rules out individual-difference and stimulus confounds.
Be ready to discuss limitations, especially the long-term cost of distraction (no habituation) and the Facebook-only stimulus set.
The deprivation procedure (changing the Facebook password) is a methodological detail worth knowing. It increased stimulus saliency but did not bias natural usage patterns.
True or False: Attentional distraction involves physically looking away from tempting stimuli.
False. It is a top-down mental strategy; gaze remains fixed.
Fill in the blank: The CDA component indexes the degree to which information is actively represented in ______.
Visual working memory.
True or False: 100% of participants showed reduced desire in the distraction condition compared to temptation.
True.
Fill in the blank: The study's within-participants design means that each participant served as their own ______.
Control.
True or False: Attentional distraction is recommended by the authors as the ideal long-term self-regulation strategy for social-media temptation.
False. They note it has long-term costs (no habituation) and suggest combining it with cognitive reappraisal.
Q: What is the main theoretical premise tested in Sternberg et al. (2021)?
A: That successful self-regulation of social-media temptations requires limiting the mental representation of those temptations in working memory. Attentional distraction achieves this by loading WM with neutral content, thereby crowding out the tempting representation.
Q: Describe the two experimental conditions and explain how they differed.
A: In the temptation condition, participants passively watched Facebook images and allowed social-media-related thoughts and desires to arise naturally. In the attentional-distraction condition, participants viewed the same images but deliberately redirected their thoughts toward neutral content (geometric shapes or daily routine activities). The stimuli were identical across conditions; only the cognitive instruction differed.
Q: What does the CDA measure, and what did the CDA results show in this study?
A: The CDA (contralateral delay activity) is a negative ERP component that indexes how much information is actively maintained in visual working memory. The study found less negative CDA amplitudes in the distraction condition (M = -0.33 µV) compared to the temptation condition (M = -1.03 µV), with Cohen's d = 0.6, indicating that distraction reduced the WM representation of social-media content.
Q: Why did the researchers deprive participants of Facebook access before the experiment, and does this deprivation bias the results?
A: Deprivation (changing the Facebook password for 48 hours) increased the saliency and motivational value of the Facebook stimuli, making the temptation manipulation more potent. Prior research (Sternberg et al., 2018) demonstrated that this procedure does not distort natural usage patterns: deprived lab usage time positively correlated with normal at-home usage time.
Q: Name one long-term cost of attentional distraction as a self-regulation strategy and suggest what the authors recommend to address it.
A: Attentional distraction prevents full processing of the temptation in working memory, which blocks habituation. Without habituation, the temptation remains equally potent each time it arises. The authors recommend combining distraction (for short-term efficacy) with cognitive reappraisal (which allows meaning-making and supports longer-term regulation).
Q: The study found that 100% of participants showed reduced desire in the distraction condition, but only 72% showed reduced LPP. What might account for this difference?
A: Self-reported desire reflects a conscious, aggregate evaluation of one's motivational state, which may be more uniformly sensitive to the distraction manipulation. The LPP is a neural measure of initial attentional allocation, which can be noisier at the individual level due to EEG signal variability and individual differences in early attentional processing. The smaller effect size for LPP (d = 0.4 vs d = 1.5 for desire) also means that smaller individual differences can push some participants against the group trend.
This study connects to broader self-regulation and emotion-regulation literature (Sheppes, 2020). If your course covers cognitive reappraisal, compare it with attentional distraction: reappraisal changes the meaning of a stimulus (effective long-term, but harder with high-intensity temptations), while distraction blocks the stimulus from entering WM in the first place (effective short-term, but prevents habituation).
The working-memory framework here links directly to Baddeley's model and to dual-process accounts of self-control (Barrett et al., 2004). If your course covers executive function or cognitive load, the CDA findings provide a concrete example of how loading WM with neutral content reduces the processing of competing information.
The study also relates to the psychology of addiction and compulsive behaviour. The Facebook deprivation procedure mirrors substance-deprivation paradigms (e.g. Grimm et al., 2001 on incubation of cocaine craving), and the reward-pathway activation triggered by Facebook stimuli (Turel et al., 2014) draws parallels with substance-use neuroscience.
self-regulation, social media addiction, Facebook temptation, attentional distraction, cognitive distraction, working memory, WM, contralateral delay activity, CDA, late positive potential, LPP, event-related potential, ERP, EEG, desire regulation, temptation regulation, emotion regulation, cognitive reappraisal, dual-process theory, executive function, cognitive load, digital self-control, screen time, smartphone addiction, social media and driving, social media and sleep, Sternberg Luria Sheppes 2021, Hofmann Van Dillen, self-control, impulse control, reward pathways, amygdala-striatum, Facebook deprivation, within-participants design, Psychology II, FIU