The study involved a total of 464 university students (185 men, 40.09%; 277 women, 59.91%; 2 non-binary, 0.2%) recruited from multiple Italian universities. Participants ranged in age from 18 to 60 years (M = 24.97, SD = 7.36). Recruitment was carried out through various online channels, including social media platforms, university mailing lists, and community forums, using announcements that described the study’s aims and procedures. The distribution of students across academic disciplines is presented in Table 1.

A priori power analysis was not conducted; however, given reviewer feedback, we conducted a sensitivity analysis using GPower 3.1* (Faul et al., 2009) to estimate the smallest effect size that could be reliably detected with the available sample size. For bivariate correlations, with α = .05 and power (1−β) = .80, the full sample of N = 464 allows detection of small-to-moderate effect sizes (r ≥ .13). For regression or path models, this sample provides sufficient power to detect small-to-moderate effects (f² ≥ .03) with up to 10 predictors. Thus, the study is adequately powered to detect effects of practical relevance, in line with those commonly observed in psychological research. A subsample of 80 participants completed the computerized cognitive tasks. A sensitivity analysis for this subset indicates that, with α = .05 and power = .80, the minimum detectable effect is r = .31 for correlational analyses (two-tailed), corresponding to a medium effect size (Cohen, 1988). Therefore, findings based on this smaller sample should be interpreted with caution, particularly where effect sizes are below this threshold.

In the present study, questionnaires were administered to assess personality traits, need for achievement, inattention, procrastination, and social media addiction.

The study was reviewed and approved by the Ethics Committee of the University of Messina and was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. Participation was entirely voluntary, and no compensation or incentives were offered. All participants received a detailed explanation of the study’s objectives, expected duration, and procedures, and provided written informed consent prior to participation. The study was conducted in two distinct phases. In the first phase, participants completed a battery of self-report questionnaires administered online via the Google Forms platform.

These questionnaires assessed personality traits, need for achievement, inattention, procrastination, and social media addiction. All participants completed the Pure Procrastination Scale (PPS; Steel, 2010) to measure their tendency to procrastinate. Based on PPS scores, a subsample of 80 students was selected and divided into two groups of 40 participants each: high procrastinators (HP) and low procrastinators (LP). Group assignment was based on scores falling within the upper and lower quartiles of the PPS distribution, in order to maximize the contrast between the two experimental conditions. In the second phase, participants from both groups took part in individual in-person experimental sessions conducted in a controlled laboratory environment. Prior to the cognitive testing, standardized instructions were provided to ensure proper understanding of the tasks. The tasks, administered via the online platform “Cognitive Fun”, included assessments of attention, inhibitory control, and working memory. Each session was conducted individually and lasted approximately 30 minutes.

Based on PPS scores, a sample of 80 students was selected and divided into two groups of 40 participants each: high procrastinators (HP) and low procrastinators (LP). Both groups completed a battery of cognitive tasks aimed at assessing attentional capacity, inhibitory control, and working memory. The order of task administration was randomized to control for sequence effects. Attention was examined with the Go/No-Go test by calculating the visual reaction time needed to provide the correct response and to suppress an incorrect impulsive action to 10 visual stimuli. The Flanker test is an interference task where different inputs compete with the target, thus slowing down response speed. This is a basic variant using arrows, in which subjects must identify the direction of the central arrow. The estimated completion time was less than 1 minute per session. Arrows were displayed in congruent or incongruent forms. For example, in one case, the central arrow pointed to the right, so participants had to press the right arrow key on the keyboard. In another case, the central arrow pointed to the left, so subjects had to press the left arrow key. The task consisted of 20 trials, which randomly included both congruent and incongruent forms.

Finally, the N-back task was used to assess working memory, originally introduced by Kirchner (1958) as a visuospatial task with four levels (“0-back” to “3-back”) and by Mackworth (1959) as a visual letter task with up to six levels of load. In the present study, 2-back tasks were continuous recognition measures presenting sequences of stimuli (20 images) for each item in the sequence. Subjects judged whether the current stimulus matched the one presented two trials earlier by pressing a computer key. Each stimulus was presented for a maximum of 2000 milliseconds, and a new stimulus appeared every 2500 milliseconds.

All analyses were conducted using SPSS (Version 28; IBM Corp.) and AMOS (Version 28; IBM Corp.) for path analysis. Descriptive statistics, including means, standard deviations, skewness, and kurtosis, as well as Pearson correlations, were first computed to assess bivariate relationships among the study variables. To examine the statistical associations between psychological traits and procrastination, we performed multiple linear regressions, including conscientiousness, inattention, hyperactivity, need for achievement, and social media addiction as predictors. A bootstrap-based indirect-effect analysis (5,000 resamples) was conducted to test whether need for achievement statistically accounted for the association between conscientiousness and procrastination. Because the data are cross-sectional, these indirect effects were interpreted as statistical rather than causal. In addition, path analysis (structural equation modeling) was employed to examine the direct and indirect associations among the predictors, procrastination, and psychological well-being. Model fit was evaluated using the chi-square statistic (χ²), the Comparative Fit Index (CFI, with values greater than .95 indicating good fit), the Root Mean Square Error of Approximation (RMSEA, with values less than .06 indicating good fit), and the Standardized Root Mean Square Residual (SRMR, with values less than .08), following the guidelines by Hu and Bentler (1999).

Furthermore, independent samples t-tests were used to compare high versus low procrastinators on executive functioning measures, and Bonferroni correction was applied for multiple comparisons. Before conducting regression and path analysis, all relevant statistical assumptions were tested. Regarding normality, skewness and kurtosis values for all continuous variables fell within the acceptable range of ±2, suggesting an approximately normal distribution (George & Mallery, 2010). Additionally, Q–Q plots and Shapiro–Wilk tests conducted on a random 10% subsample confirmed the normality of residuals.

Linearity and homoscedasticity were assessed through scatterplots of standardized residuals versus predicted values, which showed no evidence of non-linearity or heteroscedasticity. Multicollinearity was evaluated using the Variance Inflation Factor (VIF), with all predictors showing VIF values below 2.5 and tolerance values above 0.4, indicating no issues with multicollinearity (Tabachnick & Fidell, 2007).

Finally, outliers and influential cases were assessed by examining standardized residuals greater than ±3. Cook’s distance and Mahalanobis distance were also computed, and no influential cases exceeded the critical thresholds, with Cook’s D values below 1.0 and p-values greater than .001 for Mahalanobis distance.

Descriptive statistics for the study variables are presented in Table 2. Overall, the sample showed moderate to high levels of agreeableness (M = 7.01, SD = 1.71) and openness to experience (M = 7.47, SD = 1.84), and elevated levels of procrastination (M = 33.04, SD = 12.39), inattention (M = 25.19, SD = 6.72), and hyperactivity (M = 21.43, SD = 5.14). Pearson correlations revealed significant associations among variables (Table 3). Procrastination was negatively correlated with conscientiousness (r = –.51, p < .001) and need for achievement (r = –.47, p < .001), and positively correlated with inattention (r = .75, p < .001), hyperactivity (r = .56, p < .001), and social media addiction (r = .47, p < .001). Psychological well-being showed a negative correlation with procrastination (r = –.45, p < .001).

A multiple regression analysis confirmed that inattention (β = .47, p < .001) and hyperactivity (β = .26, p < .001) were significantly positively associated with procrastination, whereas conscientiousness (β = –.38, p < .001) and need for achievement (β = –.29, p < .001) were significantly negatively associated with procrastination. A bootstrap-based indirect-effect analysis indicated that need for achievement statistically accounted for part of the association between conscientiousness and procrastination.

The indirect effect was significant (indirect effect = –0.19, 95% CI [–0.28, –0.12], p < .001), suggesting that lower need for achievement is associated with higher procrastination among individuals lower in conscientiousness. Given the cross-sectional design, this indirect effect should not be interpreted as evidence of causality.

A structural equation model was tested to examine the direct and indirect relationships among personality traits, attentional difficulties, need for achievement, procrastination, and psychological well-being. The model demonstrated a good fit to the data: χ²(7) = 12.63, p = .08; CFI = .99; RMSEA = .038 [90% CI: .00, .07]; SRMR = .019. As shown in Figure 1, inattention (β = .53, p < .001), hyperactivity (β = .28, p < .001), and social media addiction (β = .21, p < .001) showed significant positive paths to procrastination. Conversely, conscientiousness (β = –.35, p < .001) and need for achievement (β = –.32, p < .001) had significant negative effects. Procrastination, in turn, was significantly associated with lower psychological well-being (β = –.45, p < .001), consistent with the hypothesized indirect association modeled through procrastination, although no causal inference is warranted.

Openness and agreeableness were excluded from the final model due to non-significant paths.

Participants in the highest and lowest quartiles of the procrastination distribution were classified as high procrastinators (HP; n = 40) and low procrastinators (LP; n = 40), respectively. Independent samples t-tests revealed that HP participants demonstrated significantly longer reaction times and lower accuracy on executive functioning tasks. In the Go/No-Go task, the average reaction time for the HP group was 695.07 milliseconds (SD = 188.25), compared to 584.79 milliseconds (SD = 96.33) for the LP group, yielding t(78) = –3.30, p = .001, and an effect size of d = 0.69.

In the Flanker task with incongruent stimuli, HP participants had an average reaction time of 926.60 milliseconds, whereas LP participants averaged 689.83 milliseconds, with t(78) = –2.92, p = .005, and d = 0.65. For the N-back task (2-back accuracy), the HP group showed a mean accuracy of .42 (SD = .13), while the LP group achieved a higher mean accuracy of .54 (SD = .24), resulting in t(78) = 2.68, p = .009, and d = 0.60. Additionally, there was a trend toward greater response variability among HP participants, although this difference did not reach statistical significance (p = .073) (see Table 4). These findings support the view that high procrastination is associated with impairments in inhibitory control, attentional stability, and working memory updating—core components of executive functioning.

Despite its strengths, this study has several limitations. Importantly, because the present study is cross-sectional, the directionality of the associations cannot be established. Although the model is theory-driven, alternative explanations are plausible, including reciprocal or bidirectional relationships (e.g., procrastination may also exacerbate attentional difficulties and reduce well-being over time). Therefore, findings should be interpreted as statistical associations rather than causal effects. Future longitudinal and experimental studies are needed to clarify temporal precedence and causal mechanisms. Second, although the sample was large and diverse in age, it consisted solely of university students, limiting generalizability to other populations. Third, while our cognitive tasks provided objective measures of executive functioning, they were administered online, which may introduce uncontrolled environmental variability.

Furthermore, the executive tasks assessed broad cognitive constructs but did not capture all relevant components of self-regulation (e.g., planning, error monitoring). Future research could benefit from using neurophysiological measures (e.g., EEG, fNIRS) or ecological momentary assessment techniques to further triangulate findings. Additionally, examining contextual moderators such as task value, deadline proximity, or academic pressure could enrich understanding of when and why procrastination occurs.

This study offers a comprehensive view of academic procrastination by integrating personality traits, motivation, attention-related difficulties, and behavioral data into a cohesive explanatory model. By bridging subjective tendencies with cognitive performance, it highlights the multifaceted and embedded nature of procrastination as a self-regulatory failure. The indirect effects reported in this study should be understood as statistical, not causal, given the cross-sectional design. These findings nonetheless underscore the importance of combining motivational and executive-level interventions to reduce procrastination and promote academic engagement and psychological well-being.