Suggested time-25 minutes.
Spend the first 10 minutes reading the source, then 15 minutes writing your response.
The following is a summary of a peer-reviewed study. Use the description of the research to answer the questions that follow.
Halverson, Pradhan, and Coelho (2023) noted that prior survey work has linked smartphone use during studying to lower test scores, but most of those studies were correlational and could not separate the effect of incoming alerts from the effect of students choosing to check their phones. The present study was designed to isolate the effect of notification rate on short-term verbal memory by using a lab-issued research phone that participants could see but were not permitted to touch. The researchers predicted that increasing the rate of incoming notifications during a brief study period would reduce the number of items participants could recall on an immediate free-recall test.
A total of 192 undergraduate students enrolled in an introductory psychology course at State University of Northern Ohio took part in the study in exchange for partial course credit. Participants ranged in age from 18 to 22 years (mean age = 19.7 years; standard deviation = 1.2 years). The sample identified as 56% women, 42% men, and 2% nonbinary. Self-reported race/ethnicity was 64% White/European American, 12% Black/African American, 11% Hispanic or Latino/a/x, 8% Asian or Asian American, 3% Multiracial, and 2% preferred not to answer. All participants reported owning a personal smartphone and using it daily.
Each participant drew a sealed envelope containing a condition assignment from a shuffled stack of envelopes upon arrival at the lab. The envelope determined which of three notification conditions the participant would complete: a no-notification condition, a low-frequency notification condition (1 banner-plus-vibration alert per minute), or a high-frequency notification condition (6 banner-plus-vibration alerts per minute). Participants were informed of the study’s general procedures and signed a document indicating their voluntary agreement to participate before any data were collected.
During the 10-minute encoding phase, each participant studied a list of 30 unrelated English nouns (e.g., “lantern,” “harbor,” “velvet”) presented one at a time on a desktop monitor for 4 seconds per word, with a 2-second interstimulus interval. A lab-issued research smartphone was placed face-up on the desk, approximately 25 cm to the right of the monitor and within the participant’s peripheral vision. The research phone was loaded only with a custom messaging app that delivered scripted alerts on a fixed schedule controlled by the experimenter; participants were told they were not to pick up or interact with the phone in any way during the session. Notification content consisted of short generic messages (e.g., “New message from Jordan,” “3 new emails in your inbox”) drawn from a pool of 80 pre-written items.
A wall-mounted, screen-recording eye tracker logged the timestamp of any glance at the research phone for use as a compliance check; data from any participant whose eye-tracking record showed sustained fixation on the phone (> 2 s) was flagged for review. After the encoding phase, participants completed a 90-second arithmetic distractor task to clear active rehearsal, then were given 4 minutes to write down as many of the 30 studied words as they could remember in any order. Recall accuracy was scored as the percentage of words correctly recalled out of the 30 studied. At the end of the session participants completed a brief post-task questionnaire about how many notifications they noticed (a manipulation check), and they were then told the full hypothesis of the study and given contact information for the lab’s institutional review board.
The manipulation check confirmed that participants in the high-frequency condition reported noticing significantly more notifications (M = 47.2, SD = 9.1) than those in the low-frequency condition (M = 8.6, SD = 2.4); the eye-tracking compliance check showed that fewer than 4% of participants in any condition produced a sustained fixation on the research phone. A one-way analysis of variance revealed a statistically significant effect of notification rate on mean recall accuracy, F(2, 189) = 19.4, p< .001. Bonferroni-corrected post-hoc comparisons indicated that the high-frequency group recalled significantly fewer words than the no-notification group (Cohen’s d = 0.52), and the low-frequency group also recalled significantly fewer words than the no-notification group, though the latter difference was smaller. Mean recall by condition is presented in the table and figure below.
| Condition | n | Mean % Correctly Recalled | Standard Deviation |
|---|---|---|---|
| No-notification | 64 | 79% | 10.2 |
| Low-frequency (1/min) | 64 | 70% | 11.4 |
| High-frequency (6/min) | 64 | 58% | 12.7 |
Figure 1. Mean recall accuracy by notification rate (error bars = +/- 1 SD).
The authors interpreted the pattern as evidence that the mere presence of incoming, unattended alerts on a nearby device can degrade performance on an immediate verbal-memory test. They speculated that the alerts intermittently captured cognitive resources that participants would otherwise have used to maintain the studied items, and they called for follow-up research using delayed-recall tests to determine whether the effect persists over longer retention intervals.
Halverson, M. T., Pradhan, R. L., & Coelho, A. F. (2023). Peripheral notifications and immediate verbal recall in a controlled laboratory paradigm. Memory & Cognition, 51(4), 812–826.
Using the source above, respond to all parts of the question.
Suggested time-45 minutes.
Spend the first 15 minutes reading the sources, then 30 minutes writing your response.
Using the sources provided, develop and justify an argument about which feature of a situation most strongly determines whether sleep deprivation affects a person’s moral judgment.
Niedermayer, K. T., Ojeda, S. R., & Faulkner, P. M. (2021). Five nights of sleep restriction shifts moral judgments on high-conflict but not low-conflict dilemmas. Sleep, 44(11), zsab132.
Introduction. Researchers examined whether several nights of sleep restriction would shift the proportion of utilitarian responses people give on classic moral dilemmas, and whether any such shift depended on the dilemma type.
Participants. A community sample of 142 healthy adults (mean age = 28.4 years, SD = 5.6; 55% women, 45% men; 61% White, 14% Black, 12% Hispanic/Latino, 9% Asian, 4% Other) was recruited through advertisements at two large universities in the Pacific Northwest. Participants with prior diagnoses of a sleep disorder, current shift work, or recent travel across time zones were excluded.
Method.Participants were assigned by sealed envelope to either a sleep-restricted condition (5 consecutive nights of 4 hours of time in bed, 02:30–06:30) or a rested condition (5 consecutive nights of 8 hours of time in bed, 23:00–07:00). Adherence was verified using wrist actigraphy worn continuously throughout the 5-night period; participants whose actigraphy showed deviations greater than 30 minutes from the assigned schedule on more than one night were excluded from analysis (n = 11). On the morning after the fifth night, participants completed a battery of 20 moral dilemmas drawn from a standardized item bank. Half (10) were classified as high-conflictdilemmas, in which the utilitarian action required directly causing harm to one person to save several others (e.g., the “footbridge” family of dilemmas). The other half (10) were low-conflict dilemmas, in which the utilitarian action did not require directly harming anyone. For each dilemma the participant chose either the utilitarian action or the alternative.
Results and Discussion. A 2 (sleep condition) x 2 (dilemma type) mixed ANOVA revealed a significant interaction, F(1, 129) = 22.7, p< .001. As shown in Figure 1, sleep-restricted participants endorsed the utilitarian action significantly less often than rested participants on high-conflict dilemmas (31% vs. 52%), but the two groups did not differ on low-conflict dilemmas (45% vs. 47%). The authors interpreted the pattern as evidence that sleep restriction selectively disrupts the deliberative processes engaged by emotionally aversive moral choices, while leaving more intuitive moral responses intact.
Figure 1. Percent utilitarian responses by sleep condition and dilemma type (error bars = +/- 1 SE).
Asplund, B. R., Berenstain, T. M., & Vlasov, I. K. (2022). Sleep loss and moral judgment: A meta-analysis with moderator analysis of harm severity. Psychological Bulletin, 148(7), 491–519.
Introduction. The authors meta-analyzed studies that experimentally manipulated sleep duration and measured a moral-judgment outcome, with the goal of estimating the overall effect and identifying which features of the moral scenarios moderated effect size.
Inclusion criteria and sample.Studies were included if they (a) used a within- or between-subjects design that manipulated sleep duration, (b) used adult participants aged 18 or older, (c) reported a quantifiable moral-judgment outcome, and (d) provided enough statistics to compute Hedges’ g. The final pool included 32 independent studies representing 4,318 participants from 11 countries. Each effect was coded by two independent raters for the severity of harm depicted in the scenarios (high vs. low) and for the type of moral content (care/harm vs. fairness vs. everyday moral lapses such as cheating on a small test).
Results.Across all 32 studies, the pooled effect of sleep loss on moral-judgment shift was Hedges’ g = 0.27, 95% CI [0.18, 0.36], p< .001, with moderate heterogeneity (I² = 48%). The moderator analysis summarized in Figure 2 showed that the effect was substantially larger for scenarios depicting high-severity harm than for those depicting low-severity harm or everyday moral lapses. The authors concluded that the severity of the harm at stake is the strongest situational moderator of when sleep loss shifts moral judgment.
Figure 2. Pooled Hedges’ g of sleep loss on moral judgment by harm-severity moderator (error bars = 95% CI).
Discussion. The authors emphasized that the harm-severity moderator was substantially larger than any methodological moderator (e.g., type of sleep manipulation, sample age range, or whether the study used dilemmas with personal versus impersonal harm). They concluded that the field should report harm-severity coding as a standard descriptor in future studies and cautioned that pooled effects that do not separate by severity may obscure when sleep loss does and does not matter. They also noted that almost all of the included studies used Western, industrialized samples and recommended cross-cultural replication.
Okonkwo, C. E., & Dagenais, L. M. (2024). Individual differences as moderators of sleep loss effects on moral judgment: A narrative review. Annual Review of Psychology, 75, 401–429.
Overview. This narrative review synthesized 47 primary studies published between 2005 and 2023 that examined how sleep loss interacts with stable individual differences to shape moral judgment. Unlike the meta-analysis in Source 2, the review did not pool effect sizes; instead, the authors organized the literature thematically and weighed evidence for the relative importance of different person-level moderators.
Key themes. The authors identified three person-level moderators that recurred across the reviewed literature, summarized in the table below. They argued that trait empathy was the most consistently supported moderator, with the largest and most replicable interaction with sleep loss.
| Moderator | Studies Reporting It | % Showing Significant Interaction with Sleep Loss | Direction of Effect |
|---|---|---|---|
| Trait empathy (IRI score) | 18 | 78% (14 of 18) | High-empathy individuals show ~2x larger judgment shift after sleep loss than low-empathy individuals |
| Trait self-control / conscientiousness | 15 | 53% (8 of 15) | Lower self-control associated with larger lapses in everyday moral judgments under sleep loss |
| Cultural background (individualist vs. collectivist self-construal) | 9 | 44% (4 of 9) | Mixed: collectivist participants showed smaller utilitarian shifts on personal-harm dilemmas in 4 of 9 studies |
Limitations. The authors flagged three limitations of the underlying literature: (a) most studies measured trait variables only once and treated them as fixed, ignoring fluctuation across the day; (b) reliance on self-report inventories such as the Interpersonal Reactivity Index for empathy raises concerns about response bias; and (c) the cultural moderator literature is small (only 9 studies) and disproportionately compares U.S. participants to East Asian samples, leaving most of the world unrepresented.
Conclusion.The authors concluded that person-level differences—particularly trait empathy—moderate sleep-loss effects on moral judgment more reliably than most situational features other than harm severity, and that individual-difference variables should be treated as primary predictors rather than nuisance variables in future designs.
Using the sources provided, develop and justify an argument about which feature of a situation most strongly determines whether sleep deprivation affects a person’s moral judgment. Cite each source you use (e.g., “...(Source 1).” or “According to Source 1...”). Write in complete sentences using appropriate psychological terminology.
(i) Support your claim using at least one piece of specific and relevant evidence from one of the sources.
(ii) Explain how the evidence from Part B(i) supports your claim using a psychological perspective, theory, concept, or research finding learned in AP Psychology.
(i) Support your claim using an additional piece of specific and relevant evidence from a different source than the one used in Part B(i).
(ii) Explain how the evidence from Part C(i) supports your claim using a different psychological perspective, theory, concept, or research finding than the one used in Part B(ii).