Every issue of the Journal of High School Social Science. All articles are open access; click a title to read the paper.
Bryan Kyung, PhD · Editor-in-Chief
Ask a high school student what a scientist does and most can describe an experiment. Ask what a social scientist does and the picture goes blurry. Yet the questions that shape our lives — why people cooperate or cheat, how policies change behavior, what makes a community trust its institutions — are social-scientific questions, and they are answered with the same tools that power any empirical field: a clear question, honest data, and a method that could prove you wrong. The Journal of High School Social Science exists because too few students ever get to use those tools on a real problem before they leave for university.
There is no shortage of programs that promise teenagers a publication. What is scarce is rigor. A great deal of "high school research" is a literature summary dressed up as a study, or an opinion essay with citations. We built JHSS around a single, deliberately narrow commitment: we publish empirical and analytical work only. Every paper in this issue collects original data or analyzes an existing dataset to answer a question the author defined and could have answered either way. That constraint is not a barrier to entry; it is the entry. It is what separates a finding from a feeling.
We also treat peer review as teaching, not gatekeeping. Each submission receives a structured report meant to leave its author a more capable researcher, whether or not the paper is accepted. Round 1 is a decisive go/no-go gate on contribution and methodology, returned within two weeks, because a student's time is worth more than a slow drip of vague feedback. Roughly three in ten submissions advance. We think an honest "no, and here is exactly why" respects a young researcher far more than carrying them through revisions that were never going to end in publication.
The ten studies that follow were written by students in grades nine through twelve. They run a field experiment in a school cafeteria, map grocery access against bus routes, analyze election turnout against the decline of local news, and test whether brighter streetlights actually reduce crime. None of them is perfect; every one names its own limitations, which is precisely what real research does. Read together, they are an argument that adolescents can do genuine social science when someone holds them to the standard and shows them how to meet it.
That is the work of this journal. To the students reading this and wondering whether their question is worth pursuing: it almost certainly is. Define it sharply, gather the evidence, and let the data surprise you. We look forward to reviewing what you find.
Abstract. Does the mere arrival of a phone notification impair adolescent working memory, even when the phone is not checked? Sixty-two students completed an operation-span task under three within-subjects conditions: phone absent, phone present and silent, and phone present with periodic notifications. Performance declined significantly when notifications arrived (a 14% drop in span score relative to phone-absent), while silent presence produced a smaller, non-significant decrement. Self-reported urge to check the phone mediated part of the effect. Findings support attentional-capture accounts over simple presence effects and suggest that notification management, not device removal alone, may matter most for adolescent concentration.
Keywords: working memory, smartphones, attentional capture, adolescents, within-subjects experiment
Abstract. This study tests whether making a vegetable side the default option increases teen uptake without reducing satisfaction. Over four weeks, a school cafeteria alternated daily between an opt-in condition (students requested a vegetable) and an opt-out default (a vegetable was plated unless declined). Across 3,180 transactions, the default raised vegetable acceptance from 31% to 58%, a 27-point increase, with no significant change in overall meal-line time or in a brief satisfaction survey. Plate-waste audits showed a modest rise in uneaten vegetables, tempering the welfare interpretation. Results replicate adult default effects in an adolescent setting and inform low-cost school nutrition policy.
Keywords: nudges, default effects, food choice, field experiment, school nutrition
Abstract. Using an anonymous peer-nomination survey of 410 students, this study maps the friendship network of a single high school and measures homophily across grade, activity, and self-reported background. Network statistics reveal strong clustering by extracurricular activity and a segregation index well above chance for background, but substantially lower segregation among students enrolled in mixed-track elective courses. An exponential random graph model indicates that shared electives predict cross-group ties even after controlling for proximity. The results suggest that institutional structures schools can actually change — course mixing — meaningfully shape who befriends whom.
Keywords: social networks, friendship, homophily, segregation, ERGM
Abstract. Do communities that lose local newspapers see lower youth turnout? This study merges a national newspaper-closure dataset with county-level voter-file estimates of turnout among 18–24-year-olds in the 2024 general election. Counties classified as "news deserts" show turnout roughly 4.6 points lower among young voters, a gap that shrinks but persists after controlling for income, education, broadband access, and prior turnout. The association is weaker where strong local digital outlets exist. While observational and not causal, the pattern is consistent with the theory that local journalism lowers the information cost of civic participation for the youngest, least-resourced voters.
Keywords: civic engagement, news deserts, voter turnout, youth politics, secondary data analysis
Abstract. This randomized trial tests whether low-stakes retrieval quizzes improve six-week retention compared with rereading. Within four biology sections (N = 96), students were randomly assigned by topic to either repeated retrieval practice or matched rereading time. On a delayed test, the retrieval condition outscored rereading by 0.41 standard deviations, with the largest gains on application items rather than recall items. A confidence-accuracy analysis showed retrieval also reduced overconfidence. The study replicates a well-established laboratory finding inside a real classroom with intact sections, supporting wider use of brief retrieval quizzing in secondary science instruction.
Keywords: retrieval practice, retention, randomized trial, science education, testing effect
Abstract. A mid-sized city upgraded streetlights in selected neighborhoods on a staggered schedule, creating a natural experiment. Using monthly open crime data for 24 neighborhoods over three years, this study applies a difference-in-differences design comparing upgraded and not-yet-upgraded areas. Nighttime property crime fell an estimated 9% in upgraded neighborhoods relative to controls, while daytime crime and violent crime showed no significant change — a pattern consistent with deterrence through visibility rather than displacement. Event-study plots show no pre-trend differences, supporting the design. Limitations include reliance on reported crime and possible spatial spillovers.
Keywords: crime, street lighting, difference-in-differences, deterrence, urban policy
Abstract. Standard food-desert measures use straight-line distance, which ignores how carless residents actually travel. This study geocodes every full-service grocery store in a city of 180,000 and computes public-transit travel time from each census block group using a transit routing API. By transit time, 22% of block groups lie more than 30 minutes from a grocery store — nearly double the share flagged by distance alone — and the transit-defined gaps concentrate sharply in lower-income areas. The analysis demonstrates that mode-aware accessibility metrics reveal inequities that conventional measures conceal, with direct implications for transit and zoning decisions.
Keywords: food deserts, GIS, accessibility, public transit, spatial analysis
Abstract. This study content-analyzes 600 climate-related Instagram posts from teen-oriented accounts, coding each for dominant frame (threat, solution, justice, or individual-action), emotional tone, and engagement. Two coders achieved strong reliability (Krippendorff's α = 0.81). Threat frames were most common (43%) but solution frames received significantly higher engagement per follower. Posts emphasizing individual action outnumbered those addressing systemic or policy change by more than three to one. The findings describe the climate information environment that adolescents actually encounter and raise questions about whether dominant framings promote efficacy or fatigue.
Keywords: content analysis, climate change, social media, framing, intercoder reliability
Abstract. Using state-level labor data, this study estimates the effect of three recent statewide minimum-wage increases on teen (16–19) employment with the synthetic control method, constructing a weighted combination of comparison states that tracks each treated state's pre-policy trajectory. Across the three cases, post-increase teen employment rates diverged only modestly from their synthetic counterparts, with two of three estimates statistically indistinguishable from zero and the third showing a small decline. Placebo tests support the inference. The results sit within the contested empirical literature and illustrate how transparent counterfactual methods can be applied to a longstanding policy debate.
Keywords: minimum wage, teen employment, synthetic control, labor economics, policy evaluation
Abstract. Does switching between languages change how a speaker is judged? In a randomized vignette experiment, 148 bilingual high school students rated a described peer who either spoke one language consistently or code-switched within a conversation. Code-switching speakers were rated as significantly more sociable and in-group, but slightly less "professional," with the professionalism penalty appearing only when the imagined setting was formal. Participants' own code-switching frequency moderated the effect, reducing the penalty. The study provides experimental evidence that evaluations of code-switching are context-dependent rather than uniformly negative, complicating deficit narratives about bilingual speech.
Keywords: code-switching, bilingualism, language attitudes, vignette experiment, sociolinguistics