Чтение и письмо: два обзора (сравнительные и причинные конструкции)
Слова урока · 4
- review /rɪˈvjuː/обзор (научный)
- selective /sɪˈlektɪv/избирательный
- pool /puːl/объединить (данные)
- bottom line /ˌbɒtəm ˈlaɪn/итог, суть
Фразы урока · 2
- A narrative review argues; a systematic review counts — a selective review cites what fits.Нарративный обзор доказывает; систематический — считает; избирательный обзор цитирует то, что подходит.
- Pooling studies with different measures muddies the estimate; the bottom line of both reviews is «small», but the tone differs.Объединение исследований с разными мерами мутит оценку; итог обоих обзоров — «мало», но тон различается.
Текст: «Два обзора: экраны и самочувствие подростков»
We reviewed 42 studies published since 2015 on digital media use and adolescent mental health. The great majority report negative associations between time spent on screens — particularly social media — and measures of well-being, depressive symptoms and sleep. Effects are consistently larger among heavy users and among girls aged 12–15, in whom several large surveys find a doubling of depressive symptoms between the lowest and highest use categories. While causation cannot be established from observational data, the consistency of the direction across studies, the plausibility of the mechanisms (displacement of sleep, social comparison) and the timing of the rise in adolescent distress since 2012 together warrant a precautionary approach. We recommend that clinicians ask about screen use and that policy-makers consider limits for younger adolescents.
REVIEW 2 — «Screen time and well-being: what the data do and do not show» (systematic review and meta-analysis, 2024; abstract and conclusion)
We identified 108 studies meeting pre-registered criteria (adolescents 10–19; a quantitative measure of screen time; a validated well-being or mental-health outcome) and pooled 84 with usable effect sizes. The pooled association between screen time and well-being is small and negative (r = −0.05, 95 % CI −0.07 to −0.03). Effects are larger for self-reported than for logged use (−0.07 vs −0.02), and larger in cross-sectional than in longitudinal designs; in the 11 longitudinal studies, the association is close to zero and its direction is unstable. Restricting to heavy users increases the estimate modestly (−0.09) but the subgroup is defined inconsistently across studies. We find no evidence of a threshold. The data are consistent with a small effect of screen use on well-being, with reverse causation (low mood leading to more use), and with confounding by sleep. They do not warrant population-level limits; they do warrant better measurement. Clinicians may reasonably ask about screen use as one factor among several.
Задание
Напишите критическое сопоставление двух обзоров (300–340 слов): определите предмет спора (at issue is; the crux; they differ on method / interpretation, not on the data); проверьте термины и определения (as defined; heavy users defined inconsistently; well-being vs mental health); сопоставьте на одной основе (comparable only if; self-report vs logged; cross-sectional vs longitudinal; subgroup); оцените соразмерность выводов каждого (warrants a precautionary approach vs do not warrant limits; is consistent with three explanations; effect size and CI); распакуйте одну плотную фразу (e.g. «population-level limits»; «displacement of sleep»); и дайте собственное справедливое прочтение (a fair reading of both is …) с указанием, какие данные решили бы спор.
Образец ответа
Two reviews, one dataset, two readings
At issue between the reviews is not the data but what the data warrant. Both accept a negative association between adolescent screen time and well-being, larger in heavy users, with causation unestablished; both are content for clinicians to ask about screen use. The crux is whether a small, uncertain association justifies precautionary limits, and the reviews differ on method and interpretation rather than on fact.
Method first. Review 1 is a narrative review of 42 studies selected by the authors; Review 2 is a pre-registered systematic review that pooled 84 of 108 studies meeting stated criteria. That difference matters for what «the great majority» means: in Review 1 it is a majority of chosen studies, in Review 2 a pooled estimate with a confidence interval (r = −0.05, −0.07 to −0.03). Review 2 also separates what Review 1 pools: self-reported use yields −0.07, logged use −0.02; cross-sectional designs yield more than longitudinal, and in the eleven longitudinal studies the association is near zero and unstable in direction. Review 1's strongest claim — a doubling of depressive symptoms between lowest and highest use among girls aged 12–15 — comes from large surveys that are, by Review 2's account, self-reported and cross-sectional; the two are comparable only when restricted to that subgroup and that measure, which Review 2 notes is «defined inconsistently».
Terms slip between the two. Review 1 moves from «well-being» to «mental health» to «distress»; Review 2 keeps to validated outcomes. «Population-level limits», unpacked, means rules applied to all adolescents regardless of their use — which is why Review 2, finding no threshold, declines to support them.
On proportion, Review 1 reaches beyond its data by adding mechanism plausibility and the timing of the post-2012 rise — both suggestive, neither shown in the reviewed studies — to reach «warrants a precautionary approach». Review 2 states what the data are consistent with — a small effect, reverse causation, confounding by sleep — and concludes they «do not warrant» limits but «do warrant» better measurement; this is the more proportionate verb.
A fair reading of both is: a small negative association, real and consistent in direction, larger in self-reported and cross-sectional data, near zero when measured over time, without a threshold. Whether that is worth acting on is a judgement each review makes differently; what would settle it is a longitudinal study with logged use and a consistently defined heavy-user subgroup.
lightbulbСопоставление обзоров: at issue is not the data but what they warrant → метод (narrative vs systematic; pooled; CI) → на одной основе (self-report vs logged; cross-sectional vs longitudinal; subgroup) → сползание терминов → соразмерность (warrants vs do not warrant; consistent with) → a fair reading of both; what would settle it.
Закрепите урок: упражнения со словами, фразами и грамматикой с проверкой и повторением.