#0132模式只在聚合之后出现Patterns only appear in aggregate

Meltdown · 说不清为什么，因为你在看单次Meltdown · You can't see it because you're looking at one at a time

「他到底为什么突然就炸了」——说不清的原因不是你不够细心，是**单次爆发本来就看不出模式**。模式只在聚合之后出现。 每次用三十秒记一条：**时间、前一秒发生了什么、你怎么处理、多久平息**。记两周之后贴进来，代码做三件事： **一、高发时段直方图。** 时段是最容易改的一个变量，因为它通常和饿、困、转场绑在一起。如果集中在傍晚或睡前，**先动的不该是你的应对方式，是那个时段的安排本身**。 **二、触发类型分布**（转场被打断／要求被拒／饿／困／挫败／环境／分离／要关注）。它告诉你注意力该放在哪一类，而不是每次从头猜。 **三、按应对方式分组的平均平息时长——这是核心。** 抱着不说话、共情、讲道理、转移注意、冷处理、满足要求、提高音量，哪一种在你家真的更快？**这是你自己的数据说的，不是育儿书说的。** 但同一段也会告诉你一件更重要的事：**因果方向可能是反的**——也许是爆发比较轻的时候你才用得出那种方式。所以正确的用法是把它当成一个值得试的方向，然后继续记。 **这是行为记录的归纳，不是发育或心理评估。** 「正常范围」因年龄与个体差异很大，工具不做任何判断；出现自伤伤人、持续时间明显变长、或你自己长期处于崩溃边缘时，请咨询专业人员——**最后这一条同样重要，而且最常被忽略**。 $3 解锁前因-行为-结果的完整交叉分析、按你家高发时段的前置改造方案、以及可打印的三十秒记录卡。"Why does he just explode?" — you can't say because **a single episode contains no pattern**. Patterns only exist in aggregate. Spend thirty seconds per episode logging four things: **when, what happened the second before, what you did, how long until it settled**. After two weeks, paste it in and the code does three things: **One: a time-of-day histogram.** Time is the most changeable variable, because it usually travels with hunger, tiredness and transitions. If episodes cluster before dinner or bed, **the thing to change first isn't your response — it's how that half hour is arranged**. **Two: trigger distribution** (interrupted transitions, refusals, hunger, tiredness, frustration, sensory load, separation, bids for attention). It tells you where to aim instead of guessing each time. **Three, the core: average settling time grouped by what you did.** Holding quietly, naming the feeling, explaining, distracting, withdrawing attention, giving in, raising your voice — which is actually faster in *your* house? **That's your own data, not a parenting book's.** The same section also says something more important: **the causality may run backwards** — perhaps you can only manage the calm response when the episode was mild to begin with. So treat it as a direction worth testing, and keep logging. **This is behavioural bookkeeping, not a developmental or psychological assessment.** Normal ranges vary enormously; no judgement is made here. If there's self-harm or harm to others, episodes lengthening markedly, or you yourself are close to breaking — **that last one matters just as much and gets ignored most** — talk to a professional. $3 unlocks the full antecedent-behaviour-consequence cross-analysis, a preventive plan built around your own peak windows, and a printable thirty-second log card.

**本工具是行为记录的统计归纳，不是发育、心理或医学评估，也不诊断任何状况。** 触发类型与应对方式按关键词匹配，是粗口径，会归错也会漏。「按应对方式的平均平息时长」是**相关性而非因果**——很可能是爆发较轻时你才用得出某种方式，请勿据此断定某种做法「有效」或「无效」。样本量小时结论极不稳定。儿童情绪爆发的频率、强度与持续时间的正常范围因年龄与个体差异很大，本工具不作任何比较或判断。**若出现自伤或伤人、持续时间明显延长、频率突然显著增加、伴随语言或社交发展方面的担忧，或照顾者自身长期处于情绪耗竭状态，请咨询儿科医生、儿童心理专业人员或当地相关服务。** 请勿填写孩子的姓名或可识别信息。你填写的内容不入库。**This is statistical summarisation of a behaviour log — not a developmental, psychological or medical assessment, and it diagnoses nothing.** Trigger and response classification is keyword-based, rough, and both mis-sorts and misses. Average settling time by response is **correlation, not causation** — you may simply be able to use the calmer response when the episode was milder. Do not conclude that any approach "works" or "doesn't" on this basis. With small samples the figures are highly unstable. Normal ranges for frequency, intensity and duration vary enormously by age and child; no comparison or judgement is offered. **If there is self-harm or harm to others, episodes lengthening markedly, a sudden sharp increase in frequency, concerns about language or social development, or sustained caregiver burnout, consult a paediatrician, a child mental health professional or local services.** Do not enter your child's name or identifying details. Nothing you enter is stored.