「依恋地图」测评的理论根基，是心理学史上最有生命力的理论之一——依恋理论（Attachment Theory）。这篇文章讲清楚它的来龙去脉，以及你在测评里答的 36 题究竟在测什么。
从婴儿到成人：理论的演化
1969 年，英国心理学家 John Bowlby 提出：婴儿与主要照顾者之间会形成一种先天预装的「依恋」联结，其质量塑造个体对「自己是否值得被爱」「他人是否可靠」的内在预期。Mary Ainsworth 随后通过「陌生情境实验」识别出婴儿的不同依恋模式。
1987 年，Hazan 和 Shaver 提出了一个划时代的假设：婴儿期的依恋模式会延续到成年亲密关系。此后，Brennan、Clark 和 Shaver 于 1998 年发表的 ECR 量表（后修订为 ECR-R）确立了成人依恋的两个基本维度：
- 焦虑维度：对被抛弃的担忧程度——「TA 会不会离开我」；
- 回避维度：对亲密的舒适程度——「我需不需要靠这么近」。
两个维度的高低组合，就是你在依恋地图里看到的四种类型：安全型（低焦虑低回避）、痴迷型（高焦虑低回避）、疏离型（低焦虑高回避）、恐惧型（高焦虑高回避）。
它能做什么，不能做什么
能做的：给你的关系模式一个结构化的语言，帮你看见「为什么我总在关系里重复同样的剧本」。大量研究显示，依恋类型与关系满意度、沟通模式存在稳定相关。
不能做的：它不是宿命。依恋类型是「当前状态的快照」而非「终身判决」——研究显示相当比例的人在几年内会发生类型转变，尤其是经历重要关系或心理咨询之后。它也不能替代临床诊断：如果你正在经历严重的关系或情绪困扰，请寻求专业帮助。
依恋地图画的是你的起点，不是你的终点。
为什么是「两个维度」，而不是「四个标签」
很多人做完测评只记住了自己的类型名，但类型其实是个简化。你真正的结果是两个分数：焦虑维度上你落在哪，回避维度上你落在哪。
用二维坐标的好处在于，它承认程度的存在。「高回避」不是开关，而是刻度——5.5 分的回避和 6.5 分的回避体验完全不同，但贴上类型标签后它们看起来一样。四个类型（安全、痴迷、疏离、恐惧）只是这张坐标图上画的四个象限，方便记忆和交流，不是为了把你钉死在某一格。
所以读完结果的正确姿势是：先看两个维度各自的分数和位置，再看类型名。维度告诉你「在哪个方向上偏了多少」，类型只负责给你一个起讨论的词汇。
36 道题在测什么：ECR 的工作方式
「依恋地图」沿用了 ECR 量表的经典结构：36 道题，一半测焦虑维度，一半测回避维度，交替出现以降低作答的套路感。
焦虑维度的题目关心的是「你对被抛弃的警觉程度」——例如对方没有及时回复时你的内心戏、你对关系承诺的渴求强度。回避维度的题目关心的是「你对亲密的舒适区」——例如被依赖时是否想逃、袒露脆弱时的难度。
计分上，你在每个维度得到一个平均分，以 4.0 为分界判断高低，四个象限交叉出类型；常模参照采用了中文样本的分布，因此你看到的「偏高」「偏低」是相对中文人群而言的。
需要知道的是：自评量表测的是你此刻如何看自己。情绪状态、最近的一段关系、甚至作答当天的吵架，都会让分数小幅移动——这也是为什么间隔一段时间重测可能得到不同结果。
什么情况下这套解读不适用
依恋类型描述的是亲密关系中的一般倾向，它有几个明确的适用边界。
首先，它不是临床诊断。如果你的困扰是持续的——长期的关系痛苦、反复陷入让你受伤的关系模式、或已经影响到工作、学业与日常生活功能——这就超出了一份自评量表能覆盖的范围，值得寻求专业的心理评估。量表是自我探索的入口，不是结论性的判断。
其次，依恋风格是关系特异性的：你对伴侣的模式，未必等同于你对父母或密友的模式。同一个人在不同关系里可以落在不同象限——这不矛盾，恰恰说明依恋是「你与某个人之间的系统」，而不全是你一个人的出厂配置。
最后，分数描述的是当下。重大关系事件、稳定的新关系或心理咨询都可能移动你的坐标——把它当成会变化的地图，而不是固定的命盘。
拿到结果之后怎么用
一张地图的价值取决于你拿它做什么。三个推荐的用法：
用来命名重复出现的剧本。当你又一次「因为对方没回消息而脑补分手」时，你能认出：这是焦虑维度在报警，不是事实在报警。能给情绪命名，就已经卸掉了一半强度。
用来和伴侣对齐语言。把「我需要空间」翻译成「我的回避刻度在响」，比「你别逼我」好接得多。共享的词汇能把两个人的冲突从「谁有问题」改写成「我们的坐标怎么相处」。
用来观察变化。隔半年重测一次，看坐标有没有移动。移动的方向，往往比你现在站在哪更重要。
常见疑问
「类型会变吗？我去年测的和今年不一样。」会。依恋类型是状态快照而非终身判决，关系经历与自我成长都可能移动坐标。结果不一致不等于之前测错了，而是你变了。
「安全型是不是就万事大吉了？」不是。安全型描述的是对亲密与独立的平均舒适水平，不是关系质量的保证书——沟通、价值观、冲突处理能力一样都不能省。
「伴侣不肯测怎么办？」不必勉强。你可以用这篇文章里的二维框架，先理清自己在关系里的坐标。理解任何一个人的模式，都比两个人都带着无名火硬碰硬要好。
想绘制自己的依恋地图？开始 36 题测评。
The theoretical foundation of the Attachment Map assessment is one of the most durable theories in the history of psychology: attachment theory. This article traces where it came from, and explains what the 36 items you answer are actually measuring.
From infants to adults: how the theory evolved
In 1969, the British psychologist John Bowlby proposed that infants form an innate, pre-wired "attachment" bond with their primary caregiver, and that the quality of that bond shapes a person's internal expectations about "whether I am worth being loved" and "whether others can be relied on." Mary Ainsworth later identified distinct infant attachment patterns through the Strange Situation procedure.
In 1987, Hazan and Shaver put forward a landmark hypothesis: infant attachment patterns carry over into adult romantic relationships. Subsequently, the ECR scale published by Brennan, Clark, and Shaver in 1998 (later revised as the ECR-R) established the two fundamental dimensions of adult attachment:
- The anxiety dimension: how much you worry about being abandoned — "will they leave me?";
- The avoidance dimension: how comfortable you are with closeness — "do I need to be this close?"
Combining high and low positions on these two dimensions yields the four types you see in the Attachment Map: secure (low anxiety, low avoidance), anxious (high anxiety, low avoidance), avoidant (low anxiety, high avoidance), and disorganized (high anxiety, high avoidance).
What it can and cannot do
What it can do: give your relationship patterns a structured vocabulary, and help you see "why I keep replaying the same script in relationships." A large body of research shows stable associations between attachment style and relationship satisfaction and communication patterns.
What it cannot do: it is not destiny. An attachment style is "a snapshot of your current state," not "a life sentence" — research shows a substantial share of people shift styles within a few years, especially after a significant relationship or psychological counseling. It also cannot substitute for a clinical diagnosis: if you are going through serious relationship or emotional distress, please seek professional help.
The attachment map draws your starting point, not your destination.
Why two dimensions, not four labels
Most people walk away from the assessment remembering only their type name. But the type is a simplification. Your actual result is two scores: where you land on anxiety, and where you land on avoidance.
Two dimensions have one big advantage: they admit degrees. "High avoidance" isn't a switch, it's a dial — a 5.5 and a 6.5 feel completely different to live with, yet they look identical once a type label is slapped on. The four types (secure, anxious, avoidant, disorganized) are just four quadrants drawn on that coordinate plane. They exist for convenience — to give you vocabulary — not to pin you permanently to one square.
So the right way to read a result: look at each dimension's score and position first, then the type name. The dimensions tell you "how far off in which direction." The type just gives you a word to start the conversation.
What the 36 items measure: how the ECR works
The Attachment Map follows the classic ECR structure: 36 items, half measuring the anxiety dimension and half measuring avoidance, interleaved to keep your answers from falling into a rut.
The anxiety items probe "how alert you are to abandonment" — your inner monologue when a reply doesn't come, how strongly you need reassurance of commitment. The avoidance items probe "how large your comfort zone around closeness is" — whether being depended on makes you want to leave, how hard it is to show vulnerability.
For scoring, you get an average on each dimension, with 4.0 as the dividing line between high and low; the four quadrants combine into a type. The norms are referenced against Chinese-language samples, so the "above average" and "below average" labels you see are relative to a Chinese-speaking population.
Worth knowing: a self-report scale measures how you see yourself right now. Your mood, your most recent relationship, even a fight on the day you take it can nudge the scores — which is why retesting after some time can produce a different result.
When this reading does not apply
Attachment style describes general tendencies in close relationships, and it has clear boundaries.
First, it is not a clinical diagnosis. If what you're dealing with is persistent — long-running relationship pain, a pattern of repeatedly ending up in relationships that hurt you, or distress that is already affecting work, school, or daily functioning — that is beyond what a self-report scale can cover, and it deserves a professional evaluation. The questionnaire is an entry point for self-exploration, not a verdict.
Second, attachment is relationship-specific: the pattern you run with a partner may not be the one you run with parents or close friends. The same person can land in different quadrants with different people — that's not a contradiction, it's evidence that attachment is a system between you and someone else, not entirely a factory setting of yours alone.
Finally, the scores describe the present. Major relationship events, a stable new relationship, or therapy can all move your coordinates — treat it as a map that gets redrawn, not a fixed birth chart.
What to do with your result
A map is only worth what you do with it. Three recommended uses:
Name the recurring script. The next time a slow reply spirals into an imagined breakup, you can recognize it: that's the anxiety dimension sounding an alarm, not the facts. Naming the emotion already drains half its charge.
Build a shared language with your partner. "My avoidance dial is going off" lands much better than "stop pressuring me." A shared vocabulary reframes a conflict from "who is defective" to "how do our coordinates get along."
Track the movement. Retake it six months later and see whether the coordinates moved. The direction of movement usually matters more than where you currently stand.
Common questions
"Can my type change? Last year's result was different." Yes. Attachment style is a snapshot of a state, not a life sentence — relationships and personal growth can both move the coordinates. A different result doesn't mean the earlier test was wrong; it means you changed.
"Is being secure a guarantee of good relationships?" No. Security describes an average comfort level with intimacy and independence — it is not a warranty on relationship quality. Communication, values and conflict skills still have to be earned.
"My partner won't take the test. What now?" Don't push. You can use the two-dimensional framework in this article to sort out your own coordinates first. Understanding even one person's pattern beats two people colliding with unnamed frustration.
Want to map your own attachment style? Take the 36-item assessment.