医学申请 UCAT 语言推理 定量与情景判断
UCAT 由一批英国大学在 2006 年联合设立，2026 年的官方时间表是 5 月 20 日开放注册、6 月 23 日 14:00 开放预约、7 月 13 日开始考试、9 月 24 日为最后一个考试日。BMAT 在 2023 年 10 月最后一次考试之后停办，原先用 BMAT 筛选的牛津、剑桥、帝国理工与 UCL 都改看 UCAT，所以现在准备路径只有一条。下面把题量、时长、分数换算、官方统计与十二周排期拆成十二个板块，每一个数字都可以到 ucat.ac.uk 的官方页面回查。
The UCAT was set up by a consortium of UK universities in 2006, and its official 2026 calendar runs from registration on 20 May to booking on 23 June, testing from 13 July and a final test day on 24 September. The BMAT had its last sitting in October 2023, so Oxford, Cambridge, Imperial and UCL now select on the UCAT and there is a single route to prepare for. The twelve blocks below take the question counts, timings, score conversion, official statistics and a twelve-week plan in turn, and every figure quoted can be checked on the consortium's own pages.
四个分项的题量与时长 / Question Counts and Timings
UCAT 由四个分别计时的分项组成，语言推理 44 题 22 分钟，决策判断 35 题 37 分钟，定量推理 36 题 26 分钟，情景判断 69 题 26 分钟，每个分项开始前还有一段单独计时的说明页。三个认知分项各换算到 300 到 900 分，总分区间是 900 到 2700，情景判断不给分数只给 1 到 4 档，第 1 档最高。2025 年起抽象推理被撤出考试，最高分从 3600 降到 2700，手里还留着旧版抽象推理题库的人可以先放一边。考生每年只能参加一次，重复参加按违规处理。整场考试不到两小时，中途不能暂停。很多学生以为分项之间可以回头改答案，实际上一个分项结束后就再也回不去了。
The test is built from four separately timed subtests: Verbal Reasoning with 44 questions in 22 minutes, Decision Making with 35 questions in 37 minutes, Quantitative Reasoning with 36 questions in 26 minutes, and Situational Judgement with 69 questions in 26 minutes; each one opens with its own timed instruction section. The three cognitive subtests are reported on a common 300 to 900 scale for a total between 900 and 2700, while Situational Judgement is banded from 1 (highest) to 4. Abstract Reasoning was withdrawn in 2025, which cut the maximum cognitive total from 3600 to 2700, and candidates may sit the test only once in any year.
2026 考试日历与报名费用 / The 2026 Calendar and Test Fees
官方时间表上的日期是：5 月 20 日 14:00 开放注册与助学金、便利安排申请；6 月 23 日 14:00 开放预约；7 月 13 日开始考试；9 月 10 日 15:00 便利安排申请截止；9 月 16 日 15:00 预约截止并且注册关闭；9 月 24 日最后一个考试日，同日 15:00 助学金申请截止；10 月 15 日是 UCAS 截止日；11 月初成绩送达到大学。英国境内考点的费用是 70 英镑，境外是 115 英镑，符合助学金条件的考生在结账时自动免单，不用先付款再退。官方明确建议预约 7 月或 8 月的考位，因为报名开放后 9 月的考位最紧张，一旦当天去不了，往后改期的空间也很小。注册与预约是两件事，注册只表示你打算参加，考位要在预约系统里另外挑。这一步最容易踩的坑是把注册成功当成考位到手，等到 9 月才发现附近考点只剩远郊的时段。
The official 2026 calendar opens registration and bursary applications on 20 May, opens booking at 14:00 on 23 June and starts testing on 13 July; the booking deadline is 15:00 on 16 September, the last test day is 24 September, and the UCAS deadline follows on 15 October with results reaching universities in early November. Fees are 70 pounds at a UK test centre and 115 pounds outside the UK, waived automatically at checkout for candidates holding a UCAT bursary voucher. Booking a July or August appointment is the safer choice, because September slots are in high demand and rescheduling options shrink near the end of the window; registering an account is not the same as holding an appointment, since the slot has to be chosen separately in the booking system.
语言推理 44 题 22 分钟 / Verbal Reasoning in 22 Minutes
语言推理给出一系列短文，每段配四道左右的题，44 道题 22 分钟，平均每题不到 30 秒，而每段材料在 200 到 400 词之间，逐字精读根本来不及。题目分成四类：判断陈述是真、假还是无法判断；判断作者是否会同意某个观点；四选一的推论题；以及完成句子的选择。真假判断的口径只看文中写明或能必然推出的内容，凡是需要课外常识去补的都应该归到无法判断。很多学生以为读懂了文章就能做对，失分其实集中在两处：把文章没写的常识当成事实，或者在一道题上反复纠结把后面题目的时间挤掉。可行的顺序是先扫题干关键词再回文定位，三道题都拿不准就先标记跳过去，把时间留给能确定的两道。官方免费备考资源里有分项练习与完整模考，用它摸清自己的读速比刷商业题更省时间。
Verbal Reasoning presents short passages with about four questions each, 44 questions in 22 minutes, which works out at under 30 seconds per question against texts of 200 to 400 words, so reading every word is not an option. Questions ask whether a statement is true, false or impossible to say from the passage, whether the writer would agree with a view, which of four conclusions follows, and how a sentence should be completed. Answers rest on what the passage states or implies, so reaching for outside knowledge pushes the right answer towards cannot tell; the usual way to lose time is arguing with one item while three easier ones slip past.
决策判断 35 题 37 分钟与部分给分 / Decision Making and Partial Credit
决策判断 35 题 37 分钟，单题平均约 63 秒，是三个认知分项里时间最宽的一档，题型包括逻辑推理题、形式逻辑的三段论、文氏图判断、概率与统计判断题，以及多陈述题。官方评分规则写得很具体：单选每题 1 分，多陈述题 2 分，多陈述题答对一部分给 1 分。这条部分给分规则直接决定策略，多陈述题不要因为凑不满三条而空着，把确定的那几条选上就能拿到 1 分。常见的错误是拿语言推理的读法来做决策判断，边读边凭语感挑选项，这类题需要先把条件写成符号或画成一行箭头，答案往往在图示完成后就露出来了。概率题先把样本空间写清楚再动笔，很多题目的陷阱在于把有先后顺序的抽样当成了无序组合。37 分钟的预算可以这样切：多陈述题与形式逻辑各留一半，中文口算类的题不超过一分钟。
Decision Making runs 35 questions in 37 minutes, about 63 seconds per item and the most generous pacing of the three cognitive subtests; it mixes logical puzzles, syllogisms, Venn diagram reasoning and probability with multi-statement questions. The official marking is specific, in that single-answer questions carry 1 mark, multi-statement questions carry 2 marks, and a partially correct response on a multi-statement question earns 1 mark, so blanks give away credit. Translating the conditions into symbols or a small diagram is faster than scanning the options, and writing out the sample space prevents the classic slip of counting ordered draws as unordered ones.
定量推理 36 题 26 分钟 / Quantitative Reasoning in 26 Minutes
定量推理 36 题 26 分钟，平均每题约 43 秒，屏幕上提供计算器，题目来自百分比变化、比例、单位换算、利率、面积体积、速度与密度的多步应用题。绝大多数题的解法不超过两步，坑通常埋在单位与百分比的方向上：从 80 涨到 100 是涨了 25%，从 100 降到 80 是降了 20%，两个百分比不能互换。动手之前先估算量级最省时间，算出来的答案与估算差一个数量级，基本可以断定是小单位或者小数点出了问题。学生最常把三道题的时间花在一道算不清的题上，可行做法是给每题设 40 秒上限，超时就标记再做下一题，因为后面的题往往更短。官方没有负分，空白和猜错的期望值不同，拿不准时也要选一个。心算与估算的提速靠每天 20 道不带计算器的练习，两周就能看出差别。
Quantitative Reasoning gives 36 questions in 26 minutes, roughly 43 seconds each, with an on-screen calculator; the items are multi-step percentage, ratio, unit conversion, interest, area, volume, speed and density problems. Most take no more than two steps, and the traps sit in units and in the direction of a change, since 80 rising to 100 is a 25% rise while 100 falling to 80 is a 20% drop. Estimating the size of the answer before calculating catches decimal errors, and because nothing is deducted for a wrong answer, an educated guess always beats a blank.
情景判断 69 题的四个档位 / Situational Judgement and Bands 1 to 4
情景判断 69 题 26 分钟，折合每题约 23 秒，是四个分项里节奏最快的一段。题目描述一个情景，然后要求判断某个行为是否恰当，或者某个考虑因素有多重要，选项按非常恰当到非常不恰当分格。成绩以 1 到 4 档报告，第 1 档最高，官方不公布原始分与档位之间的换算表。评分依据的是医生职业行为的标准口径，患者安全、尊重与保密、诚实、与同事协作和知情同意这几条几乎覆盖全部题目。很多学生以为这题考个人性格，实际考的是在给定情境下最合适的行为，判断时先问哪一项把患者安全放在最前面。错得最多的是挑最"亮眼"的答案而不是最稳妥的答案，比如病历信息可能外泄时，先按规定报告通常比自行处理更贴近评分口径。情景判断的 69 题里约有三分之一是重要程度判断，读题时先把问题句画出来能省不少时间。
Situational Judgement packs 69 questions into 26 minutes, about 23 seconds each and the fastest pace in the test; every item describes a scenario and asks how appropriate an action is or how important a consideration is, from very appropriate to very inappropriate. Results are reported as bands 1 to 4 with band 1 the highest, and the consortium does not publish the raw mark to band conversion. Scoring follows the professional standards expected of doctors, so patient safety, confidentiality, honesty and working with colleagues set the priorities, and acting through the proper channel is usually closer to the intended standard than handling a sensitive matter alone.
分数换算与大学的使用口径 / Scaled Scores and How Universities Use Them
各分项的原始分不能直接相加比较，因为题量不同，语言推理 44 题、决策判断 35 题、定量推理 36 题，官方把每个分项换算到 300 到 900 的共同刻度，三者相加得到 900 到 2700 的总分。换算表不公开，也不存在按比例推算的公式，任何声称能精确换算原始分的计算器都只是估计。大学用这份成绩的口径并不统一：有的只看认知总分划一条入围线，有的把它与 GCSE 成绩或预估成绩加权成一个综合分，有的把情景判断单独拎出来，作为面试前后的参考。门槛每年都会动，报考前要到目标学校的课程页看当年写的口径，照搬往届学生的经验值风险很大。这一步最容易犯的错是拿前一年某位学长的分数直接定选校名单，结果两头对不上。选校时把认知总分、情景判断档位与学术门槛分开列，再决定志愿顺序。
Raw marks cannot be compared between subtests because the question counts differ, so each cognitive subtest is converted to a common 300 to 900 scale and the three are added to give a total from 900 to 2700; the conversion table is not published, so any tool claiming an exact raw-to-scaled conversion is estimating. Universities differ in how they use the result, with some applying a cognitive cut-off, others folding it into a composite that includes GCSE results or predicted grades, and a few treating the Situational Judgement band separately. Thresholds move each year, which makes the course page for the year of application the only reliable source and turns a copied score profile into a weak basis for a school list.
2025 官方统计怎么读 / Reading the 2025 Test Statistics
UCAT 官方公布的 2025 年统计覆盖到 2025 年 9 月 26 日、共 41,354 份成绩，认知总分平均 1891 分，三个分项平均分分别是语言推理 602、决策判断 628、定量推理 661。官方在同一页写明，2025 年撤掉了抽象推理，所以 2025 年的分数不能与更早年份直接比较，拿 2024 年的 2500 分来对比属于换了一把尺子。平均分说明的是全体考生的位置，不等于任何一所学校的门槛，名校候选人的分数分布整体高于平均。读统计的顺序是先看平均分，再看十分位表定位自己落在哪个区间，最后回到目标学校的当年口径判断够不够。很多学生以为平均分 1891 就是某种及格线，实际上考试没有任何及格线，分数只是相对位次的一种表达。把十分位表打印出来贴在书桌前，模考后往上点一个点，进度会看得更清楚。
The consortium's published 2025 statistics cover 41,354 tests taken up to 26 September 2025, with a mean cognitive total of 1891 and subtest means of 602 for Verbal Reasoning, 628 for Decision Making and 661 for Quantitative Reasoning. Because Abstract Reasoning was removed in 2025, those figures are not comparable with earlier years, so a 2500 total from 2024 belongs to a different scale. Reading the decile table places a score against the cohort rather than against a school's threshold, and the UCAT has no pass mark at all.
十二周备考排期 / A Twelve-Week Plan
排期可以按十二周铺开。前四周只做分项熟悉与基础运算提速，每天 30 分钟语言推理、30 分钟定量推理，周末加一次分项计时练习。中四周进入整卷模考，每周两套，模考之后对每道错题写一句错因。后四周压到每周三套，并且把练习开始时间固定下来，比如预约的是上午 9 点的场次，模考就都在 9 点开始。每周留一次 90 分钟的整场模拟，按官方顺序语言推理、决策判断、定量推理、情景判断连做，中间不休息。定量推理的提速靠每日 20 道心算与估算，语言推理靠限时扫描，决策判断靠题型归类。最后两周不再碰新题，只把错题本里标记过的题反复做到不再错。
A twelve-week plan can open with four weeks of familiarisation and mental maths drills, 30 minutes each on Verbal Reasoning and Quantitative Reasoning on weekdays plus one timed subtest set at the weekend. The middle four weeks move to two full mocks a week, each followed by a one-line cause note on every wrong answer, and the final four weeks push to three mocks a week while fixing practice start times to match the booked appointment. One 90-minute simulation a week in the official order, without a break, is what builds the stamina the real session needs.
模考复盘与错题账本 / Mocks and the Error Log
模考的价值取决于复盘方式，只记总分没有用，要按分项把错题分成三类：不会做、会做但做错、没时间做。不会做的题回去补方法，会做但做错的题写清是哪一步的符号或者单位出错，没时间做的题只改配速不动知识。官方免费备考资源里有完整模考与分项练习，难度与真考最接近，商业题库往往偏难，用它估分容易低估自己。模考时把标记与跳过的动作练成习惯，真考屏幕上只有一个标记键，临时学来不及。学生最常把错题本写成摘抄本，抄了题目却没写错因，两周后再看还在同一处丢分。每周花 20 分钟数一遍三类错题各有多少道，比例变了就说明方法在起效。
The value of a mock depends on the review that follows it, so every wrong answer should be filed as one of three types: not known, known but misread, or never reached. Only the first type needs new method, the second needs a note on the exact sign or unit that slipped, and the third needs pacing changes rather than extra content. The consortium's free preparation resources include full practice tests at a difficulty closest to the live test, so a commercial bank that runs harder tends to depress a candidate's own estimate; keeping a cause log rather than a question notebook is what stops the same mistake returning.
考试日流程与只能考一次 / Test Day and the One-Sitting Rule
考试在 Pearson VUE 的考点进行，需要带符合官方照片证件政策的身份证明，考场提供可擦写板和笔，草稿写在板上。成绩在离开考场前由工作人员打印给你，同一份结果随后自动送达你所申请的大学，填报环节不需要你做任何事。整场考试不到两小时，中途不能暂停，也不允许带自己的纸笔进场。每年只能参加一次，重复参加按违规处理，这条规则决定了备考要留足余量，别把第一次真考当成摸底。如果当天身体状态不适合考试，官方有专门的 fitness to test 政策，符合条件可以申请改期，硬撑的后果通常是一个低分而且没有重考机会。考试中心偶尔会出现技术故障，遇到这种情况当场向监考人员报备并记录，比事后申诉容易处理得多。
The test is delivered at Pearson VUE centres, where candidates need photo identification that matches the official ID policy and are given a laminated notebook and pen for rough work. The score is printed before you leave the centre and is then sent to your chosen universities automatically, so there is nothing to upload after the session. Only one attempt is allowed in any year and a second sitting counts as misconduct, which is why the fitness to test policy exists, so that an unwell candidate reschedules instead of pushing through a bad day with no route to a retake.
BMAT 停办与选校口径 / The End of the BMAT and Building a School List
BMAT 在 2023 年 10 月的最后一次考试之后停办，过去靠它筛选的牛津、剑桥、帝国理工、UCL、利兹、兰卡斯特与布莱顿萨塞克斯等医学院，现在都在 UCAT 的名单里，那种在两项考试之间二选一的准备路径已经不存在。UCAT 官方名单上的英国院校超过四十所，牙科与部分生物医学课程也接受这份成绩。选校时按三条线筛：先看该校用不用 UCAT、用哪些分项，再看近几年入围分数的区间，最后核对 GCSE 与预估成绩的门槛，三条都对得上才放进志愿。常见的错误是只盯分数门槛而忽略 GCSE 要求，两头够不上的申请在初筛阶段就被刷掉。BMAT 时代的科学知识题不再考，物理与生物的知识要放到 A Level 成绩和面试里去证明。名单定下来之后，把每所学校的口径写成一行备注，九月底再核对一次课程页，比记在脑子里稳妥。
The BMAT had its final sitting in October 2023, and every school that used it, including Oxford, Cambridge, Imperial, UCL, Leeds, Lancaster and Brighton and Sussex, now appears on the UCAT list, so the old choice between two tests has gone. More than forty UK institutions are on that list, joined by dental and some biomedical courses. Building a school list means checking three things in order: whether the course uses the UCAT and which subtests it weights, where its shortlisting scores sat in recent years, and what academic requirements sit alongside them, since a strong test score cannot rescue an application that misses the GCSE or predicted-grade bar.
牛津AQA国际GCSE生物?OxfordAQA International GCSE Biology (9201): Onl?进口原版¥399.0📱 去京东购买京东精选需要一对一辅导？Need One-on-One Tutoring?
医学申请 UCAT 分项训练、A-Level 生物化学物理、个人陈述与面试准备
16621398022 同微信
visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图