IB化学 IA评分 研究设计 数据分析 满分
IB 化学新大纲(首次评估 2025)把内部评估 IA 正式更名为科学探究 Scientific Investigation,分值占总成绩的 20%,SL 与 HL 权重相同。IA 是整门课里最可控的一块分数:不需要临场发挥,不需要运气,只要你读懂评分标准并把每一项要求写进报告,高分就是水到渠成的事。这篇攻略从四个评分维度逐一拆解考官到底在看什么。
Under the IB Chemistry syllabus first assessed in 2025, the Internal Assessment is officially called the scientific investigation and counts for 20% of the final grade at both SL and HL. It is the most controllable block of marks in the whole course: no exam pressure, no luck, just a report you write over many weeks. If you understand what the descriptors actually ask for and put every requirement into a specific paragraph, top marks follow naturally. This guide breaks down exactly what the examiner looks for in each of the four criteria.
先读懂分数结构:24 分如何分布
整份报告满分 24 分,由四个各占 6 分的维度构成:研究设计 Research Design、数据分析 Data Analysis、结论 Conclusion、评价 Evaluation。最容易被忽视的事实是:结论加评价合计 12 分,正好占 IA 总分的一半。很多学生把时间全花在漂亮的实验和图表上,最后在结论与评价里丢掉大量分数,非常可惜。报告篇幅上限为 3000 词,但数据表格、方程式、计算过程、引用和参考文献都不计入词数。旧大纲里单独评分的个人参与度与交流维度已被删除,新标准更看重高阶思维:你能否用数据支撑一个化学结论,并诚实评估自己方法的局限。
The report is marked out of 24 using four equally weighted criteria: research design, data analysis, conclusion, and evaluation. The fact most students miss is that conclusion plus evaluation are worth 12 marks, exactly half of the IA. Many people pour their effort into a polished method and beautiful graphs, then bleed marks in the final two criteria. The written report has a firm maximum of 3,000 words, while data tables, equations, calculations, citations, and the bibliography are excluded from that count. Personal engagement and communication, which were separate criteria in the old syllabus, are gone. The new descriptors reward higher-order thinking: can you turn data into a defensible chemical conclusion and honestly appraise your own method.
研究设计 Research Design:让问题可回答
这一维度拿满分的关键是四个字:可答、可复现。研究问题必须同时点名自变量和因变量,界定体系边界,并且指向一个真正可测量的化学量。例如:盐酸浓度如何影响镁带与盐酸反应的平均初速率?这个问题里,浓度是自变量,初速率是因变量,温度、镁带表面积、酸液体积全部是控制变量。假设要建立在化学原理之上,比如用碰撞理论解释:浓度升高,单位体积内有效碰撞频率增大,反应速率加快。方法部分要细到任何一位同学照做都能得到同样的结果:器材清单、操作步骤、每个控制变量如何控制与测量、安全与环境注意事项,一项都不能少。背景介绍则把问题挂靠到大纲主题上,如化学动力学、酸碱化学或氧化还原,证明你清楚自己在做什么、为什么做。
Full marks here need an answerable research question and a reproducible method. State the independent variable, the dependent variable, and the boundaries of the system, all tied to a chemical quantity you can actually measure. For example: how does the concentration of hydrochloric acid affect the initial rate of reaction between magnesium ribbon and the acid? Here concentration is the independent variable, initial rate is the dependent variable, and temperature, ribbon surface area, and acid volume are all controlled variables. Build the hypothesis on chemistry, for instance collision theory: as concentration rises, effective collisions per unit volume increase, so the initial rate increases. Write the method so precisely that any classmate could repeat it and obtain the same results: full apparatus list, step-by-step procedure, how each controlled variable is controlled and measured, plus safety and environmental notes. In the background section, link the question to a syllabus topic such as kinetics, acid base chemistry, or redox, to show that you understand what you are doing and why.
数据分析 Data Analysis:用证据说话
数据部分拿满分要满足三点:原始数据完整、处理科学、呈现清晰。每张原始数据表都要标注单位、测量不确定度和一致的有效数字;每个条件至少三次重复,并报告均值与标准差。不确定度必须按规则传递,例如秒表计时与量筒量取体积各自有相对不确定度,合成后才是结果的总不确定度。作图用处理后的数据,误差棒反映随机误差,坐标轴写清变量与单位,趋势、规律和异常点都要在正文里解释,而不是把图往报告里一放就完事。有效数字前后保持一致,绝不在计算中人为增加精度。举个具体例子:测定白醋中乙酸含量,若三次滴定均值为 4.52 g/100 mL,需同时给出这组数据的标准差和滴定管 0.05 mL 估读误差的传递结果,考官才能判断你的结论有多可靠。
A full-mark data analysis shows complete raw data, sound processing, and clear presentation. Every raw data table carries units, measurement uncertainty, and consistent significant figures; take at least three repeats per condition and report the mean and standard deviation. Propagate uncertainties correctly: the relative uncertainty of stopwatch timing and that of a measuring cylinder combine to give the total uncertainty of the result. Plot processed data with error bars that reflect random error, label axes with variable and unit, and explain trends, patterns, and anomalies in the text instead of leaving the graph to speak for itself. Keep significant figures consistent and never invent extra precision. A concrete example: when finding acetic acid content in white vinegar, if the three titration trials give a mean of 4.52 g per 100 mL, you must also report the standard deviation of those trials and the propagated uncertainty from the 0.05 mL burette estimation, so the examiner can judge how reliable your conclusion really is.
结论 Conclusion:直接回答研究问题
结论必须开门见山回答研究问题,并且只用你自己的数据说话。引用具体数值:均值、斜率、活化能或含量测定结果,说明结果支持还是反驳假设,再用化学原理解释为什么。把实验值与文献值或课本值对比,计算百分比差异,并讨论差异的来源。例如乙酸含量实测 4.52 g/100 mL,文献值 5.00 g/100 mL,百分比差异为 9.6%,可以讨论温度影响、终点判断早晚或样品本身差异等因素。结论切忌过度延伸:数据只能支撑到什么程度,就说到什么程度,超出证据范围的断言会被判定为无据结论。反过来,如果数据不支持假设,诚实说明并给出可信的化学解释,同样能拿到分数。
The conclusion must answer the research question directly and rely only on your own data. Quote specific numbers: the mean, a gradient, an activation energy, or a measured content. State whether the hypothesis is supported or refuted, then explain the outcome through chemistry. Compare your experimental value with a literature or textbook value, compute the percentage difference, and discuss its sources. For example, if your measured acetic acid content is 4.52 g per 100 mL against a literature value of 5.00 g per 100 mL, the difference is 9.6 percent; discuss factors such as temperature effects, early or late end point detection, or sample variation. Do not overreach: claim only what the data can support, because assertions beyond the evidence are judged as unsupported. Conversely, if the data refute the hypothesis, saying so honestly with a credible chemical explanation still earns marks.
评价 Evaluation:具体才得分
评价维度是区分 6 分与 3 分报告的分水岭。考官最反感的是"存在人为误差"这类空话,它没有指出任何具体问题,自然一分不值。满分评价要做到:指出哪一个具体环节引入误差,判断它是随机误差还是系统误差,并说明它让结果偏大还是偏小。例如,滴定终点估读每次都不一样,属于随机误差,它让平行结果离散;而热量散失到环境使焓变测定值系统性偏低,属于系统误差。紧接着给出针对性的改进:用保温杯与高灵敏度温度探头减少热损失,用色度计代替肉眼判断终点,增加重复次数并检验离群值。最后提出可行的延伸方向,例如换用不同催化剂,或改变温度验证阿伦尼乌斯方程,展示你还能把探究带到哪里。
Evaluation is where six-mark reports separate from three-mark reports. Examiners strongly dislike empty phrases such as human error, which name no real problem and earn nothing. A full-mark evaluation does three things: identifies the exact step that introduced error, classifies it as random or systematic, and states whether it pushes results too high or too low. For instance, inconsistent reading of the titration end point is random error that widens the spread of replicates, while heat loss to the surroundings systematically lowers measured enthalpy changes. Then offer targeted improvements: an insulated calorimeter with a sensitive temperature probe to cut heat loss, a colorimeter to replace the visual end point, more repeats, and a check for outliers. Finally propose realistic extensions, such as testing different catalysts or varying temperature to test the Arrhenius equation, showing where the investigation could go next.
选题建议:什么样的探究能拿高分
选题决定了整个 IA 的天花板。最好的研究问题通常满足三个条件:只改变一个自变量,测量手段你能独立掌握,结果有明确的化学意义或现实价值。几个经过验证的方向:用碘钟反应或镁带与酸的反应测定初速率,研究浓度与温度的影响;用酸碱滴定测定市售白醋、抗酸片或果汁中的含量,并与标签值对比;用简易量热计测定不同醇的燃烧焓或中和焓,比较实验值与文献值;用色度计结合比尔-朗伯定律测定高锰酸钾或含铜离子的浓度;通过不同温度下的速率常数作图求活化能,验证阿伦尼乌斯方程。这些题目数据充足、误差可控、文献可比,每个维度都有话可说。反过来,题目太宽,比如只写"研究反应速率",会让自变量说不清;题目太窄,则数据撑不起完整的分析。
Topic choice sets the ceiling of the whole IA. The strongest questions share three features: only one independent variable changes, the measuring technique is one you can master on your own, and the result carries clear chemical meaning or everyday value. Proven directions include: measuring initial rates with the iodine clock reaction or a magnesium acid reaction under different concentrations and temperatures; titrating commercial vinegar, antacid tablets, or fruit juice and comparing the results with label values; measuring enthalpies of combustion of different alcohols or the enthalpy of neutralization in a simple calorimeter and comparing with literature values; using a colorimeter with the Beer-Lambert law to find the concentration of permanganate or copper ion solutions; and plotting rate constants against temperature to determine an activation energy and test the Arrhenius equation. These topics generate plenty of data, keep errors manageable, and offer literature values for comparison, so every criterion has material to work with. A topic that is too broad, such as simply studying reaction rate, leaves the independent variable unclear, while one that is too narrow produces too little data for a meaningful analysis.
字数分配与提交前检查清单
一份冲满分报告的参考字数分配:研究问题与背景 300 至 450 词,变量与方法设计 600 至 850 词,数据采集与处理 650 至 800 词,结果解释 250 至 350 词,结论 300 至 400 词,评价与改进 450 至 600 词,总计约 2600 至 3000 词。提交前把四个维度当成评分卡逐条自查:研究问题是否同时包含自变量与因变量;方法能否被他人完整复现;原始表是否含单位与不确定度;均值、标准差与误差棒是否齐全;结论是否引用具体数据并与文献对比;评价是否区分了系统误差与随机误差并给出可执行的改进。每一项都能在报告里找到对应的句子时,你离满分就只有一步之遥了。
A proven word budget for a top report: research question and context 300 to 450 words, variables and methodology 600 to 850, data collection and processing 650 to 800, interpretation 250 to 350, conclusion 300 to 400, evaluation and improvements 450 to 600, roughly 2,600 to 3,000 in total. Before submission, treat the four criteria as a scorecard and check line by line: does the research question name both the independent and dependent variables; can someone else fully reproduce the method; do raw tables show units and uncertainties; are means, standard deviations, and error bars all present; does the conclusion cite specific data and compare with literature; does the evaluation separate systematic from random error with workable improvements. When every item maps to a clear sentence in your report, you are one step away from full marks.
高频失分点值得单独列出对照检查:研究问题漏掉自变量或因变量,考官无法判断你在研究什么;只有一组读数、没有重复,分析就没有统计意义;图表缺单位或误差棒,数据处理被视为不完整;结论没有引用自己算出的数值,只剩一句"假设成立";评价写"人为误差"这类套话,毫无针对性。这五条每一条都对应一到两分的差距,改掉它们,报告立刻上一个台阶。
The five most common mark losers deserve a dedicated check: a research question that omits the independent or dependent variable leaves the examiner unsure what you actually studied; a single set of readings with no repeats gives the analysis no statistical meaning; graphs without units or error bars are treated as incomplete processing; a conclusion that never quotes your own calculated values is only an assertion that the hypothesis holds; and an evaluation full of generic human error phrases has no focus at all. Each of these costs one or two marks, so fixing them lifts the whole report to the next level.
海外直订AQA GCSE Chemistry Student Book AQA GCSE化学学生手册¥588.0📱 去京东购买京东精选需要一对一辅导? Need One-on-One Tutoring?
IB / A-Level / GCSE 化学等全科辅导,中英双语授课
或截图后用「扫一扫 → 相册」选择本图