IB化学 IA评分 研究设计 数据分析 结论
IB DP 化学的内部评估现在只有一份作业，官方名称是科学探究（scientific investigation），在最终成绩里占 20%，SL 与 HL 用同一套标准打分，四项标准各 6 分、权重都是 25%，合计 24 分。旧版标准里的 Personal engagement（2 分）与 Communication（4 分）在 2025 年首次评估的版本中已经取消，沟通质量改为并入 Research design 与 Data analysis 的描述符里评分。评分流程是教师先内部打分，IB 再抽样做外部调节（moderation），所以最终分数取决于描述符有没有被逐条满足，而不是老师的整体印象。24 分里有 12 分落在 Conclusion 与 Evaluation 这两项上，实验做得多漂亮与这两项拿分高低没有直接关系。很多学生以为数据好看就能拿高分，阅卷时被扣分的往往是研究问题没写清变量、不确定度没有传到最终结果、评估段全是大而化之的句子。下面按四项标准逐节拆开，每一节给出 5 到 6 分档描述符的实际要求，以及可以直接照着写进报告的句式。
IB Diploma Programme chemistry now carries a single internally assessed task, the scientific investigation, worth 20% of the final grade and marked against identical criteria at SL and HL; four criteria of six marks each, weighted equally at 25%, produce a total of 24. The personal engagement and communication marks that older cohorts knew have gone, with communication quality assessed inside research design and data analysis instead, and teachers mark internally before the IB moderates a sample. Half of those marks sit in the conclusion and evaluation, so a beautifully executed practical has little bearing on them; a research question missing its variables, uncertainty that never reaches the final value and a vague evaluation paragraph are the usual reasons a strong experiment still lands below 18 out of 24.
24 分怎么构成
四项标准的官方名称是 Research design、Data analysis、Conclusion、Evaluation，每一项满分 6 分，权重各 25%。这些分数不是按段落长度分配的，而是按描述符逐条对应，描述符用的是 described、explained、justified 这类动词，动词换了，分数档就换了。教师先按描述符内部评分，IB 再抽样做外部调节，如果学校的给分整体偏离调节结果，下一年会被要求重新校准。旧标准里的实验操作技能（manipulative skills）已经不再单独评估，安全问题改为写进 Research design 里说明如何控制。24 分里 Conclusion 与 Evaluation 占一半，这两项靠的是分析之后的解释与判断，而不是实验台前的熟练程度。学生最常把时间分配反着用，十小时里九小时在实验室，剩下一小时赶报告，最后两页纸却决定了一小半分数。更稳妥的分配是把动手部分压到四五小时，数据处理与写作留足五小时以上。
The four criteria are research design, data analysis, conclusion and evaluation, six marks each and equally weighted at 25%. Marks are not spread over paragraphs but matched to descriptors, and the descriptors turn on verbs such as described, explained and justified, so a change of verb moves the report up or down a band. Teachers apply those descriptors internally and the IB moderates a sample; a school whose marks drift away from the moderation outcome is asked to recalibrate the following year. Manipulative skills are no longer assessed separately, so safety now appears inside research design as a controlled decision rather than a lab-bench impression. Half the marks sit in the conclusion and evaluation, which reward interpretation rather than skilled handling, and the commonest time-allocation error is nine hours of practical work followed by one hour of writing; four or five hours at the bench with five or more hours left for processing and writing is far safer.
研究问题要写清变量与体系
Research design 的 5 到 6 分档要求研究问题在 specific and appropriate context 里被描述，官方澄清把这句话拆成三件事：提到自变量与因变量，或者两个相关联的变量；用一句简洁的话描述所研究的体系；给出直接相关的背景理论。只写研究问题、不给背景，或者背景泛泛谈反应速率很重要，都会停在 3 到 4 分档。可以照着写的句式是：在 25.0 摄氏度的水浴中，0.10 到 1.00 mol dm-3 的盐酸与 0.50 g 碳酸钙反应，初始反应速率如何随盐酸浓度变化？后面接一句用碰撞理论与速率方程解释预期的一级关系。常见的错误是问题写成研究反应速率，变量、范围、温度条件一个都没有，方法再完整也只能拿 4 分。体系描述不必长，一句碳酸钙悬浮在盐酸溶液中的封闭气体收集体系就够了，缺了它读者无法判断结论能用在哪里。
Research design asks for a research question described within a specific and appropriate context, which the guide unpacks into three parts: the independent and dependent variables or two correlated variables, a concise description of the system studied, and background theory of direct relevance. A question with no context, or a paragraph explaining that reaction rates matter, stays in the 3-4 band. A credited example reads as follows: how does the initial rate of reaction between 0.50 g of calcium carbonate and hydrochloric acid of 0.10 to 1.00 mol dm-3 change at 25.0 degrees Celsius, followed by an expectation of first-order behaviour justified through collision theory and the rate equation. Writing only that the investigation concerns reaction rates, with no variables, range or temperature condition, caps the criterion at four marks however complete the method looks; one sentence placing the calcium carbonate suspension in a closed gas-collection system is enough to define where the conclusion applies.
方法论考虑要解释而不是罗列
5 到 6 分档要求方法论考虑被 explained，3 到 4 分档只要求 described，一个动词之差决定了能不能拿到 5 分。官方把方法论考虑归成四类：自变量与因变量的测量方法选择；数据库或模型的选择与取样方式；测量范围与质量的决定，包括区间、间隔、频率、重复次数与精度；控制变量的识别与控制方式。写报告时每一类都要给出理由，而不是只列设备清单。选择 50.00 cm3 滴定管量取盐酸，理由可以是 0.05 cm3 的读数不确定度相对 25.00 cm3 只有 0.2%，而 100 cm3 量筒的 1 cm3 不确定度是 4%，两者相差二十倍。浓度取 0.10、0.30、0.50、0.70、1.00 mol dm-3 五个点，理由是这样才能检查图形是否线性，两个点只能连成一条直线，斜率对不对无从判断。重复三次取平均，是因为气体体积读数的随机散布比仪器分辨率大。温度用水浴控制在 25.0 摄氏度，控制变量还包括碳酸钙的粒径与质量，粒径不同表面积就不同，整条速率曲线会平移。
At the top band the methodological considerations must be explained rather than described, and a single verb separates five marks from four. The guide groups them into four families: choosing methods for measuring the independent and dependent variables, choosing a database or model and sampling data, deciding the scope, quantity and quality of measurements such as range, interval, frequency, repetition and precision, and identifying control variables together with the method used to control them. Every family needs a reason attached instead of a list of apparatus. A 50.00 cm3 burette is the sensible choice for delivering the acid because a 0.05 cm3 reading uncertainty is 0.2% of 25.00 cm3, while a 100 cm3 measuring cylinder at 1 cm3 carries 4%, which is twenty times worse. Five concentrations from 0.10 to 1.00 mol dm-3 allow a linearity check that two points can never provide, three repeats address random scatter in gas volume, and a water bath holds 25.0 degrees Celsius while particle size and calcium carbonate mass are controlled because surface area shifts the whole rate curve.
数据记录既要清晰也要精确
Data analysis 的 5 到 6 分档把三个条件并列写在一起：数据的记录与处理既清晰又精确；对不确定度有恰当的考虑；数据处理准确。清晰指读者能顺利看懂你从原始读数走到结果的全过程，精确指遵守学科惯例，包括表格与图形的标注、单位、小数位与有效数字。原始数据表没有量名和单位、图形坐标轴只写一个字、浓度写成 0.05 M 而不是 0.0500 mol dm-3，这几种写法都会直接掉档。表头按量名加单位的形式写，例如体积 V / cm3、温度 T / 摄氏度，每一列的位数与仪器分辨率对齐，多留一位是允许的，忽多忽少不行。图形上标出坐标轴量名与单位，拟合直线旁边写方程与 R2，让读者一眼看出你用的是线性拟合还是别的模型。单位换算要写进报告，把 cm3 直接当 dm3 代入 c 等于 n 除以 V 的公式，结果会差 1000 倍，而这类错误在评卷里出现的频率比计算出错高得多。有效数字全篇保持同一套规则，中间步骤多留一位，最终结果按最不精确的输入给出。
Data analysis sets three conditions side by side: recording and processing that is both clear and precise, appropriate consideration of uncertainties, and accurate processing of data relevant to the question. Clear means a reader can follow every step from raw reading to result, and precise means correct scientific conventions covering table and graph annotation, units, decimal places and significant figures. A raw table with no quantity and unit headers, an axis labelled with a single word, or a concentration written as 0.05 M instead of 0.0500 mol dm-3 each costs a band; annotate axes with quantity and unit, print the fitted equation and R2 beside the line, and keep one consistent significant-figure policy with an extra digit through intermediate steps. Unit conversion deserves a written line in the report, because substituting cm3 into c equals n divided by V as if it were dm3 produces a factor of 1000 error, and that slip appears in moderated samples far more often than arithmetic mistakes do.
不确定度要传到最终结果
恰当的不确定度考虑在化学里有具体含义：知道哪些来源主导误差，并把它们传到最终结果上去。仪器不确定度最容易拿，读仪器规格就能得到，滴定管正负 0.05 cm3、温度计正负 0.5 摄氏度、电子天平正负 0.001 g，凡是取两次读数相减的量，不确定度要相加再除以差值。化学实验里最大的随机来源常常是温度漂移，温度计读数给出的相对不确定度可能只有 1%，而同一浓度三次重复得到的初始速率能相差 10%，后者才是真实散布，报告里应该用重复数据的标准偏差说话。系统误差要单独交代，天平没有校准会让所有质量同向偏大，重复多少次都不会收敛，只能靠标准砝码或已知样品发现。处理数据时把不确定度画成误差棒，如果两个浓度点的误差棒有重叠，就不该声称两者速率存在显著差异。这句话写进结论比写误差很小有用得多，它同时满足 Conclusion 描述符里对不确定度的要求。
Appropriate consideration of uncertainty carries a specific chemical meaning: identifying which sources dominate and propagating them into the final value. Instrumental contributions come straight from the specification sheet, such as a burette at plus or minus 0.05 cm3, a thermometer at 0.5 degrees Celsius or a balance at 0.001 g, and any quantity obtained as a difference needs its uncertainties added before division. Temperature drift is usually the largest random term in a kinetics investigation, so a 1% relative uncertainty on temperature can coexist with 10% scatter between repeats of the same concentration, and the scatter is the honest figure to report. Systematic contributions behave differently: an uncalibrated balance shifts every mass the same way and repetition never removes it, so only a standard mass or certified sample reveals it. Error bars on the graph then support a claim about rate differences, since overlapping intervals mean no significant difference, and that sentence also satisfies the uncertainty requirement inside the conclusion criterion.
结论要与已发表数值对比
Conclusion 的 5 到 6 分档要求结论被 justified 并且与数据分析 fully consistent，官方澄清明确说一致性意味着对处理后的数据连同不确定度一起解释。也就是说浓度越高反应越快这种句子拿不到高分，要给出斜率、量级与不确定度：初始反应速率与盐酸浓度在 0.10 到 1.00 mol dm-3 范围内呈线性关系，拟合斜率对应一级反应的预期量级，相对不确定度约 6%。第二句要求与 accepted scientific context 对比，可用来源包括已发表论文、已发表数值、课本与课程笔记，引用要详细到能被追溯，作者、年份与出处都不能省。做中和焓实验时，把实测的 -55.8 kJ mol-1 与强酸强碱中和焓的公认值 -57.1 kJ mol-1 并排写出，再说明两者相差 2.3%，比写结果合理有用得多。常见的错误是用 R2 代替物理判断，R2 等于 0.98 只能说明点接近直线，不能说明斜率与文献值一致，这是两个独立的判断。范围外推也要交代清楚，1.00 mol dm-3 以上的浓度没有测过，结论就不应该覆盖那段区间。
A top-band conclusion is justified, fully consistent with the analysis and compared against the accepted scientific context, and the guide states that consistency means interpreting the processed data together with their uncertainties. Writing that faster reactions follow from higher concentration earns little, while a sentence reporting the slope, its magnitude and its uncertainty does real work, for example a linear dependence of initial rate on acid concentration between 0.10 and 1.00 mol dm-3 with about 6% relative uncertainty. The comparison must be traceable, drawing on published papers, published values, course notes or textbooks with author, year and source given, so setting an experimental -55.8 kJ mol-1 beside the accepted -57.1 kJ mol-1 for a strong acid and strong base and noting a 2.3% difference is exactly what the descriptor rewards. A high R2 proves linearity alone and never agreement with a literature value, and a conclusion drawn from data ending at 1.00 mol dm-3 should say that anything above that range remains untested.
评估要指出弱点的影响大小
Evaluation 是 24 分里最容易写空的一项，因为 1 到 2 分档只要求 generic weaknesses，5 到 6 分档要求解释 specific methodological weaknesses 的 relative impact。通用的句子是实验存在人为误差、需要更小心，这句话对任何实验都成立，所以不给分。具体的写法要落到环节与数字上：用 100 cm3 量筒量取盐酸带来的 1 cm3 不确定度占 5 cm3 取样体积的 20%，比浓度范围本身的影响还大，所以斜率的不确定度主要来自取样而不是计时。limitations 是另一类，指的是结论被限制的范围：数据只到 1.00 mol dm-3，向外推就不成立；体系假定温度恒定，而反应放热会让实际温度略升，反应越快到后面偏得越多。改进措施必须对着上面某一条弱点写，用滴定管替代量筒、用数据记录仪每 0.1 秒采一次温度、把浓度范围扩到 2.00 mol dm-3，这三条都算改进，而下次更仔细不算。5 到 6 分档的分界线就在这里：能不能说出哪一条弱点对结论影响最大，以及其余弱点为什么排在后面。
Evaluation is the emptiest section in most reports, because the 1-2 band accepts generic weaknesses while the 5-6 band asks for the relative impact of specific methodological ones. A sentence about human error and the need for greater care applies to every experiment ever run, which is precisely why it scores nothing; credited writing names the step and the number, such as the 1 cm3 uncertainty of a 100 cm3 measuring cylinder on a 5 cm3 sample sitting at 20%, larger than the effect of the concentration range, so the gradient uncertainty comes mainly from sampling rather than timing. Limitations form a separate family: data ending at 1.00 mol dm-3 cannot support extrapolation, and an assumed constant temperature drifts upwards as an exothermic reaction proceeds. Improvements must answer those same points, so replacing the cylinder with a burette, logging temperature every 0.1 seconds or extending the range to 2.00 mol dm-3 are genuine changes, while being more careful is not; the band boundary is whether the report ranks the weaknesses by the size of their impact on the conclusion.
3000 词与十小时的分配
报告上限是 3000 词，图表、数据表、公式与计算、引用、参考文献与页眉都不计入词数，所以可行的时间与篇幅分配是研究设计与方法约 600 词、数据处理与分析约 900 词、结论约 400 词、评估约 700 词，剩下留给引言与变量说明。报告开头要写清四样东西：探究标题、考生代码、所有小组成员的代码、词数。IB 给的时长是 10 小时，动手做实验通常占一半，另一半留给数据处理与写作，把十小时全花在实验室、最后两小时赶报告，是长文失分最集中的原因。使用二手数据是允许的，条件是研究设计里要说明数据从哪来、怎么取样，并且尽可能讨论它的不确定度，直接贴一张网上的表格而不讲来源会被当作方法描述不全。涉及人类志愿者时，IB 的 Examiner Instructions 2025 要求学校确认已取得同意，用自己唾液或汗液做样本属于例外情形。词数统计按正文算，把公式、表格与图注算进去会虚高几百词，写进报告的词数应当与正文实际长度一致。
The cap is 3000 words with charts, data tables, equations and calculations, citations, bibliography and headers excluded, which makes a workable split roughly 600 words for research design and method, 900 for processing and analysis, 400 for the conclusion and 700 for evaluation, leaving the remainder for the introduction and variable definitions. Four items belong at the very start: the title, the candidate code, the codes of any group members and the word count. Ten hours are allocated in total and the bench work usually takes half, so a report written in the final two hours is the commonest way long pieces lose marks; secondary data are permitted provided the research design explains where the data came from and how they were sampled, with uncertainty discussed as far as possible. Where human volunteers are involved the IB examiner instructions for 2025 expect consent to have been obtained, with a student's own saliva or sweat as the stated exception, and the word count should match the actual length of the prose once equations, tables and captions are left out.
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?
A-Level / GCSE / IB / KS2-KS3 全科辅导，IB Chemistry IA 选题、数据处理与报告批改
16621398022 同微信
aleveler.com
或截图后用「扫一扫 → 相册」选择本图