IB IA Data Collection：数据收集完整指南
What is IA Data Collection? | 什么是IA数据收集？
In the International Baccalaureate (IB) programme, the Internal Assessment (IA) is a compulsory piece of coursework that accounts for 20–25% of your final grade in most subjects. Data collection is the foundation of any successful IA — it is the process of gathering information, measurements, or observations to answer your research question. Whether you are conducting a science experiment, a geography field study, or a mathematics exploration, the quality of your data directly determines the strength of your analysis and the credibility of your conclusions. A well-designed data collection strategy shows the examiner that you understand the scientific method and can apply it rigorously.
在IB课程中，内部评估（IA）是占最终成绩20–25%的必修课业。数据收集是任何成功IA的基础——它是收集信息、测量数据或观察结果以回答研究问题的过程。无论你是在进行科学实验、地理实地研究还是数学探索，数据质量直接决定了分析的力度和结论的可信度。一个精心设计的数据收集策略向考官展示了你对科学方法的理解以及严谨应用的能力。
Why Data Collection Matters | 为什么数据收集至关重要
Many students underestimate the importance of the data collection phase, rushing through it to get to the "interesting" analysis part. This is a critical mistake. The IB marking criteria explicitly reward thorough, well-documented data collection. Under Criterion B (Methodology) and Criterion C (Analysis), examiners look for evidence of careful planning, appropriate sampling techniques, and sufficient data volume. Poor data collection leads to weak analysis, which drags down your overall IA score significantly — often by 3–5 marks out of the available 20–24. In a system where every mark counts, this is not a corner you can afford to cut.
很多同学低估了数据收集阶段的重要性，急于进入"有趣"的分析部分。这是一个严重的错误。IB评分标准明确奖励详尽、记录完整的数据收集。在标准B（方法论）和标准C（分析）中，考官寻找精心规划、适当抽样技术和充分数据量的证据。糟糕的数据收集导致分析薄弱，往往使IA总分下降3–5分——在每一分都至关重要的体系中，这是你承受不起的损失。
Primary vs Secondary Data | 一手数据 vs 二手数据
Understanding the distinction between primary and secondary data is essential for IA planning. Primary data is original information you collect yourself — experiment results, survey responses, interviews, field measurements, or observations. This is generally preferred in science IAs (Biology, Chemistry, Physics) and Geography, as it demonstrates personal engagement and hands-on research skills. Secondary data is information gathered from existing sources — government statistics, academic journals, databases, or historical records. This is more common in Economics, Business Management, and some Mathematics IAs where original data collection may be impractical.
理解一手数据和二手数据的区别对IA规划至关重要。一手数据是你自己收集的原始信息——实验结果、调查问卷回复、访谈、实地测量或观察。这在科学IA（生物、化学、物理）和地理中更受青睐，因为它展示了个人参与和实践研究能力。二手数据是从现有来源收集的信息——政府统计数据、学术期刊、数据库或历史记录。这在经济、商业管理和某些数学IA中更常见，因为原始数据收集可能不切实际。
Some of the strongest IAs combine both: primary data for the core investigation and secondary data to contextualise findings or validate results. For example, a Biology IA might compare personal experiment results with published literature values. Whatever your choice, justify it clearly in your methodology section — the examiners want to see that you have made a deliberate, reasoned decision.
一些最优秀的IA结合两者：核心调查使用一手数据，同时使用二手数据来背景化发现或验证结果。例如，生物IA可以将个人实验结果与已发表的文献值进行比较。无论你选择什么，在方法论部分清楚说明理由——考官希望看到你做出了有意识的、有理据的决定。
Quantitative vs Qualitative Approaches | 定量 vs 定性方法
Another key decision is whether your data will be quantitative, qualitative, or mixed. Quantitative data is numerical and measurable — test scores, reaction times, population counts, temperature readings. It allows for statistical analysis, graphs, and mathematical modelling. This is the standard for most science and mathematics IAs. Qualitative data is descriptive and categorical — interview transcripts, observational notes, survey comments. It is common in subjects like Psychology, English, and Visual Arts, where human experience and interpretation are central.
另一个关键决定是你的数据是定量、定性还是混合型。定量数据是数值型和可测量的——测试分数、反应时间、人口统计、温度读数。它允许统计分析、图表和数学建模。这是大多数科学和数学IA的标准。定性数据是描述性和分类性的——访谈记录、观察笔记、调查评论。这在心理学、英语和视觉艺术等以人类体验和诠释为核心的学科中很常见。
A mixed-methods approach — using both quantitative metrics and qualitative insights — can produce exceptionally rich IAs, but it requires more planning. If you choose this path, make sure each type of data serves a distinct purpose in answering your research question, rather than collecting both "just in case."
混合方法——同时使用定量指标和定性洞察——可以产生异常丰富的IA，但需要更多规划。如果你选择这条路，确保每种数据类型在回答研究问题中都有独特的作用，而不是"以防万一"地两者都收集。
Subject-Specific Strategies | 各学科的数据收集策略
Group 4 Sciences (Biology, Chemistry, Physics): Focus on controlled experiments with clear independent and dependent variables. Collect at least 5 data points per variable condition, and repeat each measurement 3–5 times to calculate means and uncertainties. Document all equipment specifications and calibration procedures. Photograph your experimental setup — visual evidence strengthens the "Personal Engagement" criterion.
第4学科组（生物、化学、物理）：专注于具有明确自变量和因变量的受控实验。每个变量条件至少收集5个数据点，每次测量重复3–5次以计算平均值和不确定度。记录所有设备规格和校准程序。拍摄实验设置照片——视觉证据可以加强"个人参与"评分标准。
Mathematics: Data collection in Math IAs often involves generating datasets through simulations, modelling real-world phenomena, or analysing publicly available data. Clearly state your data sources, sample sizes, and any transformations applied. If using regression or curve-fitting, ensure your dataset is large enough to produce statistically meaningful results — typically 30+ data points for meaningful correlations.
数学：数学IA中的数据收集通常涉及通过模拟生成数据集、对现实世界现象建模或分析公开数据。清楚说明数据来源、样本量和应用的任何转换。如果使用回归或曲线拟合，确保数据集足够大以产生有统计意义的结果——通常需要30个以上的数据点才能得到有意义的相关性。
Geography: Fieldwork is central to Geography IAs. Use systematic sampling (transects, quadrats) or stratified sampling to ensure representative data. Collect both quantitative measurements (slope angles, river velocity, pebble size) and qualitative observations (land use descriptions, environmental quality indices). Always include a site map and photographs of your field sites.
地理：实地考察是地理IA的核心。使用系统抽样（样线、样方）或分层抽样以确保数据具有代表性。同时收集定量测量（坡度角、河流流速、卵石大小）和定性观察（土地利用描述、环境质量指数）。务必包含实地地图和实地照片。
Sampling and Reliability | 抽样与可靠性
Your sampling method can make or break your IA. Random sampling gives every member of the population an equal chance of selection, minimising bias. Stratified sampling divides the population into subgroups and samples from each proportionally — useful when you expect different behaviours across groups. Systematic sampling selects every nth item, which is practical for fieldwork but can introduce periodic bias. Convenience sampling (using whoever is available) is the weakest method and should be avoided unless you can strongly justify it.
抽样方法可以成就或毁掉你的IA。随机抽样给总体中每个成员平等的选择机会，最小化偏差。分层抽样将总体分成子组并按比例从每组抽样——当你预期不同组有不同行为时很有用。系统抽样选择每第n个项目，对实地考察很实用但可能引入周期性偏差。便利抽样（使用任何可用的人）是最弱的方法，除非你能有力地证明其合理性，否则应避免使用。
Beyond sampling method, consider your sample size. For surveys, aim for at least 50 responses to draw meaningful conclusions. For experiments, more trials mean smaller error bars. Always discuss the limitations of your sample in the evaluation section — acknowledging weaknesses shows critical thinking and can actually improve your score under Criterion E (Evaluation).
除了抽样方法，还要考虑样本量。对于问卷调查，至少争取50份回复才能得出有意义的结论。对于实验，更多的试验意味着更小的误差线。务必在评估部分讨论样本的局限性——承认弱点展示批判性思维，实际上可以在标准E（评估）下提高分数。
Ethical Considerations | 伦理考量
IB takes research ethics seriously, particularly for IAs involving human participants. If your IA involves surveys, interviews, or experiments with people, you must obtain informed consent from all participants. This means explaining the purpose of your research, how their data will be used, and their right to withdraw at any time. For participants under 18, parental consent is required. Anonymise all personal data — use participant codes instead of names. If your experiment involves animals, check your school's ethical guidelines and IB's animal experimentation policy before proceeding.
IB非常重视研究伦理，特别是涉及人类参与者的IA。如果你的IA涉及问卷调查、访谈或与人相关的实验，你必须获得所有参与者的知情同意。这意味着解释你的研究目的、数据将如何使用以及他们有权随时退出。对于未满18岁的参与者，需要父母同意。匿名化所有个人数据——使用参与者代码代替姓名。如果你的实验涉及动物，在开始前检查学校的伦理准则和IB的动物实验政策。
Common Mistakes to Avoid | 常见错误
1. Starting data collection without a clear plan. Write a detailed protocol before you begin — specify exactly what you will measure, how, when, and with what equipment. A vague plan produces vague data.
1. 在没有明确计划的情况下开始数据收集。开始前写一份详细的方案——明确你将测量什么、如何测量、何时测量以及使用什么设备。模糊的计划产生模糊的数据。
2. Insufficient data volume. Three data points do not make a trend. If your experiment shows a pattern after 3 trials, run 5 more to confirm it is not coincidence.
2. 数据量不足。三个数据点不能形成趋势。如果你的实验在3次试验后显示某种模式，再运行5次以确认这不是巧合。
3. Ignoring uncertainties. Every measurement has uncertainty. Record instrument precision and calculate propagated uncertainties. This is explicitly required in science IAs and often separates top-scoring work from the rest.
3. 忽略不确定度。每个测量都有不确定度。记录仪器精度并计算传播不确定度。这在科学IA中有明确要求，往往将高分作品与其余作品区分开来。
4. Cherry-picking data. Never discard data points that do not fit your hypothesis. Report all data, explain anomalies, and discuss their possible causes. Integrity is more important than a perfect trend line.
4. 筛选数据。绝对不要丢弃不符合假设的数据点。报告所有数据，解释异常值并讨论其可能原因。诚信比完美的趋势线更重要。
Documentation Checklist | 文档检查清单
Before you submit, verify your data collection documentation includes: (1) A clearly stated research question with defined variables; (2) A detailed methodology section explaining sampling, equipment, and procedures; (3) Raw data tables with units and uncertainties; (4) Processed data with sample calculations; (5) Photographs or diagrams of your setup; (6) Consent forms if human participants were involved; (7) A risk assessment if your experiment involved hazards. A well-documented IA is a high-scoring IA — when examiners can easily follow your process, they reward you for it.
提交前，验证你的数据收集文档包括：(1) 清晰陈述的研究问题和明确定义的变量；(2) 详细的方法论部分，解释抽样、设备和程序；(3) 带单位和不确定度的原始数据表；(4) 带有示例计算的处理后数据；(5) 实验设置的照片或图表；(6) 如涉及人类参与者则提供知情同意书；(7) 如实验涉及危险则提供风险评估。文档完善的IA就是高分IA——当考官可以轻松追踪你的过程，他们会因此给你加分。
需要IB IA一对一辅导？
资深IB教师团队，覆盖全部学科
数据分析 + 方法论指导 + 全文润色
电话/微信：16621398022
访问 aleveler.com 了解更多
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图