GCSE地理 野外调查 抽样 数据呈现 评估
AQA Paper 3 的三段结构与分值
AQA 的 GCSE 地理 8035 一共三张卷，Paper 1 与 Paper 2 各 88 分、各占 35%，Paper 3 是 76 分、占 30%，三张卷的时长都是 1 小时 30 分钟，spec 于 2016 年 9 月起施教，2018 年 5 月和 6 月首次考试。Paper 3 分成两段，Section A 用一份提前 12 周下发的资源册做议题评估，Section B 才考野外调查。野外调查的两道题，一道给陌生的调查情境，一道问考生自己做过的调查，AQA 官网写明这两题合计占整个资格证书的 15%。AQA 官方表格把 Paper 3 的 AO1 权重写成零。很多学生以为 Paper 3 也要背大量案例，实际上这张卷子里知识回忆分几乎没有，背得再多也换不到分。评分目标由 Ofqual 统一规定，全部考试局一致：AO1 15%、AO2 25%、AO3 35%，其中 10% 落在野外调查情境，AO4 25%，其中 5% 用于回应调查数据。阅卷时最容易被看出来的，是考生能不能把资源册里的图与自己调查里遇到的问题连起来。
Every AQA GCSE Geography candidate sits three papers: Papers 1 and 2 carry 88 marks and 35 per cent each, while Paper 3 carries 76 marks and 30 per cent, all lasting 1 hour 30 minutes. Fieldwork sits in Section B of Paper 3, which asks about an unfamiliar enquiry and about the enquiry you carried out yourself, and AQA states that this part is worth 15 per cent of the qualification. Because Paper 3 carries no AO1 marks at all, revision that only memorises case studies rarely converts into credit here, and examiners reward candidates who connect the resources in the booklet with what they measured in the field.
两次教室外调查的硬指标
DfE 的 GCSE 地理学科内容要求学校让学生在教室与校园之外完成至少两次野外调查，AQA 的 spec 把这条展开成两次地理调查，每次都要用到一手数据。两次调查要在对比的环境中完成，一次偏物理地理，一次偏人文地理，其中至少一次要体现物理过程与人文过程的相互作用，例如河岸管理如何改变河道、旅游如何改变海岸。AQA 官网对场合的解释相当宽松，一个上午或一个下午的登记时段就算一次场合，所以一整天外出已经能满足两次场合的要求，前提是对比环境这一条也成立。Head of centre 每季要向 AQA 签署并提交一份 centre declaration form，写明做了哪些调查、多少学生参加，表格由考试官收集。常见的错误是学生把两次调查都安排在同一条街上，物理与人文都沾一点，结果对比环境与交互这两条都不成立，考场上答自己的调查就显得单薄。选题阶段先把环境类型写清楚，先挑两类再想方法，比先想方法再找地方省事。
DfE subject content requires fieldwork outside the classroom and school grounds on at least two occasions, and the AQA specification turns this into two enquiries that each use primary data. The two enquiries must sit in contrasting environments, cover physical and human geography, and at least one of them has to show how physical and human processes interact, such as river management changing a channel or tourism reshaping a coast. AQA treats a morning or an afternoon registration period as one occasion, so a single full day out can satisfy the two-occasion rule, although the head of centre still has to sign and submit a centre declaration form each series describing what was done and how many students took part.
可调查的问题与假设怎么写
调查的起点是一个能收集到数据的题目，不是一个宽泛的话题。稳妥的写法是先把范围缩到一条河的某一段、一个街区的三个路口、一片海滩的两条横断面，再把题目改写成可以用数字回答的句子，例如某一河段的流速是否从上游到下游递增。AQA 的考纲把选题列为调查过程的第一条，要考虑的因素包括可取得的设备、时间、安全与数据的可得性，还有支撑这个调查的地理概念。学生最常把题目写成对不对的判断，比如问河流污染严重吗，这种题目没有可测的指标，改成问某一段的垃圾密度是否高于上游，就能落到每分钟件数或每平方米件数上。假设与题目配套写，写清自变量与因变量，再写清预期出现的方向，考场落笔时这一句能省掉不少解释。选题还要能对上考纲的理论线索，河流调查通常接河流过程与地貌，海岸调查通常接侵蚀与沉积过程，资源册里的图也常从这些线索出题。
A workable enquiry starts from a question that can be answered with numbers, not from a broad topic such as river pollution. The usual route is to shrink the scale to one stretch of a river, three junctions in a neighbourhood or two transects across a beach, then phrase the aim so that the data can accept or reject it. The AQA specification lists a suitable question as the first strand of the enquiry process, and it expects students to weigh up equipment, time, safety and the availability of data as well as the geographical concept behind the question; a question that cannot be answered with numbers is not yet an enquiry.
抽样方法与样本规模
抽样要说明用什么规则把测量点或受访者选出来，而不是随手挑方便的位置。常见三类做法是随机抽样、系统抽样与分层抽样，河流调查里每隔十米取一个点属于系统抽样，按土地利用类型分块再各取若干点属于分层抽样，在街口按固定间隔拦人做问卷同样属于系统抽样。AQA 的数值技能条目要求学生在设计数据收集表时就考虑准确度、样本规模、程序、对照组与可靠性，考场答案里写明样本数量是最容易被忽略的细节。常见的错误是把样本压得太小，问卷只做十份，环境质量调查只评五个点，得出的中位数与四分位距会被个别极端值带偏，结论也就撑不住。样本规模与时间直接挂钩，一份二十分钟能完成的环境质量调查表，两个人在一条街上通常能拿到三十到四十个点，这个量级足够画散点图并比较两条街。出发前先做一次试测，把表格里含义不清的栏目改掉。
Sampling means stating the rule that picked each measurement point or respondent instead of choosing whatever was convenient. Random, systematic and stratified sampling cover most school enquiries: a point every ten metres along a river is systematic, splitting a town into land-use zones and sampling each zone is stratified, and stopping pedestrians at fixed intervals on a high street is systematic as well. AQA expects accuracy, sample size, procedures, control groups and reliability to be considered while the recording sheet is designed, and samples of five to ten readings often produce a median and interquartile range that one outlier drags sideways.
风险评估的写法与阅卷口径
风险是 Paper 3 里能直接拿分的一小块，AQA 官网给出的例子是 2019 年 6 月的一道三分题，要求写出一项物理地理调查中的风险以及它如何被降低。答题的写法是先说清危险源，再说清具体措施，例如河岸湿滑，穿防滑鞋并沿固定路线行走，只写一句注意安全拿不到分。风险要与你真正做过的调查对得上，回答海滩调查却写山上落石，阅卷时会显得是背来的模板。风险里的人和情境也要写清，在退潮时段的泥滩上取样，陷脚与涨潮是两个不同的风险，措施也不一样。很多学生以为写得越多越安全，其实三分题只给三个点，两个危险源加一个措施不会比一个危险源加两个措施多得分。这一步最容易丢分的地方是把风险写成后果，滑倒与摔伤是后果，路滑才是风险，先写风险、再写后果、最后写措施，顺序清楚就稳。
Risk is one of the few parts of Paper 3 that can be answered straight from the specification, and the example AQA publishes is a three-mark June 2019 question asking for one risk in physical geography fieldwork and how it was reduced. A full-mark answer names the hazard first and then the control, such as a slippery river bank managed with grip footwear and a fixed walking route, rather than offering a general statement about staying safe. Writing a coastal enquiry and then describing rockfall reads as a borrowed template, so the risk has to match the site your group actually visited.
物理数据：河流与海岸的测量
物理调查的数据通常来自三类测量，河道断面、流速与沉积物。用浮标法测流速时，把橙子或木片放进水里，记录它走完十米所需的时间，重复三次取平均，再用距离除以时间得到表层流速。测河床粒径可以用粒径卡片，也可以用卷尺量长轴，取样点按固定间隔排列，就能比较上下游的粒径变化。海岸调查常用两条横断面，从高潮线向低潮线每隔五米记录一次沉积物类型与剖面高度，坡度用测角器或手机倾角仪读取，剖面高度用标尺配合水准仪。测断面时两个人的分工要固定，一人读尺一人记录，中途换人容易造成读数习惯不一致，中位数会跟着漂。数据表要留出天气、日期与点位编号，同一河段两次测量的天气差异，常在评估局限时成为最好用的证据。安全信息也写进表头，水位与湿滑程度各记一行。
Physical enquiries usually rest on three kinds of measurement: channel cross-sections, velocity and sediment. Velocity with a float means timing an orange over ten metres, repeating the run three times, then dividing distance by mean time to give surface speed. Coastal work often uses two transects running from the high-water mark to the low-water mark, recording sediment type and profile height every five metres so that the two lines can be compared and the dominant process inferred.
人文数据：问卷与环境质量调查
人文调查的数据质量更依赖问卷设计，问题的措辞直接决定数据能不能用。环境质量调查是学校里最常用的一种，把噪音、垃圾、绿化、建筑维护、安全感几个栏目各设一档，从一到五打分，同一条街在同一时段由同一批人评分，总分才能互相比较。问卷题要避开双重问题与引导性措辞，问受访者本区交通是否既拥堵又不安全，没人能用一个档位回答，问是否同意限制私家车，答案会被措辞带走。样本的选取方式要写进方法说明，工作日中午的商圈与周末上午的商圈，人流结构差异很大，两份数据不能简单相加。评分尺度要有书面定义，组内成员先各自评三处再对答案，分歧记录下来往往正好是考场会问到的局限。汇总时用众数与众数组处理评分栏，用中位数处理交通计数这类容易受极端值影响的栏目。
Human enquiries live or die on questionnaire design, because the wording of each item decides whether the data can be used at all. Environmental quality surveys work well in school because five or six fixed categories, such as noise, litter, greenery and safety, are scored from one to five by the same group in the same time slot, which makes total scores comparable across streets. Items that ask two things at once, or that hint at the expected answer, produce responses that cannot be grouped into a single score, so piloting the sheet on three streets and comparing the group scores first usually exposes the ambiguous wording.
数据呈现方式的选择与理由
呈现方式的选择取决于数据类型与题目要求，AQA 的技能清单列出的图形包括折线图、条形图、饼图、象形图、等组距直方图、分段条形图、散点图与人口金字塔，地图类还包括等值线图、点状图、愿望线与比例符号。等组距直方图适合连续的粒径或坡度数据，条形图适合街区之间的人数比较，比例符号能在一张底图上同时表示位置与数量。选完图要写理由，写清这种图为什么能把你的数据讲明白，考场里被问为什么某种呈现方式适合题目给出的数据，答案就落在这里。学生最常把比例符号与象形图混用，前者用图形面积表示数量，后者用重复的小图标计数，画反了会把数量级讲颠倒。图表的标题、单位与图例缺一个都会被扣分，坐标轴刻度也要等距，缺刻度的散点图很难读出趋势。
Choosing a presentation method follows from the data type and from what the question wants shown. AQA lists line charts, bar charts, pie charts, pictograms, histograms with equal class intervals, divided bars, scattergraphs and population pyramids, plus choropleth maps, isoline maps, dot maps, desire lines, proportional symbols and flow lines. Proportional symbols show location and quantity together on one base map, while pictograms repeat a single icon, and confusing the two reverses the scale of the data.
统计技能：中位数四分位距与百分比变化
AQA 的统计技能条目要求集中趋势与离散程度里的少数几种，中位数、均值、极差、四分位与四分位距、众数与众数组，加上百分比增减与百分位的理解。野外数据通常夹着几处极端值，中位数比均值更能代表一条街的典型状况，四分位距又能说明中间一半数据的分散程度。算四分位距先把数据从小到大排列，找到下四分位与上四分位再相减，样本少时按位置公式取整数位置的数据即可，学校层面不必纠结插值细节。百分比变化用新旧差值除以旧值，回收率与客流量的变化都这样算，写成绝对值会失去比较意义。相关性在 GCSE 只要求用散点图与最佳拟合线描述，画趋势线、做预测、内插与外推，拟合线不该为了穿过原点而扭曲，也不该忽略一处明显的离群点。阅卷时常见的错误是把均值当默认选择，分布明显偏斜时仍然用均值，结论就会被几个极端读数拉走。
The statistical skills AQA requires are narrow and specific: median, mean, range, quartiles and interquartile range, mode and modal class, plus percentage increase or decrease and the idea of percentiles. Fieldwork data usually contain a few extreme readings, so the median describes a typical street better than the mean, and the interquartile range shows how tightly the middle half of the data sits. Correlation at GCSE goes no further than sketching trend lines on scattergraphs, drawing an estimated line of best fit and using it to interpolate or extrapolate; a line forced through the origin, or one that ignores an obvious outlier, misleads the reader.
异常值与散点趋势线的解释
异常值不等于坏数据，处理方式要在方法或评估里交代。判断异常值可以看读数是否偏离四分位距的一点五倍这一常用界线，也可以判断它是记录错误还是真实测量，抄错就回原始表格改，真实读数就保留并解释，例如某一采样点正好在排水口下方，垃圾密度自然偏高。AQA 把识别野外数据中的异常值列为分析过程的一条，考场上写出异常值并给出成因，比假装它不存在更容易进档。散点图上的趋势线要写成描述性的句子，说清正相关或负相关、关系的强弱以及个别偏离的点，不要只写图上有一条线。常见的错误是把相关写成因果，把商铺数量多与噪音高说成前者造成后者，加入步行人流这个第三变量后关系就会变化。错得最多的是把坡度角度与坡度百分比混用，四十五度与百分之四十五是两件事。
Outliers are not automatically bad data, but the way they are treated has to be explained in the method or in the evaluation. A reading that sits well beyond the interquartile range may be a transcription slip that should be corrected against the raw sheet, or a genuine measurement near a drain that should be kept and interpreted. AQA lists the identification of anomalies in fieldwork data as a required strand, and a trend line is best described by naming its direction, its strength and the points that break the pattern.
结论要落回原始假设
结论要回到调查最初提出的题目与假设，用数据回答它成立还是不成立，而不是把结果重新描述一遍。写结论的固定动作是引用具体数字，例如上游三个点位的平均粒径十二毫米，下游三个点位平均六毫米，粒径向下游递减的判断成立。假设被数据否掉同样是完整结论，许多高分答卷正是靠不成立这个词展示理解，硬把结果说成支持假设，反而不如承认偏差来得可信。结论与数据之间要能一步步对上，中间缺少的那一步就是阅卷时会指出的跳跃，例如从噪音评分低直接跳到社区很安静，中间还缺了评分尺度与时段这两个条件。结论的适用范围也要收住，一次调查只覆盖两条街与两个时段，写成整个城区都如此就超出了证据。审题时看清题目问的是你的调查还是资源册里的陌生调查，两种问法的落笔方式不同，前者用自己数据的数字，后者只用资源册给的信息。
A conclusion has to answer the question the enquiry began with, using the data to say whether the idea holds rather than restating the results. Quoting figures commits the answer, such as mean sediment size falling from twelve millimetres at three upstream sites to six millimetres downstream, which supports the expected trend. A hypothesis that the data reject can still earn full credit, because the rejection has to be justified from the same figures; inventing support for an idea the numbers contradict is where answers lose their credibility; examiners reward the phrase the data do not support, when the numbers really do not.
评估局限与考场写作口径
评估这一条要写四件事，数据收集方法的问题、数据的局限、还可以补什么数据、结论可靠到什么程度，AQA 的考纲把四条都列在调查过程的最后一条里。写局限要具体到你的操作，例如问卷在雨天只收到十四份、河床粒径只量了长轴，这些直接影响可靠性。考场口径上有一条容易忽略的规则，Paper 3 野外调查部分没有 AO1 得分点，考生不会被要求描述自己的数据收集方法，也不会被要求解释自己为何选某种呈现方式，能被问的是为什么某种呈现方式适合题目给出的数据。不熟悉的调查情境那道题以 AO4 为主，AQA 官网写明它由 12 分 AO4 与 3 分 AO3 组成，自己的调查那道题以 AO3 为主。Edexcel 的 1GA0/03 是 64 分、占 25%、1 小时 30 分钟，卷面分物理调查 18 分、人文调查 18 分、UK Challenges 28 分，Pearson 官方手册写明其中最多 4 分给拼写标点语法。OCR 的 J383/03 是 80 分、1 小时 30 分钟、占 40%，野外调查题全部集中在这一卷。
The evaluation strand asks for four things: problems with the data collection methods, limitations of the data, other data that might have been useful, and how reliable the conclusions are. Specific limitations such as fourteen questionnaires collected in rain, or sediment measured on the long axis only, carry more weight than a general comment about human error. One exam rule catches many candidates out, because the fieldwork section of Paper 3 carries no AO1 marks, so nobody will be asked to describe their own methods, while the unfamiliar enquiry question is built from 12 AO4 marks and 3 AO3 marks.
CASIO fx-991CN X 科学计算器 考试必备¥168📱 去京东购买京东精选16621398022 同微信
A-Level / GCSE / IB / KS2-KS3 全科辅导，地理野外调查与数据分析专项答疑
预约咨询 aleveler.com
或截图后用「扫一扫 → 相册」选择本图