Fieldwork: analysis, conclusions and evaluation

邊玩邊學

回答這些題目賺取能量，接著就能釣魚、探索。不需要帳號。

給教育者：為 Fieldwork: analysis, conclusions and evaluation（Geography、Geographical Skills）準備好可直接使用的課程投影片, 複習筆記——用於你的課程，或把這個主題當成互動班級活動，讓學習者以即時遊戲的方式進行。

課程筆記

Describe direction, strength, shape and unusual observations using figures and units. A scatter graph pairs two measurements for each observation. Do not join points as if they were a time sequence or assume a straight best-fit line suits every pattern.
Anomalies can reflect genuine local factors, random variation, instrument faults or recording errors. Check the evidence, retain an audit trail and explain any exclusion. Do not invent a cause when it is unknown.
Spearman’s rank coefficient describes the direction and strength of monotonic association in ranked paired observations. Coefficient and significance are distinct, with significance depending on sample size and an appropriate test. Neither establishes causation. Requirements and permitted techniques vary by exam board.

Answer the original enquiry and address a hypothesis if one was set. Give specific figures and units, discuss the main pattern and exceptions, then state the scope and uncertainty of the judgement. Evidence can support, challenge or leave a question unresolved.
The attached conclusion example uses an explicitly invented river dataset, not actual field results. Wider and narrower observations may support a general downstream pattern while showing local variation; discharge, geology, tributaries, management and other processes require further evidence. Correlation alone cannot isolate a cause.

Reliability concerns consistency under comparable repeated conditions. Accuracy concerns closeness to a true or reference value, which may not be known exactly in the field. Validity concerns whether the selected methods and evidence answer the enquiry. Consistency does not guarantee accuracy or validity.
Separate random variability, systematic error and genuine environmental change. A tape with a calibration problem may be consistently wrong. Repeats can reveal spread, but checking equipment and definitions addresses different weaknesses. Specify the suspected error and its effect rather than writing only human error.
A larger sample and additional well-chosen sites can improve coverage and estimates, but do not automatically remove selection bias, non-response or instrument faults. Define the target population, area, times and relevant characteristics before judging representativeness.
Evaluate each practical limit in context: restricted access may omit a river section; one weekday count may miss weekend use. Match improvements to those omissions while retaining equal durations, clear boundaries, safe methods and comparable conditions where required. Secondary evidence also needs checks of date, location, method, definition and purpose.

Present a traceable chain from question through method, data, analysis and conclusion. Label tables and axes with variables and units, distinguish measured observations from invented examples, and cite secondary sources with their dates.
Evaluation should explain how a weakness could affect the inference and what feasible change would address it. Do not claim certainty just because a graph looks neat, or dismiss all findings merely because the enquiry has limitations.

Invented counts of 18, 24 and 30 in equal ten-minute intervals have mean 24: add to obtain 72 and divide by three. The unit is people per ten-minute count, not a percentage or a daily total.
Invented widths of 2.1, 2.8, 3.5 and 3.2 m have range 1.4 m, found from maximum minus minimum. For the ordered counts 10, 11, 12, 13 and 104, the median is 12 while the mean is 30. This shows sensitivity to an extreme value, not proof that the value should be discarded.

Step through every slide for this topic — plus flashcards and revision notes — with a free account.

免費預覽——37 題中的 8 題。註冊即可查看全部。

這個主題的歷屆試題練習即將推出。

即將推出