ENV 100 Chap.11 Earth Science Inquiry and Communication
Earth Science Inquiry and Communication
This chapter unifies the five laboratory skills: observation, experiment, data, analysis, and communication. It turns field notes into measurements, plots, models, classifications, and audience-first explanations. Grain-size conversion, settling velocity, trendlines, R-squared, sorting, beach state, units, captions, uncertainty, and limitations are treated as one evidence workflow.
What this chapter covers
- 01
Observation and inference
- 02
Experimental controls
- 03
Data quality and units
- 04
Scatter plots and trendlines
- 05
Beach-state classification
- 06
Audience-first communication
Worked example · free
Compare two fitted relationships
- +1Identify the stronger fitted linear relationship.
- +1Name both variables and their units.
- +1Avoid a causal claim from fit alone.
- +1Check outliers, range, and process plausibility.
Key terms
- Controlled variable
- A controlled variable is held constant or accounted for so its influence does not confound a test.
- Trendline
- A trendline is a fitted model that summarises a relationship between plotted variables.
- R-squared
- R-squared describes the share of variation represented by a fitted model within the analysed data.
- Sorting
- Sorting describes the spread of sediment grain sizes within a sample.
- Figure caption
- A figure caption states what is shown and the relationship the reader should notice.
Earth Science Inquiry and Communication FAQ
What is the difference between observation and inference?
An observation records what was seen or measured. An inference proposes the process or history that explains it. Keeping them separate makes it possible to test alternative interpretations.
What must a scatter plot include?
Use the appropriate variables on labelled axes with units, a readable scale, and a caption that states the tested relationship. Inspect points before relying on a trendline or summary statistic.
How should science be communicated to a public audience?
Lead with the key finding, explain why it matters, support it with the most relevant evidence, and acknowledge uncertainty. Replace jargon carefully without changing scientific meaning or overstating confidence.
What can R-squared establish?
For a stated fitted model and dataset, it summarises how much variation is represented by that fit. It does not identify cause, guarantee prediction, validate the chosen model form, or remove the need to inspect range, outliers, units, and process plausibility.
Why should units be checked first?
A numerically correct substitution can still produce an invalid result when quantities use incompatible units. Recording units beside raw data, converting before calculation, and checking the final dimension makes the analytical chain auditable and helps reveal transcription or scale errors.
Assessment move
Treat the five laboratory themes as one inquiry cycle rather than separate assignments. Start with observation and make a two-column record of what was directly seen or measured and what was inferred. Convert one inference back into a testable question. For experiment design, identify the factor changed, the response measured, variables held constant or recorded, replication, and a source of uncertainty.
For data work, create a unit ledger before calculating. Put every measurement, original unit, required unit, conversion, and plausible range in one place so centimetres per second cannot enter a relationship expecting metres per second unnoticed. Build scatter plots by deciding which variable belongs on each axis, labelling units, choosing scales that show the data, inspecting points, and only then fitting a model.
Interpret slope direction and R-squared in words tied to the variables. Never use fit as a synonym for causation. Check outliers, restricted range, non-linearity, measurement error, and an Earth-science process that could plausibly connect the variables. For grain-size and beach-state work, keep raw measurements, transformed values, sorting, fall velocity, wave values, and classification steps traceable.
Compare the numerical classification with photographs and field observations instead of forcing the field evidence to agree. Practise captions as mini-arguments: identify the data, describe the relationship, and state the intended interpretation without overstating it. Communication revision begins by defining the audience and the single finding they need.
Draft the conclusion first, add the most relevant evidence, explain why it matters, and close with a limitation or next measurement. Replace jargon only after preserving the scientific distinction. Retrieval sessions should rotate among designing a test, auditing a plot, checking a calculation, evaluating a model, and translating a result for a public audience.
Before submission or a test, use a final six-part audit: variables, axes, units, model, uncertainty, and claim boundary. If any number appears without site, unit, comparison, or context, revise the sentence. The last question is always whether the available observations genuinely support the conclusion at the spatial and temporal scale claimed.
