Use with the 24-card deck. Supported selections are evidence checks, not a score. Accept equivalent observations. Free writing is not machine graded. A supported inference is not always a directly visible fact.
Eight full mini-lessons and text alternatives · Student cards · Source key
Describe what is visible before naming a process.
Supported: A, B. The sketches illustrate a fin, a window and a hoodoo. They are shape examples, not measured Bryce landmarks.
Accept also: ‘There is rock above the opening’ or ‘A is wider than C in the drawing’ are also defensible visible observations.
Limit: No real dimensions, dates or future condition can be read from these silhouettes.
Discuss: What would another viewing angle add?
Supported: A, D. Apparent outline depends on viewpoint. Multiple views can test whether a narrow outline belongs to a pillar or a wall.
Accept also: ‘View B hides one dimension’ is a useful alternative; no numerical length is required.
Limit: This is an original perspective exercise, not two photographs of a named landform.
Discuss: Which extra view would help you distinguish a fin from a pillar?
Sources: G02
Supported: B, D. NPS treats the named Natural Bridge as an arch/window and notes that process terminology can be imprecise.
Accept also: You may call the drawn feature an opening or window without claiming its origin.
Limit: Neither the drawing nor the name supplies a process history or dimensions for the real feature.
Discuss: Which part of your answer came from looking, and which from reading?
Sources: G03
Order, color and rock identity are different clues.
Supported: A, D. This summarizes the informal member order in the NPS wayside, not a complete section at every location.
Accept also: ‘No scale is given’ is a valid evidence observation.
Limit: Do not infer equal thickness, universal preservation or a contact from color alone.
Discuss: What could a measured section add that these boxes cannot?
Supported: A, C. Lighting and material are different explanations for color differences. Unit identification uses more than color.
Accept also: ‘The edge does not shift where the shading changes’ is a useful alternative observation.
Limit: The diagram is not a mineral analysis or a mapped Claron exposure.
Discuss: What would you record besides color?
Sources: G07
Supported: A, D. Superposition gives relative order under the stated conditions. It does not assign numerical ages.
Accept also: ‘C was deposited after B’ is equally valid under the same assumptions.
Limit: An original generic section, not a dated Bryce stratigraphic column.
Discuss: What evidence might help place a numerical age on a layer?
Sources: G05
Trace a weakness without inventing its history.
Supported: A, B. The drawing is consistent with a fracture without shown displacement. Absence of visible offset in one view does not prove no movement ever occurred.
Accept also: ‘We need another view or a displaced marker to test movement’ is defensible.
Limit: No real fracture is mapped or dated; out-of-plane movement is not resolved.
Discuss: What would make a fault interpretation stronger?
Supported: A, D. Connected cracks can provide access for water. A two-dimensional drawing cannot settle all three-dimensional pathways.
Accept also: ‘Water access to B is unknown’ is valid; ‘B is always dry’ is not established.
Limit: Arrows show an illustrative possibility, not measured infiltration or a real cave system.
Discuss: How would you express the unknown without filling it in?
Sources: G02
Supported: A, C. Water availability and rock thermal conditions matter. Experimental frost-fracture research also includes ice segregation, beyond a simple event counter.
Accept also: ‘There is no observed crack change’ is a valid evidence-limit statement.
Limit: No climate series, damage count or erosion rate is supplied.
Discuss: What would you measure to test whether the rock actually changed?
Sources: G08
Distinguish breaking, dissolving and moving.
Supported: A, B. The panels distinguish physical weathering from material transport.
Accept also: ‘Both could happen in the same landscape’ is a useful process interpretation, not a pictured event history.
Limit: Drawn arrows do not encode force, distance or elapsed time.
Discuss: Could a single rain event involve both kinds of action?
Sources: G01
Supported: A, D. Carbonate dissolution can remove material without a visible pebble leaving the rock.
Accept also: ‘This needs a chemical explanation, not just a shape description’ is defensible.
Limit: No pH, concentration, reaction speed or measured chemistry is represented.
Discuss: Why might a before-and-after photograph miss some material loss?
Sources: G05
Supported: A, C. Gravity is one plausible contributor; the arrangement alone does not identify each fragment’s source or travel history.
Accept also: ‘The pieces could have accumulated in multiple events’ is possible but not demonstrated.
Limit: This is not a rockfall forecast or an access recommendation.
Discuss: What information would separate one event from many?
Sources: G03
Work with a small, actual Sunset Point data slice.
Supported: A, B. The difference is a calculation on an existing data derivative, not a new field measurement or change detection.
Accept also: ‘The east end is about 148 m lower’ is equivalent; ‘the whole line is downhill’ is not true of every segment.
Limit: 2018 source, 4 m exported spacing; an upper surface with filtering/interpolation limits.
Discuss: Why is an elevation difference different from an erosion measurement?
Supported: A, B. The derived endpoint grades are about 12.4% downward for A and 65.5% for B. These are profile calculations, not measured trail grades.
Accept also: ‘B drops roughly five times as much over the same horizontal span’ is an acceptable approximate comparison.
Limit: Local slope, accessible tread and stability cannot be inferred from this two-endpoint comparison.
Discuss: What changes if the spans have different horizontal lengths?
Sources: D01
Supported: B, C. Topography and stratigraphy answer different questions. The section does not justify invented underground bands.
Accept also: ‘No rock type is identified by the surface height alone’ is valid.
Limit: The area beneath the profile is unknown in this representation, not an inferred cavity.
Discuss: What source would you seek before drawing a contact?
Ask what a dataset can—and cannot—show.
Supported: C, D. The integrated study’s upper-envelope method cannot recover a complete arch or ground-level scan.
Accept also: ‘More triangles alone cannot provide an unseen underside’ is a valid inference.
Limit: Generic diagram only; no replacement model of Thor’s Hammer or Natural Bridge is made.
Discuss: What evidence would you request before filling a hidden surface?
Sources: D01
Supported: A, B. Spacing describes the exported representation. Centimetre encoding and a dense grid do not certify source accuracy.
Accept also: ‘129 is a node count, not an interval count’ is a useful distinction.
Limit: The number line depicts exported sample geometry, not individual laser footprints.
Discuss: Which metadata would you read to learn about actual accuracy?
Sources: D01
Supported: A, D. A composite can combine evidence from different times. It does not establish what changed between those times.
Accept also: ‘The geometry is older than the color imagery’ is defensible; an exact six-year interval is not established by year labels alone.
Limit: Metadata exercise; no aerial photograph is copied and no present condition is claimed.
Discuss: Why should a model keep separate geometry and imagery dates?
Sources: D01
A sequence is not a clock.
Supported: A, D. The age of a rock is not automatically the age of the landform made from it.
Accept also: ‘The shape developed after the material in this model’ is acceptable without adding a date.
Limit: This is a teaching sequence, not a dated reconstruction of one Bryce hoodoo.
Discuss: Which event does the phrase ‘How old is it?’ actually refer to?
Sources: G01
Supported: A. The NPS description supports this broad process story, while the original panels omit alternate histories and timing.
Accept also: ‘A connection is lost between the second and third sketch’ describes a visible diagram change.
Limit: No duration, inevitability or collapse prediction is supplied.
Discuss: How would you label this figure to avoid implying a countdown?
Sources: G03
Supported: B, C. A single snapshot cannot provide an erosion rate. A comparison would still need uncertainty analysis and process interpretation.
Accept also: ‘No change rate can be calculated from this card’ is a complete valid conclusion.
Limit: No second survey, erosion value or new acquisition is supplied.
Discuss: What would make two surveys meaningfully comparable?
Sources: D01
Join a clue, a possible cause and a useful next question.
Supported: A, B. Differing resistance can contribute to ledges and recesses; the drawing does not identify a rock type or cause uniquely.
Accept also: ‘The change could also reflect fracture geometry’ is a reasonable alternative hypothesis, not a confirmed finding.
Limit: Illustrative shape, no sampled bed or strength measurement.
Discuss: What would help distinguish material differences from fracture effects?
Supported: A, B. Choose evidence that addresses the question. A precise-looking graphic does not automatically supply the missing information.
Accept also: ‘None of these sources alone establishes exact collapse timing’ is a valid conclusion.
Limit: A geological map also has scale and uncertainty limits; it is not a guarantee of any specific contact at any zoom level.
Discuss: Which source would you choose for a question about a hidden underside?
Supported: A, C. A careful caption combines the observation, evidence class and an honest limit. Several wordings can do this well.
Accept also: ‘The east endpoint is lower in this model; this is not a current survey’ is also defensible.
Limit: Your own wording is not automatically graded; discuss whether each clause is supported.
Discuss: Write a different caption that includes one observation and one limit.