Wait, what? A wildlife map shows six GPS dots from a tagged otter. A neat blue line joins the dots. One segment cuts straight across a wooded hill and then across a pond. A student traces the line with a finger and says, “So the otter walked exactly along this route.”
The map looks like a path, but it contains two different kinds of information: recorded locations and a representation connecting those locations. Unless a tracking system recorded the animal continuously at infinitely small intervals — which it did not — the line between two fixes is not direct observation of every metre the animal travelled.
Quick answer
A GPS or telemetry track usually consists of location fixes recorded at particular times. Software often joins consecutive fixes with straight line segments so the sequence is easy to see. Those segments can be useful for estimating displacement and visualising movement, but they do not prove that the animal moved in a perfectly straight line, visited every point on the line, moved at constant speed, or stayed on that path between fixes.
The learner job is simple to state and surprisingly powerful: separate the measured dots from the drawn journey between them.
This Reality Lab owns the line-between-fixes problem
This page does not re-teach coordinate systems, distance calculations, graph reading, sampling, animal behaviour or GPS engineering as standalone topics. It applies existing PSLE Science habits to a real scientific communication object: an animal-movement map whose connecting line can look more certain than the measurements underneath it.
For the distinction between observation and inference, route to How to Tell Observation, Inference, Prediction and Explanation Apart in PSLE Science. For sampling and ecological evidence, use Sampling, Field Surveys, Populations & Ecological Evidence for PSLE. For a separate GPS-quality communication object, see Reality Lab Vol.456 on HDOP.
Original field file: six fixes, one tempting story
Use this invented set of tracking data. The animal, locations and numbers are fictional.
| Fix | Time | Map position | What was actually recorded? |
|---|---|---|---|
| A | 08:00 | River bank | A location estimate at about 08:00 |
| B | 08:30 | Forest edge | A location estimate at about 08:30 |
| C | 09:00 | North clearing | A location estimate at about 09:00 |
| D | 10:00 | South bank | A location estimate at about 10:00 |
| E | 10:30 | Reed bed | A location estimate at about 10:30 |
| F | 11:00 | River bend | A location estimate at about 11:00 |
The mapping program joins A to B, B to C, C to D and so on. The segment from C to D crosses a steep hill on the map. Does that prove the animal climbed directly over the hill? No. The available observations tell us where the animal was estimated to be at C and D. Many paths between those points remain possible.
Dots are observations; lines are a representation
Movebank, a scientific platform for animal-movement data, describes displacement metrics using straight-line distances between subsequent locations. That is a calculation on recorded fixes. It is useful because it is defined and repeatable. But a straight-line displacement is not the same thing as the exact route travelled.
Imagine leaving home at 08:00 and being at school at 08:30. A map can draw a straight line between home and school. That line gives a direct point-to-point displacement. It does not reveal whether you took a curved road, walked around a building, crossed a bridge or stopped at a crossing. The endpoints are evidence; the detailed route requires more information.
Four layers hiding inside one track map
- Physical movement: the animal’s real continuous journey through space and time.
- Measurement events: the moments when a tag or tracking system obtains location information.
- Processed locations: coordinates retained after data handling, quality checks or filtering.
- Displayed track: dots, lines, symbols and map layers used to communicate the data.
A viewer may see only layer 4 and accidentally treat it as layer 1. Reality Lab reasoning works by asking how each layer was produced.
The time-gap test
Whenever you see a movement line, check the time between consecutive fixes. A five-second gap and a six-hour gap do not contain the same amount of unobserved opportunity. During a long gap an animal could turn, rest, forage, shelter, circle back or take a route around an obstacle before arriving at the next recorded location.
This is not an argument that longer gaps make the dataset useless. It is a scope rule. The coarser the temporal sampling, the less detail the data can directly support about what happened between observations.
Worked case 1: the line crosses a lake
A deer has Fix P west of a lake at 12:00 and Fix Q east of the lake at 16:00. The straight segment crosses open water. A viral caption says, “GPS proves the deer swam 2 km across the lake.”
The map proves neither that exact route nor swimming from the two endpoints alone. The deer could have walked around the shore, crossed a narrow northern section, used a causeway, or moved along another route not sampled by the fixes. To strengthen the swimming claim, look for intermediate fixes, sensor data, direct observations, shoreline constraints or a much finer track showing entry and exit points consistent with a swim.
Worked case 2: a fence creates a hidden detour
A small mammal is recorded at two points only 300 m apart, but a fence lies between them. The straight line passes through the fence. The animal may have travelled 700 m to reach a gap and then returned toward the second point. If we call 300 m “the distance travelled”, we have silently replaced route length with endpoint displacement.
The distinction matters whenever a claim uses energy, speed or effort. An animal can have a small displacement while travelling a much longer path.
Worked case 3: the one-hour gap that hides a loop
At 14:00 a bird is at Point R. At 15:00 it is only 100 m away at Point S. A student writes, “The bird travelled only 100 m during the hour.” That is one possible minimum-like point-to-point description, but not necessarily the travelled distance. The bird could have flown a 2 km loop and ended near its starting location.
When a movement dataset samples positions rather than continuous route length, the geometry of observed fixes constrains the claim. Do not make the data describe motion they never recorded.
Worked case 4: the impossible jump
A tracking map suddenly shows one location 80 km away and the next fix returns near the previous cluster. Is this proof of an extraordinary sprint? Not yet. Movebank’s quality-control guidance notes that movement data can contain outliers and location-quality problems, and its filtering guidance asks researchers to consider plausible speeds, sampling intervals and location error.
The scientific response is neither “the animal definitely teleported” nor “delete every strange point”. It is to inspect the record, tag behaviour, timestamps, location quality, deployment history and biological plausibility before deciding how to treat the point.
Worked case 5: three fixes make two perfect straight lines
A turtle is recorded at A, B and C, each 12 hours apart. On the display the three points form a sharp V. A caption says, “At B the turtle turned exactly 72 degrees.” That angle belongs to the line segments joining the sampled positions. It is not automatically the exact turning angle of the turtle’s body at one instant.
The turtle may have curved gradually, wandered near B, or changed direction several times between fixes. A geometric property of the displayed polyline can be valid as a property of the representation while still being too strong as a claim about the animal’s continuous behaviour.
Location error adds another layer
Even the dots deserve care. A GPS fix is an estimate of location, not a magically exact point with zero uncertainty. Buildings, vegetation, terrain, signal geometry, tag design and other conditions can affect location quality. Movebank explicitly lets researchers consider location-error estimates when filtering movement data.
If two consecutive fixes are only a few metres apart and each has substantial location uncertainty, a small apparent move could partly reflect measurement error. If fixes are kilometres apart, the same location uncertainty may be less important for the broad movement pattern. The effect of uncertainty depends on the size of the claim.
Missing fixes are not invisible straight travel
Tracking devices sometimes fail to obtain a fix. Batteries, signal conditions, tag settings or data transmission can create gaps. A line drawn across a gap can make the missing interval visually disappear. That is dangerous because the smooth line looks like information even when the underlying record is sparse.
A good map should make the sampling structure recoverable. At minimum, the reader should be able to tell where fixes occurred and, ideally, see timestamps or know the fix interval. If a long gap is important to the claim, inspect the underlying data rather than trusting a continuous-looking stroke.
Representation check: maps add meaning through design
Line width, map scale, projection, colour, smoothing and symbol size all affect perception. A thick line can cover roads, rivers or barriers beneath it. A zoomed-out map can make a curved route appear almost straight. A smoothed track can suppress sharp turns. A map that omits timestamps can hide long gaps.
Ask not only “Is the data real?” but also “What did the display do to help me see it?” A good representation makes patterns visible. The learner’s job is to avoid confusing that helpful representation with additional measurements.
Baseline and comparison check
Suppose two animals have tracks of similar visible length on two maps. Before deciding that they travelled similar distances, check:
- Are the maps at the same scale?
- Were fixes recorded at the same interval?
- Were missing data handled in the same way?
- Were both tracks filtered with comparable rules?
- Does “distance” mean sum of straight-line steps, modelled path length or another quantity?
- Are the dates and durations comparable?
Visual similarity does not guarantee measurement comparability.
What evidence strengthens an exact-route claim?
- Very frequent, high-quality location fixes relative to the scale of movement.
- Independent sensors or observations consistent with the route.
- Known physical constraints, such as a narrow corridor, that limit possible paths.
- Clear handling of outliers and missing fixes.
- A method that explicitly estimates a continuous path, together with its assumptions and uncertainty.
Even then, “exact” should be used carefully. More data can narrow the set of plausible paths without turning every metre of movement into a direct observation.
What weakens the claim?
- Long intervals between fixes.
- Many failed or missing fixes.
- Large location uncertainty relative to movement.
- Obstacles that make the straight segment physically implausible.
- Unknown filtering or smoothing.
- A claim about travelled distance based only on endpoint displacement.
- A screenshot that hides timestamps, scale or raw points.
How far can the conclusion travel?
With two reliable fixes, you can say the animal was estimated at Location A at one time and Location B at another. You can calculate a straight-line displacement between them. With many reliable fixes, you can describe an observed sequence of locations and estimate movement characteristics appropriate to the sampling interval. With suitable models and assumptions, you may infer more about the likely route.
But the map does not earn a continuous exact path merely because the software drew a continuous line.
Tempting reasoning that fails
- “The line is on the scientific map, so every point on it was measured.” The line may only join measured fixes.
- “The animal moved 500 m because the endpoints are 500 m apart.” That is displacement between endpoints, not necessarily route length.
- “A sharp corner in the polyline proves a sharp turn.” The corner can be created by sparse sampling.
- “A strange point proves strange behaviour.” Check for location error, timestamp problems and outliers first.
- “A smooth-looking track means continuous measurement.” Display smoothing can conceal the underlying sampling interval.
Original PSLE-style transfer case: the mangrove bird
This is original practice, not a copyrighted examination item.
A tagged mangrove bird produced four valid GPS fixes at 07:00, 07:30, 08:00 and 09:30. The mapping program joins the fixes with straight segments. The last segment crosses open water, but a chain of small islands lies nearby. A student writes: “From 08:00 to 09:30 the bird flew in a perfectly straight line across open water.”
Question 1: Which evidence is directly observed by the tracking record?
Explained answer: The valid GPS fixes provide location estimates at the stated times. The straight line between the 08:00 and 09:30 fixes is a representation connecting those observations.
Question 2: Give one alternative route consistent with the endpoints.
Explained answer: The bird could have flown via one or more nearby islands before reaching the 09:30 location.
Question 3: What evidence would strengthen the straight-flight claim?
Explained answer: Additional reliable fixes at short intervals along the segment would show whether the bird remained close to the straight route.
Question 4: Why is the 90-minute gap especially important?
Explained answer: A longer unsampled interval leaves more possible movements between observed endpoints, so the exact route is less constrained.
Delayed independent return
- What is the difference between a GPS fix and the line joining two fixes?
- Why can straight-line displacement underestimate travelled distance?
- What does a longer fix interval do to route certainty?
- Why should a surprising location be checked before interpreting surprising animal behaviour?
- Name two display choices that can make a track look more continuous than the measurements are.
Self-check: Fixes are recorded location estimates; connecting lines are representations; curved or looping travel can exceed displacement; longer intervals hide more possible movement; outliers and location error can imitate behaviour; smoothing, thick lines, zoom level and hidden timestamps can change perception.
Explained practice: three tiny datasets
Dataset A: fixes every 10 seconds around a small enclosure, with low location error. Dataset B: fixes every 6 hours across a forest. Dataset C: fixes every minute, but 40% of fixes are missing in a canyon.
Which dataset best supports fine-scale route claims? Usually A, because the temporal gap is small relative to the movement scale and location quality is good. B may support broad movement between regions but not detailed path reconstruction. C looks frequent in its settings, yet missing fixes in the difficult environment create gaps that must remain visible in the interpretation.
The point is not to rank every dataset by one number. Fitness depends on the learner question.
Parent and tutor guide: make the hidden path visible
Place six sticky notes on the floor as “GPS fixes”. Ask one child to walk from note to note however they like while another child looks away and records only when the walker reaches each note. Then draw straight lines between the notes. Compare the drawn path with what the walker actually did.
Repeat with fewer fixes. The learner will see that the same endpoints can hide many routes. Then add a chair as an obstacle. A straight line may now pass through the chair even though the walker went around it. That physical experience makes the representation problem memorable without teaching a formula.
In a three-student tutorial, assign roles: tracker records positions, animal walks an unseen route, and map reader tries to infer the route. Rotate. Each round should end with one sentence beginning, “The evidence directly shows…” and another beginning, “A possible inference is…”
Authoritative sources and curriculum frame
- SEAB — 2026 PSLE Science syllabus: includes interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning.
- MOE — 2023 Primary Science Teaching and Learning Syllabus: emphasises healthy scepticism, objectivity and evidence-based reasoning.
- Movebank — movement traits: documents straight-line displacement calculations between recorded locations and distinguishes movement summaries built from fixes.
- Movebank — tracking-data quality control: describes checking mapped tracks, deployment timing, outliers and data problems.
- Movebank — general data filters: discusses plausible speed, sampling interval and expected location error when assessing tracking data.
- Movebank Attribute Dictionary: documents tracking attributes such as GPS fix interval.
The quiet habit to keep
A line can be truthful as a drawing and still be too strong as a story. On a movement map, ask: Which points were measured? When? What happened between them that the data did not directly observe? Once you keep dots and journeys separate, the map becomes more useful, not less.
