Common Misconceptions

⚠️ Common Misconceptions

Misconception 1: "Random sampling means I choose randomly — just picking whatever looks typical"

Reality: Random sampling means that every possible sample has an equal chance of being selected — which means you must actively remove your own judgement from the process. If you pick the pebbles that "look like a good selection," you are introducing unconscious selection bias. True random pebble sampling means reaching into the river without looking and picking up whatever your hand touches first, regardless of size, shape, or colour. This prevents you from oversampling the "average" pebbles and missing the extremes that might be important for your mean calculation.

Misconception 2: "The float method gives the actual velocity of the river"

Reality: The float method gives the surface velocity only. Rivers have a velocity gradient: the water moves fastest at the surface and towards the centre of the channel, and slowest near the bed and banks where friction is greatest. The surface velocity is typically 10–25% faster than the mean velocity of the whole cross-section. This means that if you use the float method, your velocity figures will be an overestimate of true mean velocity. A flow meter held at 0.6 of the total depth (the 60% depth method) gives a more accurate estimate of mean velocity across the full column. In an evaluation, acknowledging this limitation of the float method and suggesting the flow meter as an improvement will score high marks.

Misconception 3: "Velocity always increases downstream — if mine doesn't, the data is wrong"

Reality: The Bradshaw Model is a general prediction based on average conditions across many rivers. Real rivers are affected by a wide range of local factors that can disrupt the pattern at specific sites. A waterfall will pool water and slow velocity immediately above it. A tributary joining just upstream of one site will temporarily spike discharge and velocity, creating an anomalously high reading at that site before velocity equalises further downstream. A gorge (narrow canyon) will speed up the flow at that point regardless of its position on the long profile. If your data does not perfectly match the model, this does not mean your data is wrong — it means you have found a real geographical anomaly that needs to be explained. This is far more interesting than data that perfectly fits the model, and it demonstrates your geographical thinking if you can explain why the pattern deviates.

Misconception 4: "More measurements always make the investigation better"

Reality: More data can improve reliability — but only if the additional data is collected using the same rigorous method as the original data. Adding extra sites in inaccessible locations where you have to rush measurements, or adding more pebble samples late in the day when the whole group is cold and tired and making subjective roundness judgements, can actually introduce more error than it removes. Quality of method is more important than quantity of data. A well-collected, well-documented set of 5 reliable sites will always outperform 10 sites where the method was inconsistent.