Plan and carry out investigations: identify the independent, dependent and controlled variables, use a control, and explain why repeated trials and a large sample make results more reliable (Virginia 2018 Earth Science SOL ES.1).

What this topic is asking

Virginia Earth Science SOL standard ES.1 asks you to plan and carry out investigations. On the EOC this almost never appears as "define independent variable" in isolation; instead an item describes an Earth science investigation (a stream table, dissolving salt, the cooling of a rock, the angle of sunlight) and asks you to identify the variables, spot the control, or judge whether the design is fair and reliable. Getting the variables right is the single most tested investigation skill, and it underpins many items in the other categories.

The three kinds of variable

A clean way to find them in an EOC question is the phrase "how X affects Y": X is the independent variable (the cause you change), and Y is the dependent variable (the effect you measure). Everything else named in the setup should be a controlled variable. For example, "how the angle of a light source affects the temperature of a surface" makes angle the independent variable and temperature the dependent variable, with the same lamp, the same surface and the same distance held constant.

Why you change only one variable at a time

If you changed two things at once, say the slope and the type of sediment in a stream table, and the erosion changed, you could not tell which change caused it. By holding everything constant except the independent variable, you can be confident that any change in the dependent variable was caused by the independent variable. This is what makes a test fair and lets you draw a valid conclusion about cause and effect.

The control group

Reliability: repeated trials and sample size

A single measurement can be thrown off by random error (a misread instrument, a stray gust of wind, an unusual sample). To guard against this you run repeated trials and use a large sample, then take the average (mean). The more trials, the more the random errors tend to cancel out, so the average is closer to the true value and a real trend stands out from the noise. On the EOC, "Why repeat the measurement?" or "Why use many groups' data?" almost always wants reliability / reducing random error / a more trustworthy average as the answer.

Reliability (consistent, repeatable results) is not the same as accuracy (closeness to the true value) or validity (the test actually measures what it claims). A biased instrument can give very repeatable (reliable) but inaccurate readings.

Stating a testable hypothesis

A good hypothesis is a testable, if-then prediction that names the variables: "If the slope of the stream table increases, then the water will flow faster." It must be possible to test it by measurement and to show it false. A statement you cannot test ("mountains are the most beautiful landform") is not a scientific hypothesis.

Try this

Q1. In a test of "how water temperature affects how fast sugar dissolves," name the independent and dependent variables. [2]

• Cue. Independent: water temperature (changed on purpose). Dependent: time to dissolve, or amount dissolved (measured).

Q2. Explain why a scientist repeats a measurement many times. [2]

• Cue. Repeated trials reduce the effect of random error; averaging many results gives a more reliable value and shows the real trend.