Original Scientific Investigation: Ask a Question Nobody Handed You

Overview

Every science experiment you have ever done was rigged. The teacher knew the answer, the procedure was written out for you, and the "right" result was waiting at the end like a punchline. That is fine for learning a technique, but it is not science. Science is what happens when nobody knows the answer yet — when you ask a question, design a fair test, and genuinely do not know what you will find until the data comes in. That uncertainty is the whole thing. This investigation puts you on the real side of it.

Over the next few weeks you will pick your own question, design a controlled experiment to answer it, run it enough times to trust the result, and write it up like a scientist — including the parts that failed. The result might confirm your hypothesis. It might contradict it. It might be a muddle that proves nothing. All three are real science, and the last one teaches you the most.

The Question

Your question must be testable, controllable, and answerable with the materials you can actually get. The best questions are small, specific, and surprisingly hard to predict. Stay away from "research questions" that just require looking up an answer — you want a question only an experiment can settle.

Strong, low-cost questions students have genuinely investigated:

• Does the type of liquid (water, soda, coffee, juice) affect how fast a nail rusts?
• Does music tempo affect how fast I can solve simple arithmetic problems?
• Does the angle of a paper airplane's wings change how far it flies?
• Do plants grown under different colored light grow at different rates?
• Does salt actually make water boil faster — or is that a myth?
• Does the brand of paper towel hold more water per sheet, controlling for sheet size?
• Does soil type affect how fast water drains through it?

Each of these has a clear thing-you-change, a clear thing-you-measure, and an answer you cannot be certain of in advance. Pick one, or design your own to that standard. Write it in your lab notebook as a single clean sentence.

Background

Before you can run a fair test, you need to understand the three kinds of variables. Mastering these is the difference between an experiment and a mess.

• Independent variable — the one thing you deliberately change. The liquid type, the light color, the wing angle. You should change only this.
• Dependent variable — the thing you measure to see the effect. Rust amount, growth height, flight distance. This is your data.
• Controlled variables — everything else, which you hold constant so it cannot secretly cause your result. Same temperature, same container, same starting amount, same time of day. These are the variables beginners forget, and forgetting them is how you get a result that looks real but isn't.

The core idea of a controlled experiment: change one thing, hold everything else steady, and measure the effect. If you change two things at once and the result shifts, you have no way of knowing which change caused it. Discipline about this single principle is what separates a trustworthy result from a worthless one.

Hypothesis

Before you begin, write down what you think will happen and why:

"I think _______ because _______."

Write this in your lab notebook before the first trial, and date it. The "because" matters as much as the prediction — it forces you to state your reasoning, which means that when the result comes in, you learn something whether you were right or wrong. A hypothesis without a reason is just a guess. A hypothesis with a reason is a small theory you are putting on trial.

Be willing to be wrong. The most exciting words in science are not "I knew it" — they are "huh, that's not what I expected." A surprising result is a gift. It means reality knows something you didn't.

Materials

• A bound lab notebook (loose paper gets lost; science demands a permanent, dated record)
• Your independent-variable materials (the different liquids, light colors, paper brands, etc.)
• Identical containers, holders, or test beds — so the only difference between trials is your independent variable
• A measuring instrument matched to your dependent variable, accurate enough to detect a small difference
• A timer or calendar for trials that unfold over time
• A camera for photographing the setup and results at each stage

Procedure

Setup

1. Write the full procedure in your lab notebook before you start. Number the steps. A real experiment is repeatable by a stranger — if someone else could follow your written steps and get comparable results, your procedure is good enough.
2. Identify your three variable types in writing. List the one independent variable, the one dependent variable, and every controlled variable you can think of. This list is your defense against a contaminated result.
3. Build identical test setups that differ only in the independent variable. If you are testing four liquids, use four identical containers, the same amount of liquid, the same nail from the same box, in the same location.
4. Set up a control if your question has one. The control is the "do nothing different" baseline — plain water, no music, flat wings. Without a baseline, you have nothing to compare your changes against.

Experiment

1. Run multiple trials, not one. A single trial proves nothing — it could be a fluke. Run at least three trials of each condition (more is better). If results vary wildly between identical trials, that variation is itself important data: it means your effect is small relative to the noise.
2. Change only the independent variable between conditions. Resist every temptation to "improve" the procedure partway through. If you change the method on trial four, trials one through three are no longer comparable.
3. Measure the same way every time, with the same instrument, read the same way, at the same point in the process.
4. Photograph each stage. Visual records catch things your numbers miss, and they make your write-up credible.

Record

Record everything in the lab notebook as it happens — never from memory afterward. Build a data table:

The "anything unusual" column is where honest science lives. The spilled container, the cloudy day, the timer you started late — write it down. Those notes are how you later explain a weird result instead of pretending it didn't happen.

Analysis

Work through these in your notebook, in order:

• What happened? State the results plainly, by condition. Average your trials for each condition if you have numeric data.
• Did it match your hypothesis? Compare the result to your written prediction. If yes, by how much? If no, how far off?
• Is the difference real, or is it noise? Look at how much your repeated trials of the same condition varied. If your conditions differ by less than that natural trial-to-trial variation, you cannot claim a real effect — the difference is within the noise. This is the single most important judgment in experimental science, and most people skip it.
• What might explain the results? Offer the best explanation your data supports — and only what it supports.
• What would you change if you did it again? Every experiment has flaws you discover only by running it. Name three. This is not admitting failure; it is how every real scientist thinks.

A Worked Example: The Salt-and-Boiling-Water Myth

To see how the whole method fits together, walk through one investigation start to finish. You have heard that adding salt makes water boil faster. Most people believe it. Is it true?

The question: Does adding salt to water make it reach a boil faster than plain water?

The variables:

• Independent (what you change): whether salt is added, and how much.
• Dependent (what you measure): time in seconds from heat-on to a full rolling boil.
• Controlled (held steady): the exact same pot, the same volume of water measured precisely (say 1 liter), the same starting water temperature, the same burner on the same setting, the lid on or off consistently, and the same definition of "boiling" (you must decide in advance what counts — the first bubbles, or a full rolling boil — and apply it identically every time).

The hypothesis (written first): "I think salted water will boil slightly slower than plain water, because dissolving salt raises the temperature at which water boils, so it has to get hotter to start boiling." Note that this prediction actually runs against the popular belief — and writing the reasoning forces you to think the science through before testing.

The procedure: Boil 1 liter of plain water, three separate trials, recording the time each. Then boil 1 liter of water with a tablespoon of salt, three trials. Then a liter with a quarter cup of salt, three trials. Same pot, same burner, same everything else. Nine trials total.

Sample data (illustrative):

The crucial analysis step: Average each condition (plain ≈ 364 s, 1 Tbsp ≈ 365 s, 1/4 cup ≈ 372 s). Now look at how much the plain trials varied among themselves: 358 to 371, a spread of 13 seconds, just from normal run-to-run noise. The difference between plain and 1 tablespoon of salt is about 1 second — far smaller than the noise. You cannot claim that difference is real. The 1/4-cup result is about 8 seconds slower, which is closer to the noise band but starts to suggest a small slowing effect, consistent with the hypothesis. The honest conclusion: a normal cooking pinch of salt makes no detectable difference to boiling time, and a large amount may slightly slow it — the opposite of the popular belief, and exactly what the chemistry predicts.

This example shows why the noise question matters more than the averages. A careless experimenter would have seen "365 vs 364" and announced a difference. The disciplined one looks at the spread of repeated identical trials, realizes the difference is buried inside the noise, and refuses to overclaim. That refusal is the scientific skill.

The Explanation

Once you have analyzed your own results, look up the established science behind your question. (Do this last, on purpose — looking it up first would have biased your whole investigation.) Compare what you found to what is known. If they match, you have independently reproduced a real result, which is genuinely satisfying. If they don't, the interesting question is why — was your method flawed, your sample too small, your measurement too crude, or did you stumble onto something the simple explanation misses? Reproducing known results and explaining your own deviations is exactly what working scientists spend their careers doing.

Extensions

• Change one variable: Once you have an answer, vary something you held constant — temperature, time, quantity — and see whether your effect holds up under new conditions. Robustness across conditions is what turns a one-off result into real knowledge.
• Real-world connection: Almost every kitchen, garden, garage, and gym is full of folk wisdom ("salt makes water boil faster," "plants grow better with music," "cold weather causes colds"). Pick one piece of folk wisdom and put it on trial. Testing the things everyone "knows" is some of the most satisfying science there is.
• Further reading: Look up the difference between a "controlled experiment" and an "observational study," and why scientists trust the first more. Then read about the "replication crisis" in modern science — the discovery that many published results don't hold up when repeated. It will make you take your own "run multiple trials" rule far more seriously.

Safety Notes

This investigation is rated yellow — the specific risks depend entirely on what you choose to test, so assess your own setup before you begin and have an adult review your plan.

Chemical/Material Hazards

• If your experiment involves household liquids (vinegar, bleach, ammonia, drain cleaner), never mix cleaning products — some combinations release toxic gas. Bleach plus ammonia is genuinely dangerous. Test each substance in its own isolated container, and check that any chemical you use is safe to handle before you handle it.
• If you use heat (boiling water, a stove, a lamp that gets hot), treat it as a burn hazard. Have an adult present for anything involving an open flame or sustained high heat.
• Sharp objects (nails, blades, glass containers) carry cut risk. Handle deliberately and keep a first aid kit within reach.

Disposal

• Dispose of any used chemicals according to their label, not down the drain by default. Soured liquids, rusted-metal waste, and used cleaning products may need specific disposal. When in doubt, ask an adult and check the product label.
• Compost or trash plant and food materials normally; do not dump experimental soil mixtures into a garden bed without thinking about what you added to them.

Protective Equipment

• Wear eye protection for anything involving splashing liquids, heat, or projectiles (yes, including paper airplanes launched at speed indoors near faces).
• Wear gloves when handling anything corrosive, staining, or sharp.
• Work in a ventilated space for anything with fumes, and keep the work area clear of food and drink so nothing is mistaken for a snack.