Lesson 1.1 · Unit 1 · NGSS practice 1
What makes a question scientific, and what biology cannot settle
Biology can tell you how many offspring a population produces, which allele a child inherited, and what happens to a leaf kept in the dark for a week. It cannot tell you whether a wolf is more valuable than a sheep. The difference is not that one question is harder. It is that only one of them names an observation that could come out either way.
- A testable question names an observation you could actually make and specifies what would count as each possible answer.
- An untestable question is not a bad question. Questions of value, ethics and meaning are real questions; they are simply not settled by measurement, and science claiming to settle them is science overreaching.
- A hypothesis proposes a mechanism and predicts an observation. "Plants grow taller with fertilizer" is a prediction. "Plants grow taller with fertilizer because nitrogen is required to build proteins" is a hypothesis: it says why, so it can be wrong in an informative way.
- A hypothesis must be falsifiable. You must be able to state, before you collect data, which result would make you abandon it. If no possible result would, you are not testing anything.
- Comparative words need a criterion. "Is this plant healthier?" means nothing until health is defined as height, mass, leaf count or seed yield. Fixing the criterion is part of writing the question.
- Observation and experiment answer different questions. An observational study records what is already there. An experiment changes one thing on purpose. Only the second can support a claim about cause.
Where students lose marks: calling any sentence a hypothesis. A hypothesis is not a topic ("my hypothesis is fertilizer") and not a restatement of the question. It is a proposed explanation, stated so that a specific result would contradict it.
The task. Turn the vague question "Is fertilizer good for plants?" into something a fourteen year old with a windowsill could actually answer.
Step one: find the word doing no work. "Good" is undefined. Good for what? Taller stems, greener leaves, more seeds and longer life are different outcomes, and fertilizer could improve one while harming another.
Step two: choose one measurable outcome. Pick stem height in centimeters after fourteen days. It is easy to measure, it is a number rather than a judgment, and two people measuring the same plant will agree.
Step three: name the thing you will change. Presence of fertilizer: one group gets the recommended dose, one group gets none. Everything else stays the same, which is the subject of lesson 1.2.
Step four: write the question. "Does adding the recommended dose of a nitrogen fertilizer increase the mean stem height of radish seedlings after fourteen days, compared with no fertilizer?" Every word is now checkable.
Step five: write the hypothesis, with a mechanism. "Adding nitrogen fertilizer increases mean stem height, because nitrogen is a component of the amino acids a seedling needs to build new tissue, and the soil supply limits growth." The clause after "because" is what makes it a hypothesis.
Step six: state what would falsify it. If the fertilized group's mean height is the same as or lower than the unfertilized group's, the hypothesis as stated is wrong. Writing this down before collecting data is what stops you explaining away whatever you get.
Step seven: notice what the study still cannot tell you. It cannot tell you whether using the fertilizer is a good idea. That depends on cost, on runoff into a nearby pond, and on what you want the plants for. Those are real considerations, and the experiment is silent on all of them.
Questions 1 to 5 are recall. Questions 6 to 10 ask for reasoning.
- What two things must a testable question name?
Show the full solution
An observation you could actually make, and what would count as each possible answer
- What does a hypothesis have that a prediction does not?
Show the full solution
A proposed mechanism: it says why, not only what will happen
- What does it mean for a hypothesis to be falsifiable?
Show the full solution
You can state in advance a possible result that would show it is wrong
- Which type of study can support a claim about cause, and why?
Show the full solution
An experiment, because the investigator changes one thing on purpose rather than recording what is already there
- Rewrite "Is this soil better?" so that it is measurable.
Show the full solution
Any version that fixes a criterion, for example "Does this soil produce a greater mean dry mass of bean plants after thirty days?"
- "Some species are more important than others." Explain why biology cannot
settle this, and what it can contribute.
Show the full solution
Importance is a judgment about value, and no measurement returns a value. Biology can measure how many other species depend on a given one, what happens to a community when it is removed, and how much of an ecosystem process it performs, all of which are real and relevant findings. But turning those numbers into "more important" requires deciding what we are trying to protect and why, which is a question about goals rather than about organisms. Science informs the decision without making it. Value is not measurable, though the consequences that inform the judgment are
- A student writes: "My hypothesis is that I will test whether light affects
algae." Explain what is wrong with it and fix it.
Show the full solution
This states an intention, not an explanation. It predicts nothing, so no result could contradict it, and a claim that cannot be contradicted teaches nothing when the data arrive. A hypothesis must propose a mechanism and commit to an outcome. A corrected version: "Increasing light intensity increases the rate of oxygen production in pond algae, because light supplies the energy for the light-dependent reactions of photosynthesis, and light is the limiting factor at low intensity." It states an intention rather than a falsifiable explanation
- Why is it important to write down the falsifying result before collecting data?
Show the full solution
Because afterwards it is very easy to reinterpret whatever happened as support. If the fertilized plants grow less, an investigator who has not committed in advance can decide the dose was wrong, the season was unusual or the effect takes longer, and keep the hypothesis regardless of the evidence. Fixing the falsifying result first means the data get a genuine chance to overturn the idea, which is the only way a test is informative. It prevents reinterpreting any outcome as support after the fact
- A study finds that towns with more veterinarians have more cases of a cattle
disease. Explain why this does not show that veterinarians cause the disease.
Show the full solution
Nothing was changed on purpose, so this is an observational study and a third factor can produce both patterns at once. Towns with more cattle will have both more veterinarians and more cases of a cattle disease, and herd size is doing the work in both columns. The two measurements rise together without either causing the other. Distinguishing the possibilities requires either holding herd size constant or changing veterinarian numbers deliberately, which nobody would do. A third factor, herd size, produces both patterns
- Give a question about photosynthesis that biology can answer and one it cannot,
and say what makes the difference.
Show the full solution
Answerable: "Does raising carbon dioxide concentration from 0.04 percent to 0.10 percent increase the rate of oxygen production in pond weed at constant light and temperature?" It names a change, a measurement and a comparison. Not answerable by measurement: "Should we raise carbon dioxide concentration in greenhouses?" That depends on cost, on energy use and on what else the greenhouse is for. The difference is that the first specifies an observation that could come out either way, while the second requires weighing goods against each other. The first names a possible observation; the second requires a value judgment