Sitonce
Country: US
Show exams for United States Hong Kong
Sign in

NYSTCE 245: Scientific Inquiry, Evidence and Experimental Design

Updated 9 min read
Key takeaway

Scientific inquiry questions ask candidates to connect a testable question, variables, a fair method, observations, and a conclusion supported by evidence.

  • On NYSTCE 245, practice identifying what changes and what is measured, reading data carefully, and avoiding causal claims the design cannot support.
On this page13 sections
  1. Start with a question that can be investigated
  2. Recognize variables and fair tests
  3. Distinguish observations from inferences
  4. Understand correlation and causation
  5. Separate measurement from interpretation
  6. Reason from patterns, models, and systems
  7. Evaluate claims and communicate conclusions
  8. Common experimental-design errors
  9. Apply inquiry reasoning to a new scenario
  10. Prepare for inquiry items on Field 245
  11. Sample size and representativeness are different questions
  12. Use uncertainty honestly
  13. Choose a comparison that answers the question
Core skill
Use evidence to investigate a question and support a conclusion
Useful distinctions
Observation vs. inference; independent vs. dependent variable; correlation vs. causation
Test scope
Part of Science and Technology, 40% of Field 245
Practice
Interpret original scenarios, tables, and experimental designs

Start with a question that can be investigated

Scientific inquiry begins by turning curiosity into a question that evidence can address. A question such as “Which paper towel is best?” is too broad until best is defined. A more testable version might ask which of three brands absorbs the greatest volume of water under the same conditions. This version identifies a comparison and an observable outcome. On a selected-response item, notice whether the prompt asks for a question, a hypothesis, a method, a result, or a conclusion. Each stage calls for different reasoning.

A hypothesis is a proposed explanation or prediction that can be checked against observations. It is not simply a guess: it should connect variables or describe an expected pattern. A prediction might state that a towel with a thicker weave will absorb more water in a fixed time. The investigation could show otherwise. A result that fails to support a hypothesis is still useful evidence, because scientific reasoning requires revising an explanation when observations do not match it.

Recognize variables and fair tests

The independent variable is the factor deliberately changed, while the dependent variable is the outcome measured. In the towel example, brand is changed and absorbed water is measured. Controlled variables are conditions kept consistent, such as sheet size, water volume, timing, and test surface. If several conditions change at once, it becomes difficult to attribute a difference to the factor of interest. Questions often ask which plan best tests one possible cause; choose the design that isolates that factor while measuring the outcome consistently.

A fair comparison also needs a suitable procedure. If one brand is tested with one sheet and another with three, the amount of material differs. If one sample is soaked for a minute and another for ten seconds, the timing creates a competing explanation. Repeating trials and using enough observations help reveal whether a pattern is consistent or due to chance variation. A larger sample does not repair a badly controlled design, but replication can improve confidence when the method is sound.

Distinguish observations from inferences

An observation is information recorded directly, such as a temperature reading, number of seeds that germinated, or color change. An inference is an interpretation based on observations and background knowledge. “The solution turned blue” is an observation; “a chemical reaction occurred” is an inference. A question may ask what the evidence shows, so avoid adding facts the scenario does not establish. Inference is essential to science, but the strength of the inference should match the evidence and the possible alternative explanations.

Tables and graphs present evidence in compressed form. Read the title, labels, units, scale, and legend before interpreting a trend. Check whether a difference is large, small, increasing, decreasing, or stable, and whether the graph begins at zero. A visual can exaggerate or conceal change when its scale is unfamiliar. If measurements have repeated trials, compare the pattern across observations rather than selecting one unusual point. State only the conclusion supported by the displayed data.

Understand correlation and causation

Two variables can change together without one causing the other. For example, ice cream sales and swimming activity may both rise in warm weather; temperature can influence both. A correlation suggests a relationship worth investigating, but it does not by itself establish a causal mechanism. Stronger causal evidence comes from a controlled design, a plausible mechanism, consistent observations, and attention to confounding variables. A test item may present a before-and-after trend and ask what is supported; unless the design rules out alternatives, describe the pattern rather than asserting cause.

A confounding variable is an additional factor that changes with the independent variable and could explain the outcome. In a study of fertilizer and plant growth, differences in sunlight or watering can confound the comparison. Random assignment, matched conditions, replication, and transparent measurement can reduce some sources of bias. These tools do not guarantee perfect certainty, but they make the evidence more interpretable. When evaluating a claim, ask what else could produce the observed result.

Separate measurement from interpretation

Measurement uses a defined method and units. A vague measure such as “plant health” may lead different observers to record different outcomes. A clearer procedure might count leaves, measure stem height, or record mass at a consistent time. Some concepts are not directly visible and require an operational definition: a precise rule for how the concept will be observed or measured. An item can test whether a proposed measurement actually matches the question. If the question is about growth, a one-time leaf color observation may not answer it.

Precision and accuracy are related but distinct. Repeated values close together are precise; measurements near the accepted value are accurate. A tool that is consistently miscalibrated can be precise without being accurate. When data vary, consider both the instrument and natural variation. Do not assume that more decimal places mean better measurement. The method, calibration, units, and repeatability matter more than a display that looks exact.

Reason from patterns, models, and systems

Scientific explanations connect evidence with a mechanism or model. A model may represent a process that is too small, large, slow, or complex to observe directly. Models simplify; they are useful when they explain key relationships, but they do not include every detail of reality. If a diagram shows a system, use its labels and arrows to identify parts and relationships. Do not infer an unshown process merely because it is familiar from another context.

Many investigations examine systems rather than isolated objects. A change in one part can affect another part through flows of matter, energy, or information. A food-web change, for instance, can affect organisms connected by feeding relationships. Trace the links shown in the prompt, identify direct effects, and then be cautious about long-term outcomes that require additional information. Questions often reward a bounded conclusion: the most direct consequence supported by the stated relationship.

Evaluate claims and communicate conclusions

A conclusion should answer the investigation question and refer to evidence. It should also acknowledge limits when the data are narrow. If an experiment compares two brands of towel in one trial, it cannot establish which brand is best under all conditions. A careful statement describes what happened in the tested setup and identifies what further evidence would improve confidence. Scientific communication favors clarity about method and uncertainty over confident language that exceeds the observations.

When two explanations compete, identify the evidence each predicts and ask what observation would distinguish them. This turns disagreement into a testable next step. A strong investigation is reproducible: another person can understand the procedure well enough to repeat it. Records should include conditions, units, observations, and any deviation from the plan. On test questions, look for choices that preserve the evidence and avoid claims of certainty where the design has limitations.

Common experimental-design errors

Changing multiple variables at once, using unequal samples, measuring inconsistently, and ignoring a comparison group can make results difficult to interpret. A sample that is too small may miss natural variation, while a biased sample may not represent the group being described. Selecting only favorable results after an investigation also distorts the conclusion. In evaluating a method, identify the question first; then check whether the changed factor, measured outcome, controls, sample, and recording procedure actually answer it.

Another common error is confusing a control group with a controlled variable. A control group is a comparison condition, often one that does not receive the treatment. A controlled variable is a condition kept the same across groups. A design may need both. If an item asks how to improve reliability, repeating trials may help. If it asks how to isolate a cause, controlling other factors and using an appropriate comparison are central. Match the proposed improvement to the problem described.

Apply inquiry reasoning to a new scenario

Suppose a student claims that a certain study strategy improves quiz performance. A useful investigation would define the strategy, identify what performance means, compare suitable groups, keep quiz conditions similar, and collect results from enough students to interpret a pattern. The design should account for prior knowledge, study time, and other factors that could influence performance. A single student’s improvement may be meaningful to that student, but it does not by itself establish a general effect.

This example illustrates a transferable sequence: clarify the question, identify variables, select a fair method, collect data, interpret the pattern, evaluate alternatives, and communicate a bounded conclusion. Use it with physical science, life science, environmental, or technology contexts. The exact content varies, but the logic of matching evidence to a claim remains useful across the Science and Technology competency.

Prepare for inquiry items on Field 245

The official NYSTCE framework includes scientific inquiry within Science and Technology, one of the two 40-percent competencies in Field 245. The official sample guide shows that science questions may place a concept in an experimental setup or ask candidates to interpret a relationship. Read the full prompt, identify what is measured and changed, and eliminate choices that rely on unsupported assumptions. The framework and current sample guide define scope; this article provides original explanations and examples rather than reproducing official questions.

Practice with fresh scenarios and explain your reasoning in a sentence. If you miss an item, label the error: variable confusion, data reading, causal overclaim, measurement choice, or unfamiliar concept. Review the relevant idea, then solve a changed example. Mixed practice later helps you recognize inquiry reasoning when it appears beside content knowledge. The aim is to understand why evidence supports a conclusion and where that conclusion stops.

Sample size and representativeness are different questions

Increasing the number of observations can make a result less sensitive to individual variation, but it does not automatically make a sample representative. A large survey of volunteers from one school may still exclude people with different experiences. Ask both how many observations were collected and how they were selected. For a controlled experiment, consider whether repeated trials were consistent; for a claim about a population, consider whether the sample reflects that population.

Use uncertainty honestly

Measurements have limits, and observations can vary. A careful conclusion reports the pattern and conditions rather than pretending that every result is exact. If two groups show a small difference, ask whether the method can distinguish that difference from ordinary variation. Avoid saying “proves” when the evidence only supports a tentative explanation. This does not weaken scientific reasoning; it tells readers what the evidence can responsibly support and what further test could clarify.

Choose a comparison that answers the question

A comparison group helps show what might have happened without the changed condition. It should be similar enough to the treatment group that the planned factor is the meaningful difference. A weak comparison can mislead even when measurements are precise. Ask whether the groups differ in age, starting condition, environment, or handling, and whether those differences could affect the outcome. Matching the design to the research question is more important than adding complexity for its own sake. carefully.

Common questions

What is the independent variable?

It is the factor deliberately changed in an investigation.

How do I identify a fair test?

Look for one planned change, a consistently measured outcome, and other relevant conditions kept comparable.

Does correlation prove causation?

No. A relationship between variables can suggest a hypothesis but does not by itself establish cause.

How can I use graphs in science questions?

Read labels, units, scale, and legend, then state only the trend and conclusion supported by the data.