NGSA Science common mistakes
NGSA Science · Understand the traps and how to avoid them.
When a question gives a data table or graph and asks for a conclusion, the candidate answers from prior knowledge or a remembered fact instead of reading the trend the data actually shows, sometimes contradicting the supplied numbers.
Candidates study science content as a body of facts and default to recall; they do not treat the prompt's data as the authoritative evidence, and under time pressure they pattern-match to a familiar scenario rather than checking the axes, units, or comparison group.
What to do instead
Before answering any data-based item, state to yourself in one sentence what the graph or table shows (what is compared, what changes, in which direction). Then test each option against that sentence and eliminate any option that the data do not support, even if it is a true science fact. Practice with released-style data items where the correct answer is the least exciting one supported by the trend.
The candidate identifies a real-world phenomenon (e.g., a dropped object, a moving car, a hot pan) and answers from everyday intuition rather than from the physical model the question specifies, such as assuming heavier objects fall faster or that a constant push is needed to keep something moving.
Everyday experience is dominated by friction and air resistance, so intuitive predictions conflict with idealized Newtonian models; candidates do not notice cue words like 'ignore air resistance' or 'net force' that signal which model applies.
What to do instead
Rewrite each prompt into a model statement before choosing: list the forces acting, mark whether the system is at rest or moving at constant velocity (net force zero) or accelerating (net force nonzero), and explicitly cancel friction or drag when the prompt says to. Drill the classic contrasts: constant velocity versus constant acceleration, mass versus weight, and force versus momentum.
The candidate treats a correlation shown in data or a described scenario as proof of causation, or overlooks a confounding variable, and selects a conclusion that claims one factor caused another.
Causal language in everyday speech is loose, and observational data are often presented in a narrative that implies cause; candidates also skip checking whether the study manipulated a variable or merely measured existing groups.
What to do instead
Ask two questions of every conclusion option: was the variable manipulated by the researcher, and is there a plausible third factor that could explain the pattern? Reject options containing causal verbs (causes, leads to, results in) unless the design was a controlled experiment with the independent variable changed.
The candidate confuses the roles of independent, dependent, and controlled variables, or proposes an investigation that changes several variables at once, so the chosen experimental design cannot isolate the effect being tested.
Textbook definitions are memorized as labels but not operationalized; candidates do not translate 'what I change' and 'what I measure' into the concrete setup described in the prompt, especially when the prompt describes a realistic procedure with several measured quantities.
What to do instead
For every experimental-design item, write a three-line plan: what I deliberately change (independent), what I measure as the outcome (dependent), and what I hold constant (controls). Eliminate any option that alters more than one factor at a time or that measures something unrelated to the stated question.
The candidate selects an answer that describes a mechanism at the wrong scale or level of organization, for example explaining a population-level pattern with an individual organism's behavior, or explaining a cell process with organ-level language.
Life-science content spans molecules, cells, organs, organisms, populations, and ecosystems, and the same word (e.g., 'adaptation', 'respiration', 'competition') carries different meanings at each level; candidates match on the keyword rather than on the level the question targets.
What to do instead
Label the level of organization in the prompt before reading the options (molecule, cell, tissue, organ, organism, population, community, ecosystem). Eliminate options pitched at a different level, and for the word 'adaptation' decide whether the question means a trait shaped over generations or a short-term physiological response.
The candidate explains inheritance or variation using acquired characteristics or blending, saying offspring will inherit a trait the parent developed during its lifetime, or that offspring are a simple average of the two parents.
Blending inheritance and Lamarckian inheritance are intuitive and appear in casual speech about 'taking after' a parent; candidates also do not distinguish genotype from phenotype or dominant from recessive expression.
What to do instead
Apply a fixed rule: only genetic information in gametes is inherited, so changes to body cells or acquired traits are not passed on. For any inheritance item, set up a quick genotype cross (letters for alleles) and derive phenotypes from the cross rather than from intuition about the parents' appearance.
The candidate applies a familiar Earth or space pattern to the wrong scale or cause, for example attributing seasons to Earth's changing distance from the Sun, treating the Moon's phases as caused by Earth's shadow, or confusing weather with climate.
Diagrams and everyday language reinforce these misconceptions (the Sun looks closer in summer, the crescent looks 'cut off'), and candidates do not check the mechanism against the geometry of tilt, orbit, and shadow.
What to do instead
Memorize and apply three mechanism checks: seasons come from axial tilt changing the angle and duration of sunlight, not distance; lunar phases come from the changing viewing angle of the sunlit half, with eclipses as a separate alignment event; climate is a long-term distribution while weather is a short-term state. For each option, name the mechanism and reject options whose mechanism is wrong even if the outcome sounds plausible.
The candidate mixes up the direction and products of energy transfer in Earth systems, such as reversing the order of processes in the rock cycle, confusing convection cells with plate boundaries, or misreading a food web's arrow direction as energy flowing to the prey.
Cyclic processes are learned as lists of stages without a direction, and arrows in diagrams are read as 'eats' rather than 'energy flows to'; candidates also do not link the process to its energy source.
What to do instead
Redraw each cycle as a directed loop with labels on the arrows (melting, cooling, weathering, deposition; or energy flow from producer to consumer) and state the energy source at each step (solar, gravitational, or Earth's internal heat). Answer only from the redrawn loop.
The candidate proposes a solution to an engineering problem without checking it against the stated constraints and criteria, choosing the most impressive-sounding design rather than the one that meets the given limits on cost, materials, time, or safety.
Engineering items read like general science questions, so candidates look for the technically sophisticated option and skip the constraint list embedded in the prompt; they also confuse a prototype with a final design and ignore trade-offs.
What to do instead
Before evaluating options, extract the constraints (limits that must not be violated) and criteria (measures of success) into a short list, then score each option against that list. Reject any option that violates a stated constraint regardless of its performance, and prefer options that describe iterative testing and revision over one-shot perfect solutions.
The candidate treats a scientific model, theory, or law as a fixed final truth, or answers that a new finding 'proves a theory wrong' rather than evaluating the quality of evidence, sample size, and whether the claim was peer reviewed.
School science often presents conclusions as settled facts, so candidates do not practice evaluating evidence quality, and the word 'theory' in everyday use means a guess, which reverses its scientific meaning.
What to do instead
Use a fixed evidence checklist for any claim-evaluation item: sample size and representativeness, presence of a control or comparison, whether measurements were repeated, and whether the claim was independently reviewed. Treat theories as well-supported explanatory frameworks that are revised when evidence accumulates, and reject options that describe a single study as overturning an established framework.