AP Biology · Strategy 05 · Scientific Investigation

Experimental Design

Q3 on every AP Biology exam tests experimental design. This module covers the 7 required elements, the most-tested conceptual distinction (control group vs. controlled variables), hypothesis writing, and a fill-in FRQ template that works for any design question.

5.1

Where Experimental Design Appears

Experimental design questions appear in multiple locations on the exam. Q3 is the dedicated short FRQ, but design skills appear in MCQ and long FRQ as well.

LocationHow It AppearsWhat Is Tested
Q3 Short FRQ"Design an experiment to test whether…" or "A researcher wants to determine if… Describe an experimental design."Full experimental design: IV, DV, control, controlled variables, hypothesis, measurable outcome
Q1 / Q2 Long FRQOften the final sub-part: "Design a follow-up experiment to test…"Extending an existing experimental context; one new IV manipulated while others held constant
MCQ Stimulus SetsQuestions about a described experiment: "What is the control group?" or "Which modification would improve this design?"Identify variables, evaluate flaws, suggest improvements
Discrete MCQ"Which experimental design best tests the hypothesis that…"Choose the design that correctly isolates the variable of interest
The Core Experimental Design Principle

A valid experiment changes exactly one variable (the independent variable) while holding all other conditions constant, and measures a specific, quantifiable outcome (the dependent variable). Any design that changes more than one variable simultaneously cannot establish which change caused the observed effect. This is the fundamental logic AP readers test in every design question.

5.2

The 7 Elements of a Valid Experiment

Every experimental design question on the AP Biology exam scores points based on these seven elements. You do not need all seven for every question — check the point value and sub-part structure — but knowing all seven lets you write a complete design for any question.

1
Hypothesis
A testable, directional prediction of the expected outcome. Written as an If–then–because statement or as a direct prediction. Must be falsifiable — there must be a possible result that would prove it wrong.
Example: "If bean plants are exposed to increased CO₂ concentrations, then their rate of photosynthesis will increase, because CO₂ is a substrate for RuBisCO in the Calvin cycle."
❌ Trap: Writing a question instead of a prediction. "Will higher CO₂ increase photosynthesis?" is not a hypothesis.
2
Independent Variable (IV)
The variable the researcher deliberately manipulates. There is exactly ONE independent variable per experiment. It must be clearly defined with specific levels or values.
Example: "The independent variable is CO₂ concentration, tested at 200, 400, 600, and 800 ppm."
❌ Trap: Naming the general topic instead of the specific variable. "Light" is not the IV — "light intensity in μmol photons/m²/s" is the IV.
3
Dependent Variable (DV)
The variable measured as the experimental outcome. Must be quantifiable (expressed as a number with units). The DV is what you observe or measure in response to the IV.
Example: "The dependent variable is net photosynthesis rate, measured as μmol O₂ produced per minute per gram of leaf tissue."
❌ Trap: Naming an outcome that cannot be measured numerically. "How healthy the plant looks" is not a valid DV.
4
Control Group
The group that receives no treatment (or the standard/baseline treatment). Provides the reference point against which experimental groups are compared. Without a control, you cannot determine whether any observed change is due to the treatment.
Example: "The control group consists of plants grown at ambient CO₂ (400 ppm) under identical light and temperature conditions."
❌ Trap: Confusing control group with controlled variables. The control group is a set of subjects. Controlled variables are conditions kept constant across ALL groups including the control.
5
Controlled Variables
All conditions held constant across all groups (including the control) to ensure that only the IV differs between groups. Must name at least 2–3 specific variables, not just "everything else."
Example: "Controlled variables include light intensity (600 μmol photons/m²/s), temperature (25°C), watering frequency, soil composition, and plant species."
❌ Trap: Writing "all other variables are kept the same" without naming them. AP Readers require specific named variables.
6
Replication
Using multiple subjects per treatment group (n ≥ 3) to reduce the effect of random variation and increase the reliability of results. A single subject per group cannot distinguish a treatment effect from individual variation.
Example: "Each CO₂ concentration is tested with 10 plants (n = 10 per group). The experiment is repeated three times."
❌ Trap: Describing replication as "repeating the experiment once." Replication means multiple subjects per group within a single experiment run.
7
Measurable Outcome
A specific description of how the DV will be measured, including the method, units, and timeframe. This element transforms a vague design into an actionable protocol.
Example: "Net photosynthesis rate is measured by counting the number of leaf disk flotations per 10 minutes in each CO₂ treatment, with one flotation representing a defined amount of O₂ produced."
❌ Trap: Naming the DV but not explaining how it will be measured. "Measure photosynthesis" is incomplete — say how.
5.3

Control Group vs. Controlled Variables

This is the single most commonly confused distinction in AP Biology experimental design. AP Readers specifically look for correct use of both terms.

Control Group

A set of subjects that receives no treatment, a placebo, or the standard baseline condition. It exists as a reference point for comparison.

There is one control group per experiment (sometimes called "negative control"). A positive control (known to produce the expected result) may also be included.

  • It is a group of organisms, cells, or samples
  • It receives zero treatment or standard conditions
  • You compare experimental group results to the control group
  • Example: "Plants receiving 0 mg/L fertilizer"
Controlled Variables

The experimental conditions that are held constant across all groups — including the control group. They are not the same as the control group.

Controlled variables ensure that only the IV differs between groups, so any difference in the DV can be attributed to the IV alone.

  • They are conditions, not groups of subjects
  • They apply to ALL groups (experimental AND control)
  • You must name specific variables, not say "everything else"
  • Example: "Temperature (25°C), light intensity, soil type, watering volume"
The Most Common Mistake

Students write: "The control group is the group where temperature is kept constant." This is wrong. Temperature kept constant is a controlled variable. The control group is the set of plants/cells/organisms that receives no experimental treatment. A single experiment has one control group but many controlled variables.

ScenarioControl GroupControlled Variables (examples)
Testing effect of caffeine on heart rate in DaphniaDaphnia in plain spring water (0 mg/L caffeine)Water temperature, light level, Daphnia size, time of measurement
Testing effect of substrate concentration on enzyme activityEnzyme solution with no substrate added (0 mM)Temperature, pH, enzyme concentration, reaction time
Testing effect of antibiotic on bacterial growthBacteria on plates with no antibiotic (DMSO solvent only)Bacterial strain, incubation temperature, growth medium, inoculum density
5.4

Hypothesis Writing

AP Biology tests two types of hypotheses: the alternative hypothesis (your prediction) and the null hypothesis (used in statistical testing). Know when each is required.

Alternative Hypothesis (H₁)
Your Directional Prediction

States that there IS a relationship between IV and DV, and predicts the direction. Written as an if–then statement. Must be falsifiable and specific.

Format: "If [IV is manipulated in this way], then [DV will change in this direction], because [biological mechanism]."

Example: "If temperature increases from 20–40°C, then the rate of enzyme-catalyzed reactions will increase, because higher temperatures increase the kinetic energy of substrate molecules, increasing the frequency of productive collisions with the active site."

Null Hypothesis (H₀)
The No-Effect Statement

States that there is NO relationship between IV and DV — any observed difference is due to chance. Used as the starting assumption in statistical testing (chi-square, t-test). Rejected when p < 0.05.

Format: "[IV] has no effect on [DV]." OR "There is no statistically significant difference in [DV] between treatment groups."

Example: "Temperature has no effect on the rate of the enzyme-catalyzed reaction. Any observed differences between temperature groups are due to random variation."

When AP Asks for Each

• "State a hypothesis" or "Write a hypothesis" → write the alternative hypothesis (directional prediction with mechanism)
• "State a null hypothesis" → write the null hypothesis (no-effect statement)
• "The researcher uses a chi-square test. What is the null hypothesis?" → always the null hypothesis, stating observed frequencies do not differ from expected
• If not specified, default to the alternative hypothesis

5.5

FRQ Design Template

Use this fill-in template for any "design an experiment" sub-part. Each row corresponds to one or more scoring points on the AP rubric. Fill in the italicized placeholders with content specific to the question.

AP Biology · Experimental Design Template
Hypothesis
If [IV is changed in this specific way], then [DV will change in this direction], because [named biological mechanism linking IV to DV].
Independent Variable
The independent variable is [specific variable name], which will be tested at the following levels: [list 3+ specific values with units].
Dependent Variable
The dependent variable is [specific measurable outcome], measured in [units] using [specific method or instrument].
Control Group
The control group consists of [organism/cells/samples] that receive [no treatment / standard condition / 0 level of IV]. This establishes the baseline value of the DV in the absence of the experimental treatment.
Controlled Variables
The following variables are held constant across all groups: [specific variable 1 with value], [specific variable 2 with value], [specific variable 3], and [organism type, age, sex if applicable].
Replication
Each treatment group contains [n ≥ 3] subjects. The experiment is replicated [number] times to reduce the effect of random variation.
Predicted Result
If the hypothesis is supported, the [experimental group(s)] will show [higher/lower/no change in DV] compared to the control group. [State specific directional relationship.]

Worked Example — Template Applied

Question: Design an experiment to test whether gibberellin (GA) promotes stem elongation in dwarf pea plants.

Hypothesis: If dwarf pea plants are treated with gibberellin, then their stem length will increase compared to untreated plants, because gibberellin promotes cell elongation in the stem by stimulating the loosening of cell walls and increasing vacuolar water uptake.

IV: Presence or absence of gibberellin, tested at: 0 μM GA (control), 10 μM GA, 50 μM GA, 100 μM GA.

DV: Stem length (cm), measured from the soil surface to the apical meristem using a ruler after 14 days.

Control group: Dwarf pea plants treated with distilled water only (0 μM GA). This controls for the effect of adding liquid to the plant independent of the hormone.

Controlled variables: Pot size and soil composition, watering volume (50 mL every 2 days), light intensity (16h/8h light/dark at 200 μmol photons/m²/s), temperature (22°C), plant age at start (7-day-old seedlings), pea cultivar (same dwarf variety throughout).

Replication: n = 10 plants per treatment group; experiment repeated twice.

Predicted result: Plants treated with higher GA concentrations will show greater stem elongation than control plants, with a dose-dependent increase in stem length.

5.6

Evaluating Existing Experimental Designs

Some Q3 questions (and Q1/Q2 sub-parts) ask you to evaluate a described design rather than create one. The same 7 elements are the checklist — identify which element is missing or flawed.

Common FlawWhy It’s a ProblemHow to Fix It
No control groupCannot establish baseline; impossible to attribute the change in DV to the IVAdd a group that receives no treatment or the zero level of the IV
Only one subject per group (n=1)Cannot distinguish treatment effect from individual variation; results are not replicableIncrease sample size to n ≥ 3 per group
Multiple variables changed simultaneouslyImpossible to determine which variable caused the observed outcome (confounding variables)Change only the IV; hold all other conditions constant
DV is not quantifiable"How well the plants grow" cannot be statistically analyzed or objectively comparedReplace with a specific measurable outcome: height (cm), mass (g), O₂ production rate (μmol/min)
Controlled variables not namedReader cannot verify the design is valid; uncontrolled variables may have caused the resultExplicitly list at least 3 specific conditions held constant with values
No stated time frameUnclear when measurements are taken; results from different time points are not comparableSpecify exact duration and measurement schedule: "after 14 days," "every 2 hours for 24 hours"
5.7

Common Design Traps

Top 6 Experimental Design Mistakes
  • Confusing control group with controlled variables — the most common error. Control group = a set of subjects. Controlled variables = conditions kept constant. These are not the same thing.
  • Changing two variables at once — e.g., increasing both temperature AND pH simultaneously. This creates a confounded design where neither variable can be said to cause the observed effect.
  • Unmeasurable dependent variable — "observe how the organism responds" is not a DV. Every DV must produce a number with units.
  • No control group for the solvent/vehicle — if a treatment is dissolved in DMSO, the control must receive DMSO only (not plain water), to control for DMSO’s own biological effects.
  • Circular hypothesis — "If gibberellin is added, then gibberellin will cause elongation" restates the question without making a testable biological prediction. The because-clause must name the specific mechanism.
  • Describing the expected result instead of the design — "The plants given more fertilizer will grow taller" describes an outcome, not a procedure. The design must describe what you will do, not what you expect to observe.
5.8

Practice Questions

MCQ · Discrete · SP 3 · Unit 4

A researcher wants to test whether a protein kinase inhibitor (Drug X) reduces the rate of cell proliferation in cancer cells. Which experimental design best tests this hypothesis?

  • (A) Treat one flask of cancer cells with Drug X and count the cells after 48 hours
  • (B) Treat cancer cells with Drug X at three concentrations and observe cell appearance under a microscope
  • (C) Treat three groups of cancer cells with 0, 10, and 50 μM Drug X (dissolved in DMSO), with a DMSO-only control group, and measure cell number at 0 and 48 hours in each group
  • (D) Treat cancer cells with Drug X and normal cells with no treatment and compare the results
Answer: (C) — (C) includes: IV (Drug X concentration at 0, 10, 50 μM), DV (quantifiable: cell number), a proper solvent control (DMSO only, controlling for solvent effects), replication (three groups), and a quantitative measurable outcome (change in cell number from 0 to 48 hours). (A) lacks a control group — impossible to know if cell number changed due to Drug X or normal growth. (B) uses a non-quantifiable DV ("cell appearance"). (D) compares cancer cells with drug to normal cells without drug — two variables differ (cell type AND treatment), making it impossible to attribute any difference to Drug X alone.
MCQ · Discrete · SP 3 · Controlled Variables

A student designs an experiment to test the effect of soil pH on bean plant growth. She grows 5 plants in soil at pH 5, 5 plants in soil at pH 7, and 5 plants in soil at pH 9. She places pH 5 plants near a south-facing window, pH 7 plants under a grow light, and pH 9 plants in a shaded corner. Which statement best identifies the flaw in this design?

  • (A) The sample size of 5 plants per group is too small to draw valid conclusions
  • (B) Light intensity is not controlled across groups, so any difference in growth could be due to light rather than soil pH
  • (C) The experiment should include more than three pH levels to be valid
  • (D) Bean plants are not an appropriate organism for studying the effect of pH on plant growth
Answer: (B) — The critical flaw is that light intensity differs across groups — this is a confounding variable. If plant growth differs between groups, it is impossible to determine whether the cause is soil pH (the intended IV) or light intensity (an uncontrolled variable). This violates the core principle of experimental design: only one variable can differ between groups. (A) is a valid concern but not the primary flaw — n = 5 is actually reasonable for this type of experiment. (C) is incorrect: three levels are sufficient to assess a trend.
Q3 Short FRQ · Scientific Investigation · SP 3 · Unit 3 · 4 pts

A student hypothesizes that increasing light intensity increases the rate of photosynthesis in aquatic plants. Design an experiment using Elodea (an aquatic plant) to test this hypothesis.

(a) State the null hypothesis for this experiment. [1 pt]
(b) Identify the independent variable and the dependent variable, including how the dependent variable will be measured. [1 pt]
(c) Describe the control group and identify two controlled variables. [1 pt]
(d) Predict the result if the hypothesis is supported. [1 pt]
(a) Null hypothesis [1 pt]: Light intensity has no effect on the rate of photosynthesis in Elodea. Any observed differences in oxygen production between light intensity groups are due to random variation. [Must state no effect on DV; must reference chance/random variation for full credit]
(b) Variables [1 pt]: Independent variable: light intensity (μmol photons/m²/s), tested at 0, 50, 100, 200, and 400 μmol photons/m²/s by placing lamps at different distances from the plant. Dependent variable: rate of photosynthesis, measured as the number of oxygen bubbles produced per minute (or O₂ concentration change measured by oxygen probe in μmol/L/min). [1 pt: both IV and DV named with units and measurement method]
(c) Control and controlled variables [1 pt]: Control group: Elodea sprigs placed in darkness (0 μmol photons/m²/s) in sodium bicarbonate solution, establishing the baseline O₂ production (cellular respiration only, no net photosynthesis). Controlled variables: CO₂ concentration (maintained by 0.2% sodium bicarbonate solution), water temperature (22°C ± 0.5°C), Elodea sprig length (5 cm), and measurement duration (5 min per light level). [1 pt: valid control group + at least 2 named controlled variables with values]
(d) Predicted result [1 pt]: If the hypothesis is supported, the rate of oxygen production will increase as light intensity increases from 0 to some maximum value, then plateau as another factor (CO₂ concentration or enzyme capacity) becomes limiting. The control group (0 light) will show the lowest O₂ production rate. [1 pt: correct direction + reference to control as comparison baseline]
AP® BiologySophriva · sophriva.com