Sophriva  ·  AP Biology  ·  Free sprint review
AP Biology Strategy · Statistics & Claims  ·  Strategy Playbook  ·  Free

Statistics, Error Bars & Evidence Claims

From data to defensible claim — writing conclusions that earn points

Eight pages on writing the conclusion sentence FRQ rubrics actually award. The bottleneck isn't computing chi-square — it's choosing which words are scientifically defensible and which words ('prove', 'definitely') are landmines.

Download the 10-page PDF Get your AP Bio FRQ marked free
Exam Weight
Q1·Q2·Q3·Q6 follow-ups
MCQs
n/a
FRQ Appearance
Every data FRQ
Sprint Time
~30 min

Quick Glance — all topics in this unit

Priority, exam format, and the trap that most often costs the point. Transcribed from page 1 so this content is searchable.

TopicPriorityFormatKey trap / must-know
P1 Data → Claim Pipeline★★★Q1·Q2·Q3·Q6Data → analysis → conclusion sentence — the conclusion is the rubric-earning step.
P2 SD vs. SE vs. 95% CI★★★All data FRQsSD = spread; SE = uncertainty in the mean; CI = range likely to contain true mean.
P3 Error Bar Overlap — 3 Cases★★★All data FRQsNo overlap = significant; partial = inconclusive; full overlap = not significant.
P4 Chi-Square 6-Step Protocol★★★Genetics FRQsHypothesis → expected → χ² calc → df → critical value → conclusion language.
P5 p-Value Language Traps★★★All data FRQsNever 'proves'; use 'supports' or 'fails to reject the null'.
P6 Correlation ≠ Causation★★Q3·Q6Strong correlation alone never establishes causation.
P7 CER Templates★★★Q3·Q6Claim · Evidence · Reasoning — three templates: support, refute, inconclusive.
P8 Practice + Cheat Card★★All data FRQsSix must-know phrasings the rubric explicitly rewards.

Topic-by-topic visual review

Nine pages, one topic each. Open the PDF for print quality, or scroll to study on screen.

Cover for AP Biology FRQ statistics, error bars and evidence claims. Subtitle: from data to defensible claim — writing statistical conclusions that earn full points. Tags: Q1, Q2, Q3, Q6 follow-ups, Skills 5 and 6, statistical reasoning, argumentation. The bottleneck is the conclusion sentence, not the math.
Cover  ·  Statistics, error bars & evidence claims — writing conclusion sentences that earn FRQ rubric points.
Data to claim pipeline and SD versus SE versus 95% CI. Pipeline: raw data, descriptive statistics (mean and variability), analysis (test statistic and p-value or CI), conclusion sentence using rubric language. Standard deviation SD measures spread within a sample. Standard error SE measures uncertainty in the estimate of the mean — gets smaller with larger sample size. 95% confidence interval CI is the range likely to contain the true population mean 95% of the time.
Data → Claim Pipeline + SD vs. SE vs. 95% CI  ·  How AP grades go from raw data to written claim — and what each common variability statistic actually measures.
Error bar overlap three cases. Case 1 non-overlapping bars: difference between means is likely significant, claim 'a significant difference exists.' Case 2 partial overlap: inconclusive without a formal test, claim 'difference may or may not be significant; statistical test required.' Case 3 fully overlapping: difference between means is not statistically significant, claim 'no significant difference' or 'cannot reject the null hypothesis.'
Error Bar Overlap — 3 Cases  ·  Non-overlapping = likely significant. Partial = inconclusive. Fully overlapping = not significant. Plus rubric language for each.
Chi-square six-step protocol with worked Mendelian cross example. Step 1 state null hypothesis: observed and expected ratios match. Step 2 compute expected frequencies from genetic ratio. Step 3 calculate chi-square as sum of (observed minus expected)^2 divided by expected. Step 4 determine degrees of freedom equals categories minus one. Step 5 look up critical value at p equals 0.05. Step 6 state conclusion: if chi-square less than critical value, fail to reject the null.
Chi-Square: The 6-Step Protocol with Worked Example  ·  State hypothesis → compute expected → calculate χ² → determine df → look up critical value → state conclusion in rubric language.
p-Value language traps and sample size. Avoid landmine words: prove, definitely, certainly, always. Use rubric-safe language: supports, is consistent with, fails to reject the null hypothesis, suggests, provides evidence for. Sample size and standard error inverse relationship: SE equals SD divided by square root of n; larger n means smaller SE, tighter confidence interval.
p-Value Language Traps + Sample Size  ·  Never 'prove' or 'definitely.' Use 'supports', 'consistent with', 'fails to reject the null'. Plus the SE-vs-n inverse relationship.
Correlation does not equal causation. Strong correlation between two variables does not establish that one causes the other. Confounding third-variable problem: lurking variable causes both A and B. To establish causation requires controlled experimental manipulation, not observational correlation. Rubric-safe sentence: variables are correlated; further controlled experiments are needed to establish causation.
Correlation ≠ Causation  ·  Strong correlation alone never establishes causation. The third-variable problem and how to write the careful conclusion.
Claim Evidence Reasoning CER templates. Template A supporting a claim with data: claim that variable A causes change; evidence with specific data points; reasoning linking evidence to claim. Template B refuting a claim: state competing claim; evidence contradicting; reasoning explaining why the data does not support. Template C inconclusive: state the claim that cannot be supported or refuted; evidence showing overlap or insufficient n; reasoning recommending additional experiments.
CER Templates — Support · Refute · Inconclusive  ·  Three Claim-Evidence-Reasoning templates that earn rubric points: supporting a claim, refuting a claim, and writing the careful inconclusive case.
Practice problems and one-page cheat card. Practice 1 error bar interpretation with worked answer. Practice 2 chi-square problem with full six-step protocol. Practice 3 correlation versus causation reasoning prompt. Cheat card with the six must-know phrasings: supports the hypothesis, fails to reject the null, consistent with, statistically significant at p less than 0.05, error bars overlap (or do not overlap), correlation not causation.
Practice + One-Page Cheat Card  ·  Three practice items (error bars, chi-square, correlation) with model answers, plus the six must-know FRQ phrasings.

Done with the sheet? Now get marked.

Upload your own AP Bio FRQ and we score it like an AP Reader — point by point — and show you exactly which rubric points you missed.

Get your first FRQ marked free →
AP® BiologySophriva · sophriva.com