Clue 1: Orient yourself
Which problem changes the measurement itself?
Use the labels and the picture's left-to-right, near-to-far, or before-and-after order. Name only what you can point to.
The question
Which problem changes the measurement itself?
Why it matters: Clinical recommendations affect real children and families. Fair comparisons, bias control, ethical limits, and honest uncertainty keep a promising result from becoming a harmful claim. Today you practice the professional reasoning behind that work: Bias is directional design or measurement error; confounding mixes effects; each threat needs a matched safeguard.
On your WebXam
Distinguishing confounding from measurement bias and matching the right defense
For life
To find the cause, change one thing and watch what changes.
Principle: Same look, different cause
Five principles we return to
Having it is not using it
Same instructions, different switches
Same look, different cause
Change one thing and watch
Boss or doorman?
Decides the result or only allows it
Held down, not gone
Remove the brake and it returns
Many roads, one ending
One result can begin many ways
Try the everyday version first
A crooked mirror and a hidden magnet distort results in different ways
A crooked mirror bends every reflection in a steady direction. A hidden magnet may pull only certain objects, creating a different pattern of error.
Do not jump to the biology yet. Treat the picture as a small system. Track its parts, follow one change at a time, and keep more than one explanation open until the picture supplies a way to separate them.
Clue 2: Trace one change
Which adds another cause linked to both sides?
Follow one object, stage, or path. Point to the first place where the situation changes instead of jumping to the ending.
Clue 3: Keep the cause open
Would a larger sample fix either problem?
List more than one explanation that still fits. Name the extra observation that would help you separate those possibilities.
Now inspect the illustration
Work from the visible evidence. A useful answer names the part of the picture that supports it and leaves unknown causes open.
- 1Which problem changes the measurement itself?
- 2Which adds another cause linked to both sides?
- 3Would a larger sample fix either problem?
Tier 1 check
Finish with the everyday model
Use the everyday picture to answer today's question in plain words: Which problem changes the measurement itself?
You can complete today's required check without opening the technical details below.
Ready for the real names? Optional tier 2
Technical rules and limits
Rule 1: Name the direction and stage of each bias.
Rule 2: Draw the confounder linked to exposure and outcome.
Rule 3: Choose a safeguard that fits the threat.
Where the analogy stops: Real studies can have many interacting biases and unmeasured confounders, not one visible mirror or magnet.
Carry the previous idea forward
An operational definition makes speech scoring repeatable, while patient-reported outcomes capture effects clinician scores can miss.
Today's technical takeaway
Bias is directional design or measurement error; confounding mixes effects; each threat needs a matched safeguard.
Now map the same rules onto biology
Red-team a speech study for selection, measurement, and confounding
Crooked mirror
Measurement bias
Missing participants
Selection and attrition bias
Hidden magnet
Confounder linked to exposure and outcome
Educational illustration, not a clinical photograph or a patient-specific study plan. Use the supplied evidence cards and claim ceiling.
Mateo's case file: evidence supplied in this lesson
EXP16-E1
Earlier-repair participants are more likely to receive care at high-volume centers with strong speech services.
Why it matters: Center resources could confound an observational comparison.
EXP16-E2
Assessors know each child's group and expect earlier repair to work.
Why it matters: Expectation could create directional measurement bias.
EXP16-E3
Follow-up is 92 percent in one group and 71 percent in the other, with access-related reasons.
Why it matters: Differential attrition could change the comparison.
Make the clinical decision
You are the adversarial methods reviewer before data lock.
The team says a sample of 5,000 makes bias and confounding disappear.
AMatch each supplied threat with a design, measurement, or analysis safeguard.
BAccept the claim because large samples remove systematic error.
CList every possible bias without changing the protocol.
Choose the red-team response and propose one matched safeguard per evidence card.
Evidence required
EXP16-E1 + EXP16-E2 Claim ceiling
You may reduce named threats. You may not claim adjustment removes all unmeasured bias or confounding.Go deeper Optional tier 3
Everything required for today is above. Open these only if you want the explainer, source trail, or download files.
The everyday model and Tier 1 check are the complete required path for this lesson.
Use these checks to keep your place. They are not turned in through the portal.
Check off as you finish
- Worked through the everyday picture and answered its three questions.
- Completed the Tier 1 check in plain words.
Turn in: Experimental Design lesson 16: What Could Fool Us Into the Wrong Conclusion?
Go to Schoology to turn this in.
Submit one PDF. Put your first and last name in the document header. Name the file: FirstName LastName - Assignment Title - YYYY-MM-DD.pdf.
Open Schoology PDF upload helpIf you cannot get in, see Mr. Mendoza. Do not skip the work.
Optional legacy technical materials Open only if you want the original notes, vocabulary, artifact, and CER work
Before this page, you should know
- An is the specific, defined thing a study counts; an is the exact written rule for what you observe and how you score it.
- TOPS did not score 'good speech'; it scored a velopharyngeal composite (VPC-Sum) from 0 to 6 on a fixed single-word test, with 4 or higher defined as insufficiency.
If not, review
Today's new idea is only
Bias is a built-in one-direction error in how a study is run or measured, while confounding is a hidden third variable; good design defends with standardization, , randomization, and matching.
Goal: Identify bias, confounding, and the surgeon-as-confounder problem in a real study, and name the design defenses (matching, randomization, ).
Know by the end
- Bias is a built-in error in how a study is run or measured that pushes results one direction; (a measurement bias) and survival bias (a selection bias) are two examples.
- Confounding happens when a hidden third variable is tangled with both the thing you changed and the thing you measured, so it can masquerade as the cause.
- In a real UCLP trial the dental-arch outcome tracked with which surgeon operated, not the technique, the surgeon-as-confounder problem; a skilled surgeon got good results with either technique.
- Design defenses differ by trap: standardize and blind the measurement to fight bias; randomize, match, or measure and adjust to fight confounding; you can blind the judge but never the surgeon.
1
Bias and confounding, kept separate
Model start: A study can give the wrong answer in classic ways, and good design fights each one. Anchor the difference: bias is in HOW you run or measure the study; confounding is a real third variable.
- Bias is a built-in error that pushes results one direction; mothers of affected babies recalling exposures more intensely is ____ bias, a kind of measurement bias.
- Confounding happens when a hidden ____ variable is tangled with both the thing you changed and the thing you measured.
- In the surgeon trial, the ____ was tangled with the technique and was the real driver of dental-arch growth (the surgeon-as-confounder problem).
2
Matching defenses to traps
- To fight bias (a built-in, one-direction error), standardize and ____ the measurement.
- To fight confounding (a hidden third variable), ____, match, or measure and adjust for it.
- The hardest honest truth in surgery: you can blind the judge, but you cannot blind the ____, who always knows the operation they are doing.
Everything you need for today is on this page. These links are optional.
What to read
Read the short plain-language explanation written for this lesson. Plain-language explainer for this lesson
Why this source matters
This explanation gives you the background for today's idea without making you decode a research paper: Bias is a built-in one-direction error in how a study is run or measured, while confounding is a hidden third variable; good design defends with standardization, , randomization, and matching.
Words to unlock first
biasconfoundingblindingrecall biassurgeon-as-confounder
Reading moves
- Skim the title and abstract first to get the gist.
- Circle the one sentence that states the main claim.
- Box the evidence the authors give for that claim.
- Mark one sentence that confuses you, and move on.
Stop point
Stop after the final 'Use it now' section. The research citations are available separately for advanced readers.
Your output
Write one claim-evidence sentence: state the main idea, then name the example or evidence that supports it.
Where this fits
Tested on (Ohio WebXam)
Genetics of Disease · 072130
PLTW lesson
MI · Experimental Design domain · Bias, confounding, and the surgeon-as-confounder problem; matching design defenses to traps
WebXam domain
Molecular and Genetic Technology
Evidence to produce
You are reviewing a draft claim about Mateo's care: 'Technique A repairs cleft palates better than Technique B.' As biostatistician, do three things: (1) name one confounder that could fake this result (hint: think surgeon); (2) name one design step that would defend against it; (3) decide whether the team should publish as written or revise, and write one sentence of advice to the PI.
Lab / skill
Biomedical Innovations (BI) · AP Biology
Everything required for today's decision is already in the case file and plain-language explainer. The links below are original papers and database records for teachers and advanced readers, not assigned student reading.
- STROBE. Reporting guidance for observational studies.
- Ozawa et al. 2020, Surgical technique and timing in UCLP (surgeon as driver, Orthod Craniofac Res)
- Sabbagh et al. 2023, COVID-19 risk factors and orofacial clefts, five-country case-control (BMC Oral Health)
- Grosen et al. 2011, Risk of oral clefts in twins (survival bias in registries, Epidemiology)
- Claim: In a study of two surgical techniques, the result we see may be caused by something other than the technique.
- Evidence: In a real UCLP trial, dental-arch outcome tracked with ____ rather than technique.
- Reasoning: Explain how this confounder could fool a team, and name one design defense that would have caught it.
How this is graded (rubric)
For: You are reviewing a draft claim about Mateo's care: 'Technique A repairs cleft palates better than Technique B.' As biostatistician, do three things: (1) name one confounder that could fake this result (hint: think surgeon); (2) name one design step that would defend against it; (3) decide whether the team should publish as written or revise, and write one sentence of advice to the PI.
| Criterion | Proficient | Developing | Beginning |
|---|---|---|---|
| Complete | Every required part of the artifact is present and filled in. | Most parts are present, but one is missing or left blank. | Several parts are missing. |
| Accurate | The science and data are correct and match the evidence. | Mostly correct, with a small factual slip. | Key science or data is wrong. |
| Scientific reasoning (CER) | States a claim, backs it with specific evidence, and explains the reasoning. | Has a claim and evidence, but the reasoning is thin or missing. | Gives an answer with no evidence or reasoning. |
| Professional communication | Clear, organized, and labeled the way a clinician or scientist would write it. | Readable but disorganized or missing labels. | Hard to follow. |
| Submitted | Turned in through the route named under Submit here and confirmed. | Turned in, but in the wrong place or unconfirmed. | Not turned in. |
How the model answer scores against this rubric
- CompleteProficient: Nothing is left blank: the model fills every part of "You are reviewing a draft claim about Mateo's care: 'Technique A repairs cleft palates better than Technique B.' As biostatistician, do three things: (1) name one confounder that could fake this result (hint: think surgeon); (2) name one design step that would defend against it; (3) decide whether the team should publish as written or revise, and write one sentence of advice to the PI.".
- AccurateProficient: Every number and claim matches the case evidence.
- Scientific reasoning (CER)Proficient: It names a claim, cites the specific evidence, and explains the reasoning, not just the answer.
- Professional communicationProficient: It is organized and labeled like a real chart note.
- SubmittedProficient: It would be attached to your class form or handed in, and confirmed.
What's next: We answered today's question: bias, confounding, and weak can all fool us. But even a careful single study can be a fluke or can disagree with the next study. How do we gather every study on a question and combine them into one trustworthy answer? We chase that next time.