Testing Claims
An investigation must be judged on whether it actually measures what it intends to measure.
Validity
Too small a sample size fails to represent the population; too large a sample size becomes this.
Expensive and time‑consuming
A harmless treatment with no active ingredient used to compare against real treatments.
Placebo
Workers changing behaviour because they know they are being observed is known as this effect.
Hawthorne effect
Using emotional persuasion rather than evidence to sell products is known as this type of advertising.
Emotive advertising
Repeating trials, removing outliers and averaging results improves this quality of data.
Reliability
Choosing participants in a way that favours a particular outcome introduces this.
Sample selection bias
This effect occurs when patients improve simply because they believe they are receiving treatment.
Placebo effect
A 1991 study incorrectly linked hormone replacement therapy to coronary heart disease due to this mistake.
Confusing correlation with causation
This phenomenon occurs when positive perceptions of a person or brand cause rejection of valid alternatives.
Halo effect
This type of error occurs due to human mistakes such as transcription or timing errors.
Random error
Suppressing contradictory evidence or removing significant outliers is an example of this bias.
Cherry‑picking
Neither participants nor researchers know who receives the real treatment in this type of trial.
Double‑blind trial
This famous claim suggested listening to music could improve child development, despite flawed evidence.
The Mozart Effect
Astrology, numerology and iridology are examples of this type of claim.
Pseudo‑scientific claims
This type of error occurs when equipment is not calibrated or consistently biased.
Systematic error
Delaying publication because results are unfavourable is an example of this academic bias.
Publication bias
This group receives no treatment and provides a baseline for comparison.
Control group
False or humorous statistical relationships, such as “ice cream sales and shark attacks,” are known as this.
Spurious correlations
This scientific process helps detect fraud by requiring scrutiny, replication and evaluation.
Peer review
A sample size of one, often used in advertising or testimonials, is known as this.
Anecdotal evidence
Funding from organisations with vested interests can distort results through this type of bias.
Funding bias
This gold‑standard clinical trial design combines randomisation, blinding and placebo controls.
Randomised double‑blind placebo‑controlled trial (RDBPC)
Failing to control confounding variables leads to this incorrect conclusion about cause and effect.
Misinterpreting correlation as causation
Fake journals that accept nonsense submissions, such as the “Maggie Simpson” paper, contribute to this public issue.
Damage to public perception of science