A researcher believes participants in a therapy study might improve just from knowing they're receiving treatment. She ensures participants don't know which condition they're in (treatment vs. placebo). Has she eliminated all psychologically generated bias?
AYes — single-blind design eliminates all expectancy effects in a study
BNo — she also needs to blind herself and other experimenters to condition assignment, to prevent their expectations from influencing data collection and interpretation
CNo — she should have used a within-subjects design to eliminate individual differences
DYes — as long as participants cannot guess their condition, demand characteristics are fully controlled
Single-blind design addresses demand characteristics (participants' expectations) but leaves experimenter bias unaddressed. Rosenthal's research showed that experimenters who know condition assignments can subtly influence outcomes through tone, timing, and data recording — all unconsciously. Double-blind is required precisely because the human capacity to generate self-fulfilling expectations operates on both sides of the experiment.
Question 2 Multiple Choice
Which of the following best explains why demand characteristics threaten validity even when participants sincerely try to be accurate?
AParticipants deliberately misreport their experiences once they figure out the hypothesis
BDemand characteristics only affect observable behavior, not self-reported measures
CParticipants unconsciously alter their behavior to match perceived expectations, making the dependent variable measure social compliance alongside — or instead of — the treatment effect
DDemand characteristics are only a problem in laboratory settings, not in naturalistic studies
The insidious feature of demand characteristics is that they operate automatically, not through deception. Participants are socially intelligent humans who pick up on cues and adapt their behavior accordingly — not because they're lying, but because this is how humans function in social contexts. The result is a confound: the dependent variable reflects the participant's reading of the situation, not just the treatment. Option A is wrong because dishonesty is not required.
Question 3 True / False
Experimenter bias is mainly a threat when researchers deliberately try to influence the outcome of their study.
TTrue
FFalse
Answer: False
Rosenthal's classic experiments showed that experimenter effects operate largely unconsciously. Researchers who expected better maze performance from their rats produced better results through subtle differences in handling and timing — not through deliberate manipulation. In human studies, this extends to tone of voice, phrasing of questions, and interpretive choices in coding responses. Unconscious bias is more dangerous than deliberate bias precisely because it is harder to detect and correct.
Question 4 True / False
Double-blind procedures are considered the gold standard for eliminating expectancy effects because they prevent both participants and experimenters from knowing condition assignments.
TTrue
FFalse
Answer: True
Single-blind addresses one source of bias (participants); double-blind addresses both. The logic is that psychologically generated confounds can arise from either side — participants who know their condition may respond to the expectation rather than the treatment, and experimenters who know condition assignments may inadvertently transmit expectations or differentially interpret ambiguous data. Blocking knowledge on both sides removes both pathways.
Question 5 Short Answer
Why is double-blind considered a stronger design than single-blind, and what specific mechanism does the added layer of blinding address?
Think about your answer, then reveal below.
Model answer: Single-blind keeps participants ignorant of their condition to control demand characteristics. Double-blind adds experimenter ignorance to also control experimenter bias — the tendency for researchers' expectations to subtly influence their behavior, data collection, and interpretation. The added layer addresses the Rosenthal effect: unconscious expectancy effects transmitted from experimenter to participant through tone, phrasing, and handling. Because this mechanism operates below conscious awareness, it cannot be corrected through good intentions alone — structural ignorance is required.
The practical implication is that double-blind designs require coded materials, standardized scripts, and separate chains of knowledge so that the person measuring outcomes does not know what the intervention was. This complexity is what makes double-blind studies expensive and difficult — and also why they are trusted over single-blind designs for high-stakes claims.