BIAS & CAUSALITY
What is bias in psychological research?
A systematic error that distorts results or interpretations.
What is construct validity?
The extent to which a measurement actually assesses what it claims to measure.
Give one example of bias that creates ethical concern.
Cultural or gender bias, which can lead to stereotyping or unfair conclusions.
What ethical issue arises if participants are not informed about how their data are measured?
consent
What does determinism mean in psychology?
The belief that behavior is caused by biological or environmental factors rather than free will.
what is selection bias?
Selection bias occurs when participants are not randomly assigned or are systematically different across groups in a study, which can distort the results.
In other words, they're not representative of the larger population.
What design is best for studying behavioral change over time?
Longitudinal design.
Name and explain two ethical principles researchers must uphold.
Informed consent, right to withdraw, protection from harm, confidentiality, debriefing.
Informed Consent – Participants must be fully informed about the study’s purpose, procedures, and any risks, and agree to take part voluntarily.
Right to Withdraw – Participants can leave the study at any time without penalty, even after agreeing to take part.
Protection from Harm – Researchers must ensure participants are not exposed to physical or psychological harm beyond everyday life experiences.
Confidentiality – Participants’ data and personal information must be kept private and not shared without permission.
Debriefing – After the study, participants are informed about the true purpose of the research, any deception is explained, and questions are answered.
Explain why ensuring anonymity is part of researchers’ responsibility.
It protects privacy and prevents harm from personal data exposure.
Why can’t a correlational study establish causality?
Because it doesn’t manipulate variables or control for confounding factors.
Only experimental designs that manipulate the IV and control for confounds can establish causality.
Example: There is a positive correlation between ice cream sales and sunburns.
This does not mean eating ice cream causes sunburn.
A third variable, hot sunny weather, affects both ice cream sales and sunburns.
Key point: Correlation = relationship, not causation.
What's the difference between extraneous variables and confounding variables?
Extraneous Variables:
These are any variables other than the independent variable (IV) that could affect the dependent variable (DV).
They may or may not actually influence the DV, but they are potential sources of error that researchers try to control.
Example: Room temperature, participant mood, or time of day in an experiment.
Confounding Variables:
A type of extraneous variable that does systematically affect the DV and is linked to the IV, making it impossible to tell whether the IV or the confound caused the effect.
Example: In a study on energy drinks and memory, if the energy drink group is always tested in the morning and the control in the afternoon, time of day is a confounding variable because it varies systematically with the IV.
All confounding variables are extraneous, but not all extraneous variables are confounding variables.
Define internal validity and external validity in one sentence each.
Internal validity = accuracy of causal inference; External validity = ability to generalize findings.
True or False: Avoiding deception is always more ethical, even if it introduces bias.
False – sometimes minimal deception is justified if participants are debriefed.
Explain the difference between nominal, ordinal, and ratio scales in quantitative measurement.
Nominal = categories; Ordinal = ordered ranks; Ratio = continuous with a true zero.
Explain how cultural relativism challenges universalism when interpreting research findings.
Relativism argues behavior must be understood in its cultural context; universalism assumes general laws apply to all, risking cultural bias.
Explain how participant expectations can interfere with establishing causality. Give an example
they may alter behavior (demand characteristics), creating false cause–effect links.
Differentiate between ecological validity and population validity and give an example
Ecological = realism of setting; Population = representativeness of sample
How can cultural or gender bias each reduce the validity and generalizability of psychological research findings?
Cultural bias occurs when research methods, interpretations, or conclusions reflect the norms and values of one culture. This can reduce construct validity because the behavior measured may not have the same meaning in other cultures. It also limits generalizability, since findings may not apply to people from different cultural backgrounds.
Gender bias happens when research disproportionately focuses on one gender or interprets results from a gendered perspective. This can reduce internal and external validity, as the study may not accurately reflect how the phenomenon occurs in all genders, and findings may not generalize beyond the studied group.
If a study finds a p-value of 0.03, what does this say about statistical significance and the null hypothesis?
The result is statistically significant (p < 0.05); the null hypothesis is rejected.
Give one factor that threatens internal validity and therefore weakens causality.
Confounding variables, lack of random assignment, or poor control.
In a memory study, one group listens to classical music while learning words, and another studies in silence. The music group is tested in the morning, while the control group is tested in the afternoon.
Discuss how bias and confounding variables affect causality and internal validity in this study.
Confounding Variable – Time of Day:
The music group is tested in the morning, and the control group in the afternoon.
Time of day can affect alertness, concentration, and memory performance.
This confound makes it unclear whether differences in memory are due to the music or the time of testing, reducing internal validity.
Potential Participant Bias:
Participants might behave differently because they know they are being observed (reactivity) or want to perform well (social desirability), although this is less directly indicated in the scenario.
Impact on Causality:
Because of the confounding variable, we cannot confidently claim a causal relationship between listening to music and memory performance.
The study’s design does not isolate the independent variable (music), so causality is compromised.
A researcher measures self-esteem using one scale before therapy and a different one after. Scores improve slightly.
Discuss whether real change occurred and how measurement issues affect interpretation.
Construct Validity:
Using two different scales may measure slightly different aspects of self-esteem.
Improvement in scores might reflect differences between the scales rather than actual change in self-esteem.
Reliability:
If the scales are not equally reliable, fluctuations could be due to measurement error rather than true change.
Internal Validity / Change:
Because the measurement method changes, it’s unclear whether the observed improvement represents real change in participants’ self-esteem or is just an artifact of the instruments used.
A researcher studying intelligence excludes participants from rural schools because “they might not understand instructions.”
Discuss how bias and ethical responsibility are intertwined in this decision.
Bias:
Excluding participants from rural schools introduces sampling bias, because the sample is no longer representative of the population.
It may also reflect cultural bias, assuming that rural students are less capable of understanding, which is unjustified and stereotyping.
Ethical Responsibility:
Researchers have a duty to treat participants fairly and avoid discrimination.
Excluding participants based on assumptions about their background violates ethical principles of justice and equality.
Interconnection:
Bias in selection directly conflicts with ethical responsibility, as the researcher’s assumptions not only distort results (reducing validity and generalizability) but also lead to unfair treatment of participants.
A company collects employees’ “happiness scores” through an app but doesn’t explain how scores are calculated or used. Some employees feel anxious about being monitored.
Discuss this situation in terms of ethical responsibility and measurement.
Ethical Responsibility: The company fails to provide informed consent and protect participants from psychological harm, since employees feel anxious about being monitored.
Measurement: Lack of transparency about how happiness scores are calculated reduces construct validity, making it unclear if the scores truly reflect employee well-being.
Overall: Ethical responsibility and measurement are interconnected—transparent, valid measures are necessary to respect participants and produce meaningful data.
A study concludes that depression is caused solely by low serotonin levels, ignoring life events and social support.
Discuss how the chosen perspective limits causal explanation and how integrating perspectives could improve validity.
Limitations of the Chosen Perspective:
Focusing only on biological factors (low serotonin) is a reductionist approach, ignoring environmental, cognitive, and sociocultural influences.
This limits causal explanation, because depression is likely influenced by multiple interacting factors, not just biology.
Improving Validity through Integration:
Incorporating other perspectives—e.g., cognitive (thought patterns), sociocultural (life events, social support)—provides a more holistic understanding.
This enhances internal validity (by considering multiple causes) and external validity (by making findings more generalizable across contexts).