Repeated Measures (Within-Subjects) Design – Q&A NOTES

(Based on Repeated Measures Design PDF)

SECTION 1: What an Experimental Design Does

Q1. What is an experimental design in research methods?

A: An experimental design is the plan that explains exactly how a study tests whether an
independent variable (IV) produces changes in a dependent variable (DV). Its
purpose is to isolate the effect of the IV while minimising alternative explanations.

Q2. What is internal validity?

A: Internal validity refers to the degree to which we can confidently conclude that
changes in the dependent variable were caused by the independent variable rather than
by extraneous factors.

Q3. How does experimental design affect statistical power?

A: Design decisions influence statistical power, which is the ability to detect a true
effect. Designs that reduce error variance, such as repeated measures designs,
generally increase power.

Q4. What are the two most common experimental designs in psychology?

A:

1. Between-subjects designs – different participants are used in each condition

2. Within-subjects (repeated measures) designs – the same participants take
part in all conditions

SECTION 2: Independent and Dependent Variables

Q5. What is the independent variable (IV)?

A: The independent variable is the condition, treatment, or situation that is

systematically manipulated by the researcher (e.g. teaching method, drug dosage).

Q6. What is the dependent variable (DV)?

A: The dependent variable is the measurable outcome used to assess the effect of the
independent variable (e.g. test scores, reaction time, symptom severity).

Q7. How is the DV used in a repeated measures design?

A: In a repeated measures design, the same DV is measured multiple times for the
same participants, allowing direct comparison of performance across conditions.

Q8. Why is timing important in repeated measures designs?

A: Timing strongly influences results. If the IV is delivered inconsistently or the DV is

measured differently across time, observed differences may reflect measurement
problems rather than true treatment effects.

SECTION 3: Within-Subjects (Repeated Measures) Design

Q9. What is a within-subjects (repeated measures) design?

A: A within-subjects design compares two or more conditions by measuring the same

group of individuals under every condition being studied.

Q10. Why are comparisons made “within” subjects rather than “between”
subjects?

A: Because each participant experiences all conditions, comparisons are made within
the same individual rather than across different groups.

Q11. How complex can repeated measures designs be?

A: They can include:

• Two or more levels of one IV

• Three or more levels of one IV
• Two or more IVs (factorial designs)
• More than one DV (multivariate designs)

SECTION 4: Between-Subjects vs Within-Subjects Designs

Q12. What is the key difference between between-subjects and within-subjects

designs?

A:

• Between-subjects: different people experience different conditions

 • Within-subjects: the same people experience all conditions

Q13. What is a major advantage of within-subjects designs?

A: Stable individual differences (e.g. intelligence, baseline ability) are automatically

controlled because the same participants are compared to themselves.

Q14. What is a major disadvantage of within-subjects designs?

A: Performance in later conditions may be influenced by earlier conditions, creating

order effects and carryover effects that threaten internal validity.

Q15. How should researchers choose between the two designs?

A: The choice depends on whether time-related problems (order, fatigue, carryover) can
be managed without introducing new confounds.

SECTION 5: Individual Differences and “Noise”

Q16. What are individual differences in experiments?

A: Individual differences include baseline ability, motivation, anxiety, fatigue, and

familiarity with tasks, all of which can influence outcomes.

Q17. Why do individual differences create “noise” in between-subjects designs?

A: Because differences between participants add error variance, making it harder to

detect true effects of the independent variable.

Q18. How do repeated measures designs reduce noise?

A: Each participant is compared to themselves, so baseline differences do not distort

comparisons across conditions.

Q19. What type of noise remains in repeated measures designs?

A: Time-based confounds such as:

• Order effects
• Practice effects
• Fatigue effects

Repeated Measures (Within-Subjects) Design – Q&A NOTES (Continued)

(Based on Repeated Measures Design PDF)

SECTION 6: Core Logic of Repeated Measures

Q20. What is the core logic behind repeated measures designs?

A: The core logic is that comparisons are based on change within the same individual,
such as performance in Condition A versus Condition B for the same person.

Q21. Why does using participants as their own controls strengthen comparisons?

A: Because background traits like intelligence, prior knowledge, and motivation remain
constant within a person, differences across conditions are less likely to be caused by
individual differences.

Q22. How do repeated measures designs affect statistical power?

A: They usually increase statistical power because removing variability due to individual
differences makes real treatment effects easier to detect with fewer participants.

Q23. When does the logic of repeated measures break down?

A: The logic breaks down when earlier conditions contaminate later conditions,
making comparisons unfair and threatening internal validity.

SECTION 7: Ordered vs Unordered Levels

Q24. What are ordered levels in a repeated measures design?

A: Ordered levels have a meaningful sequence, such as time points (Week 1, Week 2,
Week 3) or increasing dosage.

Q25. Why are ordered designs useful?

A: They allow researchers to examine trends such as improvement, decline, or plateaus

across time or exposure.

Q26. What is the main problem with ordered levels?

A: Time itself introduces changes such as fatigue, familiarity, and external events,
which can confound treatment effects.

Q27. Why is counterbalancing often impossible with ordered levels?

A: Because reversing the order would change the research question, especially when
time or dosage is the independent variable.

Q28. How should researchers handle threats in ordered designs?

A: By keeping timing consistent, maintaining stable measurement, and interpreting

results cautiously with awareness of history and maturation effects.

Q29. What are unordered levels in repeated measures designs?

A: Unordered levels have no meaningful sequence, such as completing different

questionnaires or tasks.

Q30. Why are unordered designs vulnerable to practice effects?

A: Because participants may learn strategies from earlier conditions that improve
performance in later ones.

Q31. What are contrast effects?

A: Contrast effects occur when judgments in one condition are influenced by what was
experienced in the previous condition.

Q32. How can researchers reduce problems in unordered designs?

A: By using counterbalancing, including breaks, or adding washout periods between

conditions.

SECTION 8: When Repeated Measures Is a Poor Choice

Q33. Why is repeated measures often inappropriate in early education research?

A: Young children change rapidly due to maturation and learning, so performance

improvements may reflect development rather than the treatment.

Q34. Why is reading instruction a classic poor fit for repeated measures?

A: Once children learn decoding skills, those skills cannot be removed, creating
irreversible carryover effects.

Q35. Why is a between-subjects design better in these cases?

A: It prevents one condition from benefiting unfairly from being taught second and
reduces time-related confounds.
SECTION 9: Types of Repeated Measures Designs

Q36. What is a simple repeated measures design?

A: A design with one independent variable and two levels, such as pre-test vs post-test.

Q37. What makes a repeated measures design “complex”?

A: Complexity increases when there are:

• More than two levels of the IV

• More than one IV (factorial designs)
• More than one DV (multivariate designs)

Q38. Why does complexity increase methodological risk?

A: Because order effects multiply and statistical assumptions such as sphericity

become more important.

SECTION 10: Single-Group Time-Series Design

Q39. What is a single-group time-series design?

A: A within-subjects design where the DV is measured multiple times before and after a
treatment is introduced.

Q40. Why are multiple pre-treatment measurements important?

A: They establish a baseline pattern, making it easier to see whether change occurs
only after treatment begins.

Q41. What advantage does time-series have over simple pre-post designs?

A: It allows detection of trends, gradual change, and inconsistent patterns that simple
pre-post designs cannot reveal.

Q42. Why does a time-series design still have weak internal validity?

A: Because external events (history threats) may occur around the same time as the
treatment and explain observed changes.

SECTION 11: Sequential Treatment Designs

Q43. What is a sequential treatment repeated measures design?

A: A design where two or more levels of the independent variable are applied
sequentially to the same participants, with the DV measured after each.

Q44. Why are sequential designs risky?

A: Because later treatments are tested on participants who have already been
influenced by earlier treatments.

Q45. What is a carryover effect?

A: A carryover effect occurs when the influence of one condition persists into later
conditions and affects performance.

Q46. How do carryover effects threaten internal validity?

A: They provide alternative explanations for differences between conditions that are
unrelated to the independent variable.

Repeated Measures (Within-Subjects) Design – Q&A NOTES (Final)

(Based on Repeated Measures Design PDF)

SECTION 12: Order Effects

Q47. What are order effects?

A: Order effects occur when the order in which conditions are experienced influences
participants’ performance, independent of the treatment itself.

Q48. What are the main types of order effects?

A: The main types are:

• Practice effects – performance improves due to familiarity or learning

• Fatigue effects – performance worsens due to tiredness or boredom
• Carryover effects – effects of one condition persist into the next

Q49. Why are order effects a serious threat in repeated measures designs?

A: Because differences between conditions may reflect order-related changes rather

than the effect of the independent variable.

SECTION 13: Counterbalancing

Q50. What is counterbalancing?

A: Counterbalancing is a technique used to control order effects by varying the

sequence in which participants experience conditions.

Q51. What is complete counterbalancing?

A: Complete counterbalancing involves using all possible orders of conditions

across participants.

Q52. Why is complete counterbalancing often impractical?

A: Because the number of possible orders increases factorially (e.g. 3 conditions = 6

orders; 4 conditions = 24 orders), requiring too many participants.

Q53. What is partial counterbalancing?

A: Partial counterbalancing uses only some possible orders, reducing participant

requirements while still distributing order effects.

Q54. What is reverse counterbalancing (ABBA)?

A: Reverse counterbalancing involves one group experiencing conditions in one order

(A–B) and another group in the reverse order (B–A).

Q55. When is counterbalancing inappropriate?

A: Counterbalancing is inappropriate when:

• The IV has ordered levels (e.g. time, learning stages)

• The treatment effect is irreversible

SECTION 14: Washout and Rest Periods

Q56. What is a washout period?

A: A washout period is a delay between conditions designed to allow effects of one

condition to dissipate before the next begins.

Q57. When are washout periods most useful?

A: They are especially useful in pharmacological or physiological studies where

treatment effects may linger.
Q58. What are the limitations of washout periods?

A: Some effects cannot be “washed out,” especially learning, attitude change, or

emotional reactions.

SECTION 15: Advantages of Repeated Measures Designs

Q59. What is the biggest advantage of repeated measures designs?

A: They control individual differences, reducing error variance and increasing

statistical power.

Q60. Why do repeated measures designs often require fewer participants?

A: Because each participant contributes data to all conditions, reducing the total
sample size needed.

Q61. Why are repeated measures designs efficient?

A: They are:

• Cost-effective
• Time-efficient
• Ethically preferable when participant recruitment is difficult

SECTION 16: Disadvantages of Repeated Measures Designs

Q62. What is the main disadvantage of repeated measures designs?

A: Vulnerability to order effects, fatigue, and carryover effects.

Q63. Why can repeated measures designs increase participant attrition?

A: Participants may drop out due to longer study duration or repeated testing demands.

Q64. Why can repeated testing distort measurement?

A: Repeated exposure can change how participants respond, especially on

questionnaires or cognitive tasks.

SECTION 17: Statistical Considerations

Q65. What statistical test is commonly used for repeated measures designs?
A: Repeated measures ANOVA is commonly used when comparing means across
multiple conditions.

Q66. What assumption is critical in repeated measures ANOVA?

A: The assumption of sphericity, meaning equal variances of the differences between

all condition pairs.

Q67. What happens if sphericity is violated?

A: It inflates Type I error rates, increasing the chance of false positives.

Q68. How can violations of sphericity be addressed?

A: By applying corrections such as Greenhouse–Geisser or Huynh–Feldt adjustments.

SECTION 18: Summary and Design Choice

Q69. When should a repeated measures design be used?

A: When:

• Individual differences are large

• Order effects can be controlled
• The treatment effect is reversible

Q70. When should a repeated measures design be avoided?

A: When:

• Learning or maturation is irreversible

• Carryover effects dominate
• Time itself is the confounding variable

Q71. How should repeated measures designs be justified in exams or reports?

A: By explicitly stating:

• Why individual differences needed control

• How order effects were addressed
• Why alternatives were less suitable
Experimental design is a fundamental component of psychological research, as it
determines how researchers test whether an independent variable produces changes
in a dependent variable. One commonly used design in psychology is the repeated
measures, or within-subjects, design. This design involves measuring the same
participants under all conditions of the independent variable. This essay explains the
logic of repeated measures designs, compares them with between-subjects designs,
and evaluates their advantages, limitations, and appropriate applications in research.

An experimental design specifies how a study isolates the effect of an independent

variable while minimising alternative explanations for observed changes in the
dependent variable. A key concern in experimental research is internal validity, which
refers to the extent to which changes in the dependent variable can confidently be
attributed to the independent variable rather than extraneous factors. Design choices
also influence statistical power, or the ability to detect a true effect when one exists.
Designs that reduce error variance generally increase statistical power, making
repeated measures designs particularly attractive.

In a repeated measures design, the same participants take part in all experimental
conditions, and the dependent variable is measured multiple times for each individual.
Comparisons are therefore made within participants rather than between different
groups. This approach contrasts with between-subjects designs, where different
participants are assigned to different conditions. Repeated measures designs can
range from simple pre-test–post-test designs to more complex arrangements involving
multiple levels of an independent variable, multiple independent variables, or multiple
dependent variables.

The core logic of repeated measures designs lies in using participants as their own
controls. Because the same individuals are measured across conditions, stable
individual differences such as intelligence, baseline ability, motivation, or personality
are automatically controlled. This reduction in individual-difference “noise” lowers
error variance and strengthens comparisons between conditions. As a result, repeated
measures designs typically have greater statistical power than between-subjects
designs and often require fewer participants to detect meaningful effects.

Despite these advantages, repeated measures designs are vulnerable to time-related

confounds. Order effects occur when the sequence in which conditions are
experienced influences performance independently of the treatment itself. Common
order effects include practice effects, where performance improves due to familiarity
with the task, fatigue effects, where performance declines due to tiredness or
boredom, and carryover effects, where the influence of one condition persists into
subsequent conditions. These effects threaten internal validity by providing alternative
explanations for observed differences between conditions.

The risk of order effects varies depending on whether the levels of the independent
variable are ordered or unordered. Ordered levels, such as time points or increasing
dosage, have a meaningful sequence and are often used to examine trends over time.
However, time itself introduces confounds such as maturation, history, and fatigue,
and counterbalancing is often impossible because reversing the order would change
the research question. Unordered levels, such as completing different tasks or
questionnaires, allow for counterbalancing but remain vulnerable to practice and
contrast effects, where responses in one condition are influenced by the previous
condition.

Several strategies can be used to reduce order effects in repeated measures designs.
Counterbalancing involves varying the order in which participants experience
conditions so that order effects are distributed evenly across conditions. Complete
counterbalancing uses all possible orders but is often impractical due to the rapid
increase in required sample size. Partial counterbalancing and reverse
counterbalancing provide more feasible alternatives. Washout or rest periods may also
be introduced between conditions to allow temporary effects to dissipate, although
some effects, such as learning, cannot be reversed.

Repeated measures designs are not always appropriate. They are a poor choice when
the effects of a condition are irreversible, such as in reading instruction or skill
acquisition, where learning in one condition inevitably benefits later conditions. They
may also be unsuitable in populations that change rapidly due to maturation, such as
young children, because improvements may reflect development rather than the
treatment itself. In such cases, between-subjects designs provide stronger internal
validity.

From a statistical perspective, repeated measures designs are commonly analysed

using repeated measures ANOVA when comparing means across multiple conditions.
A key assumption of this analysis is sphericity, which requires equal variances of the
differences between all pairs of conditions. Violations of sphericity inflate Type I error
rates and must be addressed using corrections such as Greenhouse–Geisser or
Huynh–Feldt adjustments.

In conclusion, repeated measures designs are a powerful and efficient experimental

approach in psychology. By controlling individual differences and increasing statistical
power, they allow researchers to detect effects with fewer participants. However, their
vulnerability to order effects, carryover effects, and time-related confounds requires
careful planning and justification. Repeated measures designs are most appropriate
when treatment effects are reversible and order effects can be controlled, and they
should be avoided when learning, maturation, or irreversible change threatens internal
validity.