Causality in Psychological Research

Understanding and Determining Causality

  • A complex topic discussed throughout psychology degrees.
  • Understanding the foundations of causality is essential.

Why Study Psychology?

  • Interested in people, brains, and behavior.
  • Understanding behavior involves asking "why" questions.
    • Why are we prejudiced?
    • Why do we treat people differently?
    • Where do attitudes come from?
    • Why do individual differences exist?

How of Behavior

  • Explaining how we perform tasks or engage in activities.
    • How do we remember names?
    • How do children learn to talk?
    • Focus on the processes enabling these actions.

Understanding Mechanisms of the Mind

  • Goal: understand mental processes and preconditions that influence behavior.
  • "Why" and "how" are linked to understanding cause and causality.

Determinism and Causality

  • Determinism: all events have a cause.
    • Precedent event creates the subsequent event.
    • Psychology operates under this assumption.
  • Sequential timing of events:
    • Event A precedes Event B; A potentially causes B.

Common Sense Perspective

  • Events have a sequence, challenge arises when sequence is unclear

Examples of Causality

  • Physics example: Kicking a ball.
    • Speed and force of the kick determine the ball's trajectory.
  • Camping example: Rain pooling on a tarp, leading to collapse.
  • Supermarket flooding due to heavy rain.
  • Psychological example: Traumatic event leading to new symptoms.
    • Phobia development: Dog barking at a child leads to fear of dogs.

Natural Inclination to Attribute Cause

  • Attributing cause even when it may not exist.
  • Infrequent events and food consumption:
    • Eating unusual food at a fair, feeling sick after a ride, attributing sickness to food when it was due to the ride.
  • Biological tendency to avoid foods associated with sickness.
  • Consumption of spirits:
    • Trying a new spirit, drinking too much, feeling sick, and blaming the spirit instead of the alcohol.
    • Alcohol is tasteless, so the cause is attributed to what was tasted.
    • This bias is to keep us safe but shows our inclination to associate events.

Challenges in Identifying Cause

  • Complicated world with multiple variables.
  • Identifying a relationship between factors in this matrix is challenging.
  • Not all of psychology works on a simple A causes B model.

Personality Traits

  • Stable tendencies influencing behavior.
  • Extroversion influencing social engagement.
  • Variables are related to each other.
  • Change in one variable is associated with change in another.

Measurement

  • Relationships require measurement.
  • Charting relationships requires measuring variables.

Example: Height and Weight

  • Height in meters vs. weight in kilograms.
  • Taller people tend to be heavier.
  • Positive, linear relationship.

Psychological Variables

  • Challenges in measuring psychological variables.
  • Reliability and validity of measurement.

Example: Extroversion and Social Events

  • Replacing height with extroversion and weight with the number of parties attended.
  • More extroverted individuals attend more social events.

Clinical Example: Behavioral Response

  • Measurement across two phases: baseline and post-treatment.
  • Example: Agoraphobia and going outside.
  • Treatment increases effort to go outside.

Importance of Measurement

  • Essential for charting relationships and applying interventions.
  • If variables cannot be reliably measured, charting is pointless.

Experimental Designs

  • Investigating causality.
  • Example: Noise level and text comprehension.

Approaches

  • Naturalistic observation:
    • Identifying different environments with different noise levels.
    • Participants read and answer questions; compare performance in different environments.
  • Controlled environment:
    • Standardized lab setting.
    • Manipulate noise levels, group comparison of performance.
    • Infer relationship between noise and comprehension.

Experiments and Causality

  • Experiments investigate causality.
  • Predisposing factor changes performance or mental processes.
  • Deliberate attempts to change a variable and observe change in a second variable.

Time-Based Frame

  • Change input variable (noise level) and identify change in comprehension.

Random Allocation

  • Important to eliminate systematic bias.
  • Ensures groups are similar, varying on more than just baseline comprehension.
  • Even out error from uncontrolled factors.

Quasi-Experiments/Differential Research

  • Cannot randomly allocate to groups.

Examples

  • Sex or gender groups.
  • People of different religions.
  • Ethical issues: comparing smokers vs. non-smokers.
  • Effects of domestic violence.

Limitations

  • Quasi-experiments are not as strong as true experiments.
  • Other factors associated with groups might contribute to differences.

Example: Smokers vs. Non-Smokers

  • Argument: chemicals from smoking are protective for cognition.
  • Alternative: smokers take more breaks, influencing focus.
  • Most research investigates relationships between variables and causality.

Causal vs. Non-Causal Relationships

  • Some relationships are causal, others are not.
  • Example: Marks in two different courses; correlation exists because of the individual, but one does not causally affect the other.

Laws in Psychology

  • Laws are mostly due to behaviorism.
  • Identifying orderly associations between events or variables.
  • Focus on what people do, not why.
  • Strong relationship between behavior and environmental contingencies.

Law of Distributed Practice

  • How we arrange learning relates to how much is remembered.
  • Spaced study beats cramming, applies across many contexts.
  • Clear, um, so what the study is, the more interspersed episodes of study that you do.
  • For any particular topic. The more solid you're learning, right? If you cram and you cram at last minute, you're gonna lose a lot of what you thought you had by cramming, because it's not as effective as, um, um, a learning strategy. OK.

Relationships Between Variables

  • Variation in one variable is related to variation in the second variable.

Example: Self-Esteem and School Achievement

  • Initial observation: higher self-esteem correlated with higher achievement.
  • Assumption: encouraging self-esteem would improve achievement.
  • Strategy failed because the relationship was misunderstood.
  • Achievement leads to self-esteem, not the other way around.
  • Lacked experimental evidence and misinterpreted the results.

Compare with: Background Noise and Comprehension

  • Experiment provides stronger evidence for noise level as a factor in comprehension.
  • Deliberate change in noise level shows impact on comprehension performance.

Nature of Relationships

  • Relationship does not imply temporal order.
  • Association observed, nature unknown.
  • Functionally relate variables: level on one is a function of the other.

Mathematical Representation

  • y = f(x)
  • Reading comprehension performance is a function of noise level.

Complexity

  • Reading comprehension also depends on vocabulary and reading ability.
  • Reading comprehension is a function of noise, vocabulary, and other factors.
  • Better picture requires considering multiple variables.

Importance of Measurement

  • All of this relies on good measures of the variables.

Example: Aggressive Behavior

  • Level of aggressive behavior is a function of various factors.

Shout it out

  • What do you think it might be a function of?
  • Personalaty, Mood, Stress.

Potential Variables

  • Tolerance of frustration
  • Ability to inhibit urges
  • Upbringing
  • Family dynamics
  • Cultural dynamics
  • Childhood experiences

Breaking Down Variables

  • Upbringing broken down into specific, measurable aspects
  • Need decent measures for both aggressive behavior and related variables

Conditions vs. Causes

  • Psychology focuses on conditions contributing to behavior.
  • Different from physical cause-and-effect.

Necessary and Sufficient Conditions

  • Being deprived of oxygen is sufficient to cause death, but not necessary.
  • Availability of oxygen is necessary for fire, but not sufficient.

Conditions for Behavior

  • Particular behavior requires certain conditions to be present.
  • Examples: genetics, environment, brain disorder.

Genetics and Environment

  • Genetic variation can be sufficient and necessary for a disorder.
    • Down syndrome (Chromosome 23 issue)

Mental Health

  • Genetic propensity + environmental circumstances can lead to mental health issues.
    • Example: Flu during first trimester increases likelihood of schizophrenia, but it's neither necessary nor sufficient.

JS Mill's Methods for Determining Cause

  • 19th-century philosopher.
  • Can only determine probable cause due to uncontrollable factors.
  • Can determine what is not the cause.
  • Experiments use a method of difference.
  • Cannot directly observe cause, must observe indirectly.

Method of Difference

  • Conditions A, B, C, D, E lead to phenomenon P.
  • If removing A results in no P, A is likely a probable cause.

Challenges in Determining Cause

  • Relationships imply mechanisms.
  • Important to explain how things happen.
  • Knowing why and how can help treat mental health disorders

Challenges

  • Need to be careful if relationship is not experimental because If we've got the relationship the wrong way around, then it's not gonna work.
  • Experimental evidence for a cause does not necessarily mean that variable A directly causes variable B.
  • Intervening variables may exist.

Traffic Light Problem

  • Amber proceeds red but does not cause it.
  • Internal timing mechanism.
  • Correlation does not show causation.

Directionality

  • Consider directionality when considering cause.
  • Self-esteem example.

3rd Variable Problem

  • A 3rd variable influences both variables, causing them to appear related.

Tobacco Industry Example

  • Tobacco industry argued against smoking causing cancer, suggesting a third variable (anxiety) might influence both behavior and condition, rather than a causal link directly from smoking to cancer