PR

Research Methods in the Study of Abnormal Behavior

Research Methods in the Study of Abnormal Behaviour

What is Science?

  • The universe operates according to certain natural laws.
  • Things happen in an orderly way, and we can determine cause and effect.
  • These laws are discoverable and testable.
  • We can use these laws to make predictions and then experiment to see if we are right.
  • Results can be replicated and must occur under prescribed circumstances, not just once, but repeatedly.
  • Essential Beliefs

Psychology as a Science

  • Psychology is the science of studying mental processes and behaviors.
  • Pursuit of four main goals:
    • Describe
    • Explain
    • Predict
    • Control

Scientific Method

  1. Step 1: Identify questions of interest and consult literature.
  2. Step 2: Develop a testable hypothesis (must be operationally defined).
  3. Step 3: Select a research method, choose participants, and collect data.
  4. Step 4: Analyze the data and accept or reject the hypothesis.
  5. Step 5: Seek scientific review, publish, and replicate the study.
  6. Step 6: Build a theory.

Theories vs. Hypotheses

  • Theory
    • A well-developed set of ideas that aims to explain an observable phenomena.
    • Widely accepted among the scientific community and supported by data.
    • Primary goal of science is to advance theories to account for data, often by proposing cause-effect relationships.
  • Hypothesis
    • A testable prediction or proposed explanation for an observable phenomenon.

Qualitative vs. Quantitative Research

  • Qualitative Research
    • Does not test the relationship between variables.
    • Focuses on understanding the "what".
    • Explores new or uncommon phenomena.
    • Looks at the messiness of being human.
    • Aims to get the stories behind the numbers.
    • Not just words - artifacts like pictures, meeting minutes, diaries, music, art, etc.
  • Quantitative Research
    • Includes both correlational and experimental research.
    • Focuses on gathering numbers.
    • Analyzes data through statistical computations.
    • Can determine relationships and cause and effect.

Nomothetic vs. Idiographic Research

  • Nomothetic Research
    • Measuring a group of people on a number of variables.
    • Examines the relationship among the variables.
    • Correlational methods are an example.
    • Variable-centered.
  • Idiographic Research
    • Detailed understanding of the individual.
    • Case studies and qualitative methods are examples.
    • Person-centered.

Case Studies

  • The detailed study of one individual.
  • Providing detailed descriptions.
  • Collecting historical and biographical information.
  • Often includes details of therapy sessions.
  • Several case studies can be compared and analyzed for common elements through a specific method.
  • Can provide detailed insight and explanation into therapeutic techniques.
  • Example: Little Albert.

Epidemiology

  • Examines rates of occurrence of abnormal behavior in the population as a whole and in various subgroups (e.g., race, ethnicity, gender, age, or social class).
  • Provides a general picture of a disorder.
    • Prevalence: Proportion of a population that has the disorder at a given point or period of time.
    • Incidence: The number of new cases of the disorder that occur in some period, usually a year.
    • Risk factors: Conditions or variables that, if present, increase the likelihood of developing the disorder.

Correlational Research

  • Aims to examine the strength and direction of a relationship between two variables, focusing on how they covary or vary together.

Variables

  • A variable is any characteristic, number, or quantity that can be measured or counted.
  • It is the condition, event, or situation being studied.

Correlation vs. Causation

  • Understanding the difference between correlation and causation is crucial.

Correlation Coefficients

  • Examine the strength and direction of a relationship between two variables, focusing on how they covary or vary together.
  • Linear Relationships: Types of Linear Relationships.

Non-Linear Relationships

  • In a nonlinear, or curvilinear, relationship, as the X scores change, the Y scores do not tend to only increase or only decrease; at some point, the Y scores alter their direction of change.

Interpreting Relationships

  • Correlation coefficients may range between -1 and +1.
  • The closer to ±1 the coefficient is, the stronger the relationship.
  • The closer to 0 the coefficient is, the weaker the relationship.

Using Correlations

  • Correlation does not imply causation.
  • Third variables may be at play.
  • If a causal relationship does exist, be cautious of directionality.
  • Despite their limitations correlations are still important sources of information.
  • Smoking and lung cancer research started with a correlation.
  • Allows us to make predictions.

Experimental Research

  • Researcher typically begins with an experimental hypothesis.
  • Investigator chooses an independent variable (IV) that can be manipulated (different conditions – often experimental vs. control).
  • Participants are assigned to the conditions by random assignment.
  • Researcher arranges for the measurement of a dependent variable (DV).
  • Analyze the data to determine if there has been an experimental effect.

Independent vs. Dependent Variables

  • Independent Variable (IV): The variable that you manipulate.
  • Dependent Variable (DV): The variable that you measure (or the variable that is changed by the IV).

Experimental Design

  • Experimental group: Exposed to independent variable (new therapy)
  • Control group: (on wait list or original therapy)
  • Random assignment of participants
  • Measure dependent variable (pre-test) (depression scores)
  • Measure dependent variable (post-test)
  • Compare.

Statistical Significance in Experimental Research

  • Within-group variance: The difference within the groups
  • Between-group variance: The difference between groups
  • The experimental effect
  • Statistical significance is tested by dividing the between group variance by a measure of the within-group variance.
  • When the average difference between the two groups is large relative to the within-group variance, the result is more likely to be statistically significant.

Assessing Difference

  • We can also use tests that allow us to compare the means of groups to see if the differences are statistically significant (meaningful).
  • Tests of statistical significance: P-value or Probability Statistic
    • A number describing how likely it is that your data would have occurred by random chance.
    • A p-value less than 0.05 (typically ≤ 0.05) is statistically significant.
    • If you want to know how big the effect is - need to calculate the effect size.

Effect Size

  • P-value tells us how sure we can be that one thing is related to another or how sure we can be that two groups are different (the results are from group differences, not just random chance).
    • Example: happiness and parenting status - can we be sure they’re related?
    • Example: parents vs not parents - can we be sure that these two groups are actually different?
  • Effect size:
    • How strong is the relationship between the two variables?
    • How big is that difference?
    • We can use r (for correlations) or Cohen’s d (for group differences).

Examples:

  • Literacy and hours in training: r=0.78
  • Support skills and hours in training: r=0.58

Cohen's d

  • We use Cohen’s d to assess effect size when comparing the means (averages) of two experimental groups (e.g., treatment group vs. control group) or comparing two independent groups (e.g., men vs. women, before vs. after intervention).

Interpreting Cohen's d

  • Small effect: d=0.2, overlap = 83%
  • Medium effect: d=0.5, overlap = 67%
  • Large effect: d=0.8, overlap = 53%
  • Very large effect: d=1.4, overlap = 19%

Placebo Effect

  • An improvement in a physical or psychological condition that is attributable to a client’s expectations of help rather than to any specific active ingredient in a treatment.
  • Single-blind Procedures: When the patient or client is unaware of what group they have been placed in (placebo or treatment).
  • Double-blind Procedures: When neither the researchers nor the clients are aware of who has been placed in the treatment or placebo control groups.

Validity

  • Internal validity: Extent to which effect can be confidently attributed to manipulation of IV; Inclusion of at least one control group
  • Confounders: Effects are intermixed with the effects of the IV leading to internally invalid studies
    • passage of time could change scores on a stress scale; changes may not be due to the treatment
  • External validity: Can the results be generalized beyond immediate study?
  • Analogue experiments: The use of a related phenomenon (an analogue) in the lab
    • behaviour is rendered temporarily abnormal through experimental manipulations.

Meta-Analysis

  • Involves the review of many studies in order to determine the effects of treatment
  • Examine published studies, combine the results into a common format and then determine the extent of improvement (effect size).
  • Limitations:
    • It is a complicated process that requires decisions at each of numerous phases or steps.
    • Results of a meta-analysis are difficult to interpret.
    • Need to take into account moderator variables (for example, gender) that may influence or qualify the results in some meaningful way.