The Scientific Method

Psychology Questions

•      We don’t always know how people will behave or what they think. We need to do research.

The Scientific Method

The scientific method is the way that scientific psychologists gain knowledge about behavior and mental processes.

–  The scientific method is not a particular technique or tool.

–  Instead, it is a general approach to gaining knowledge.

The scientific method (continued)

We can compare the scientific method to our “everyday,” nonscientific ways of gaining knowledge on several dimensions:

general approach                instruments

observation                         measurement

reporting                             hypotheses

concepts                             attitude

General Approach

Nonscientific

•      Intuitive

–  judgments and decisions are based on what “feels right.”

Scientific

•      Empirical

–  judgments and decisions are based on direct observation and experimentation.

Observation

Nonscientific

•      Casual, uncontrolled

–  personal biases and other factors influence observation.

Scientific

•      Systematic, controlled

–  control is the essential ingredient of science.

–  Scientists gain the greatest control when they conduct an experiment.

Observation (continued)

•      Control: Scientists investigate the effect of various factors one at a time in an experiment.

•      An experiment has at least one independent variable and at least one dependent variable.

–  Independent Variable (IV): A factor that researchers control or manipulate in order to determine the effect on behavior.

•   A minimum of two levels: The treatment (experimental) condition and the control condition

–   Example: In the Pennebaker and Francis (1996) study, the independent variable was whether students wrote about adjusting to college (experimental condition) or about superficial topics (control condition).

Observation (continued)

–  Dependent Variable (DV): The measure of behavior that is used to assess the effect of the independent variable.

•   Example:  In the Pennebaker and Francis (1996) study on the effects of emotional writing compared to superficial writing, one dependent variable was students’ Grade Point Average (GPA).

•   In most psychology research, several dependent variables are measured to assess the effects of the independent variable.

–   For example, Pennebaker and Francis also measured students’ health.

Reporting

Nonscientific

•      Biased, subjective

–  Personal impressions are reported.

Scientific

•      Unbiased, objective

–  observations and inferences are separate.

–  interobserver agreement is important.

Concepts

Nonscientific

•      Ambiguous

–  We aren’t clear in the meaning of the words we use.

–  For example, what do we mean by “intelligence”?

Scientific

•      Clear definitions

–  Define specifically what we mean by our concepts

–  A construct is a concept or idea used in psychological theories.

Constructs

•      There are many psychological constructs.

     Examples: aggression, depression, emotion, intelligence, memory, personality, stress, well-being.

•      An operational definition is the specific procedure used to produce and measure a construct.

Constructs (continued)

•     Advantages of operational definitions:

–  Allow scientists to define specifically what they mean by their construct

–  Allow clear communication among scientists.

•      Disadvantages:

–  A potentially limitless number of operational definitions exists for any particular construct.

–  Some operational definitions may be meaningless.

Instruments

Nonscientific

•      Inaccurate, imprecise

–  for example, diaries, gossip, rumors and opinions

Scientific

•      Accurate, precise

–  Accuracy: difference between what an instrument says and what is actually true

–  Precision: measures have different levels of precision.

Measurements

Nonscientific

•      Not valid or reliable

–  measures of our concepts that are inaccurate or inconsistent.

Scientific

•      Valid and reliable

–  valid measures get at the truth,

–  reliable measures are consistent.

Measurements (continued)

•      Physical measurement involves dimensions that have agreed-upon standards and instruments.

–  Examples: length, weight, time

•      Psychological measurement is used to measure constructs for which there is no agreed upon standard or instrument.

–  Are there agreed upon standards for what is considered beauty, intelligence, or aggression?

•      Psychologists develop measures to assess these and other psychological constructs.

Measurements (continued)

•      Measures must be valid and reliable.

–  Validity refers to truthfulness; a valid measure is one that measures what it claims to measure.

•   Example: Does the SAT actually measures skills necessary for success in college?

–  Reliability refers to the consistency of a measure.

•   For example, a measure is considered reliable when different observers consistently agree about an observation.

–  Note that a measure may be reliable but not valid.

•   For example, a scale that consistently underreports someone’s weight is reliable but not valid.

Hypotheses

Nonscientific

•      Untestable

–   concepts not defined clearly,

–   circular,

–   appeals to ideas outside realm of science.

Scientific

•      Testable

–   concepts are clearly defined and can be measured.

Hypotheses (continued)

•      Hypotheses are not testable if they have any of these three characteristics:

–   Constructs are not adequately defined.

•    Example: People become aggressive following exposure to media violence because the violence is “disturbing.”

–   The hypothesis is circular — the event itself is used as an explanation of the event.

•    Example: People become aggressive following exposure to media violence because they become verbally or physically abusive.

–   The hypothesis appeals to ideas or forces that are not recognized by science.

•    Example: People become aggressive following exposure to media violence because of psychic influences.

Attitude

Nonscientific

•      Uncritical, accepting

–  accept claims with insufficient evidence, ignore contradictory evidence

Scientific

•      Critical, skeptical

–  behavior and mental processes are complex,

–  human mistakes are made (even in science).

Goals of the Scientific Method

•      Researchers use the scientific method to meet four research goals:

–  description

–  prediction

–  explanation

–  application

Description

•      Researchers define, classify, catalogue, or categorize events and their relationships to describe mental processes and behavior.

–  Example: Psychologists describe symptoms of depression. One operational definition of depression comes from the list of symptoms in the Diagnostic and Statistical Manual.

Description (continued)

•      Most psychology research is nomothetic rather than idiographic.

–   Nomothetic: large sample sizes, “average” performance of a group

–   Idiographic: individual case studies

–   Nomothetic researchers appreciate there are important differences among individuals; they seek, however, to emphasize similarities among individuals.

•      Most psychology research is quantitative rather than qualitative.

–   Quantitative: statistical summaries of performance

–   Qualitative: verbal summaries of research findings

Prediction

•      When researchers identify correlations (relationships) among variables, they are able to predict mental processes and behavior.

–  Example: As level of depression increases, individuals exhibit more helplessness (failure to initiate activities and pessimism regarding the future).

Prediction (continued)

•      A variable is a dimension on which people differ, or vary.

–  Examples: childhood loss of parent (yes/no), symptoms of depression, aggressiveness, age, emotional problems, stressful life events, physical illness

Prediction (continued)

•     A correlation occurs when two measures of the same people, events, or things vary together or go together.

–  Example: The more stressful life events a person experiences (one variable), the more likely they are to experience physical illness (a second variable).

Prediction (continued)

•      When two variables are correlated, if we know people’s scores for one variable, we can statistically compute (predict) their scores for the second variable.

–   For example, if we know the extent to which someone has experienced life stress, we can compute their likelihood of experiencing physical illness (and predict stress based on illness).

–   Because test scores (SAT, GRE) are correlated with grades, we can predict students’ grades based on knowing their test scores (and predict test scores from grades).

Prediction (continued)

•      Correlation does not imply causation. We don’t know why the variables are correlated.

•      For example, there’s a correlation between the amount of hair in one’s ear and the presence of heart disease. (true)

–   Does this mean that having hair in one’s ears causes heart disease?

–   Does this mean that having heart disease causes hair to grow in the ears?

–   Or is there some other variable that accounts for the relationship between ear hair and heart disease?

Explanation

•      Researchers understand and can explain a phenomenon when they can identify its cause(s).

–  Example: Research participants exposed to unsolvable problems become more pessimistic and less willing to do new tasks (i.e., they become helpless) than participants who are asked to do solvable problems.

Explanation (continued)

•      Researchers conduct controlled experiments to identify the causes of a phenomenon.

•      Control requires that researchers manipulate factors, one at a time, to determine their effect on the event of interest — these are independent variables.

•      Researchers observe the effect of the independent variable by measuring dependent variables.

•      Remember: The word “experiment” is often used in everyday language to mean the same thing as “research,” but the word experiment refers to a very specific type of research study.

Explanation (continued)

•      Using experiments, researchers can make causal inferences — statements about the cause of an event or behavior.

•      Three conditions for making a causal inference:

1. Covariation of events: If one event causes the other, the two events must vary together (when one changes, the other must change also).

2. Time-order relationship: The presumed cause must occur before the presumed effect.

3. Elimination of plausible alternative causes: We accept a causal explanation only when other possible causes of the effect have been ruled out.

Explanation (continued)

•      Example of a causal inference based on research findings:

            Exposure to media violence causes an increase in the likelihood of aggressive and violent thoughts, emotions, and behaviors immediately after the exposure.

•      Based on this research, we know

–   Exposure to media violence and aggression vary together.

–   Aggression follows after the exposure (not before).

–   Other explanation for the relationship between exposure to media violence and aggression have been ruled out.

Explanation (continued)

•      Causal Inferences

–   Scientific control requires that the effects of independent variables are isolated.

–   A confounding occurs when two potentially effective independent variables are allowed to vary together simultaneously — we cannot determine which variable caused the effect on the dependent variable.

–   When an experiment is free of confoundings, we can make a causal inference regarding the effect of the independent variable on the dependent variable.

Generalization

•      Researchers are not interested just in the one sample of people or the one set of circumstances they tested.

•      They wish to generalize a study’s findings to different populations, settings, and conditions beyond those used in the specific study.

–   Can we generalize or apply the findings from psychology studies with college students samples to other people?

–   Can we generalize the findings of highly controlled laboratory studies to real-world settings?

–   For example, can a study that examines conditions of aggression in the lab with college students be used to understand real-life conditions of aggression?

Application

•      Psychologists apply their knowledge and research methods to improve people’s lives.

–  Example: Treatment that encourages depressed individuals to attempt tasks that can be mastered or easily attained decreases depressives’ helplessness and pessimism.

•      Basic and Applied Research

–   Applied Research: Psychologists conduct research to change people’s lives for the better.

    Applied research is often conducted in “real world” or natural settings.

–   Basic Research: Psychologists conduct research to understand behavior and mental processes — “seeking knowledge for its own sake.”

    Basic research is often carried out in laboratory settings with the goal of testing theories.

•      Both basic and applied research studies are needed.

Scientific Theory Construction and Testing

•      Theories are proposed explanations for the causes of phenomena.

–  Theories attempt to explain the who, what, when, where, how, and why of people’s behavior and mental processes.

–  A theory is a logically organized set of statements that

•   define events (concepts),

•   describe relationships among these events, and

•   explain the occurrence of these events.

Psychological Theories

–  Theories vary in their scope and complexity.

–  Successful theories

•   organize what we know about a behavior or mental process (empirical knowledge),

•   guide future research by suggesting testable hypotheses, and

•   survive rigorous testing (such as falsification).

–  Good theories are logical and internally consistent, precise, and parsimonious.

Theories are Parsimonious

•      No way to win in defining scope of study

•      If theory is too broad

–  Not easy to measure constructs

–  Not easy to put into practice

•      If theory is too narrow

–  May not explain all necessary events

Intervening Variables

–  Theories often propose intervening processes or mechanisms to explain the relationship between an independent variable and a dependent variable.

–  Intervening variables are “hidden” processes that are represented by psychological constructs.