Scientific Method Uniquely Distinguishes Science

1
Scientific Method Uniquely Distinguishes Science
KEY FEATURE EMPIRICISM Data gathered by
observation Phenomenon is observable and
measurable in some way compare temperature
with memory   1. OBJECTIVE -individual
biases, expectations or goals must not affect
measure -measures outcomes not idiosyncratic
to a given observer -data not selectively used
to fit pre-conceived notion 2.
SYSTEMATIC   -same method or instrument of
measurement applied -predetermined, unbiased
scheme of making measurements MANIPULATE/CONTROL
ANTECEDENT CONDITIONS (Cause) MEASURE/EVALUATE
CONSEQUENT CONDITIONS (Effect)
2
Explanation in Psychology
Once observations have been made or data
gathered, many explanations are may be possible.
e.g., correlation of grades with seating
pattern in class (think of a number of
explanations)

• PARSIMONY All other things being equal, it is a
  basic tenet of science that one should use the
  explanation that is
• simplest
• easiest to apply to phenomenon you want to
  explain
• clearest to understand
• usually the most precise and easiest to test

Use simplest explanation unless ruled out by
conflicting data (Clever Hans)
3
Levels Of Explanation

• LAW principle that is repeatedly tested and
  verified by many experimental findings across all
  situations intended to explain the endpoint of
  experimentation
• THEORY an interim explanation
• --pulls together several observations
• --organizes information into a unified scheme
• --should generate novel predictions
• --predictions should apply to a variety of
  conditions
• --predictions are specific theory is general
• HYPOTHESIS a specific prediction about the
  relationship between two or more variables
• what is tested directly in a scientific study

4
Building Scientific Knowledge

• FORMULATING TESTING GOOD HYPOTHESES
• TESTABLE
•   VERIFIABLE
•   FALSIFIABLE
• Empirical (relies on data and observation)
• Self Correcting
•  
• RECOGNIZING NON-SCIENTIFIC KNOWLEDGE
• FIXED BELIEF
• GRANDMOTHER KNOWLEDGE
• COMMON SENSE FALLACY
• NOMINAL FALLACY

5
Generating Testing Hypotheses

• Derive from a formal theory (prediction of new
  data) deduction reasoning (from general theory
  to specific)
• Predict from a specific observation, data, or
  experience inductive reasoning (from specific
  data to general theory)
• Correct predictions (verification, increased
  confidence
• Falsifiable theories can be proven wrong, but
  not proven right (Karl Popper)

6
Experimental Variables
Measurement

• Systematic and Objective estimation of the
  quantity or quality of an observable event
• Variable anything that can take on different
  values or a range of values along a given
  dimension.
• Physical variables scales with units,
  instruments of measurement
• What about mental states, emotions, mental
  processes?

• Operational Definition Specifying the precise
  meaning of a variable within an experiment by
  defining the variable in terms of the procedure
  or manner in which it is measured or manipulated
  
• e.g., Effect of anxiety on test performance
• Independent Variables (IVs) defined by values
  chosen by the experimenter those variables that
  are manipulated or imposed
• Dependent Variables DVs defined by the
  measures that are taken free to vary

7
EXPERIMENTAL APPROACH Manipulate Independent
Variable Measure Dependent Variable Control All
Other Relevant Variables Cause-Effect
Relationships MANIPULATE/CONTROL ANTECEDENT
CONDITIONS (Cause) MEASURE/EVALUATE CONSEQUENT
CONDITIONS (Effect)
8
(No Transcript)
9
Common Nonexperimental Approaches

• Phenomenological studies
• Naturalistic observation
• Case studies
• Field studies
• Survey research
• Correlational studies
• Ex post facto studies

10
Correlational Studies
A measure of the degree of relationship between
two variables, both free to vary and are not
usually controlled by the experimenter

• Measured by a correlation coefficient (Pearson
  r)
• r can range from 1.0 to 1.0
• Accuracy of one variable in predicting the other
• Strength of relationship from absolute value of r
• Direction of relationship from the sign ( or -)
• Positive increase together negative inverse
  changes
• Data best shown with bivariate scatterplots
• r must be mathematically calculated from
  quantitative data

11
Strong positive (direct) correlation
Strong negative (inverse) correlation
No correlationr ? 0