Docstoc

Experimental Design

Document Sample
Experimental Design Powered By Docstoc
					Experimental Design



     The Research Process
 Defining a Research Question
Research & the scientific
       process

What is the scientific process?
   Rationalism
   Empiricism
    Scientific theories
Inductive theory
   Specific to general
Deductive theory
   General to specific
Functional theory
   Elements of both
Models
   “mini-theories”
Inductive approach

       Theory

      Hypothesis

       Pattern

     Observations
Deductive approach


           Theory

         Hypotheses

        Observation

Confirmation/Non-confirmation
Phases of a research study

1.   Idea-generating
2.   Problem definition
3.   Design of procedures
4.   Data collection
5.   Data analysis
6.   Interpretation
      Research Designs*
   Naturalistic observation
   Case study




                                                           Constraint level
   Correlational
   Differential
   Experimental


•Taken from Graziano
•Not all research studies fit neatly into one of these categories
Strengths of low constraint
         research

Can be used to generate hypotheses
Can be used to negate a proposition
Can be used to identify contingent
relationships
Limitations of low constraint
          research

 Cannot be used to test hypotheses
 Poor representativeness
 Poor replicability
 Observer bias
 Ex post facto fallacy
           Strengths of
correlational/differential research

  Good for situations where manipulation
  of an independent variable is not
  practical or ethical!
  Higher constraint than observations or
  case studies
          Limitations of
correlational/differential research

  Influence of confounding variables
  Correlation does not imply causation
     A causes B, B causes A, some other factor
      causes A and B
  The researcher measures but does not
  manipulate the variables
      Strengths of
  experimental designs

Causation can be determined (if
properly designed)
The researcher has considerable control
over the variables of interest
Can be designed to evaluate multiple
independent variables
      Limitations of
   experimental designs

Not ethical in many situations
Often more difficult and costly
        Developing the research
          question/hypothesis
                            Initial idea


              Initial
                                              Literature search
           observations



                          Problem statement
                                                      Operational
                                                       definitions


(Graziano, 2000)      Research hypothesis
Good characteristics of a
  problem statement

States the expected relationship
between variables
The problem should be in the form of a
question
Implies the possibility of an empirical
test of the problem
Problem statements
Observations & Case studies
   Given A what is the probability of B?
Correlational research
   Is variable A correlated to a specific change in
    variable B
Differential research
   Will group A differ from group B by variable X?
Experimental design
   Does variable A cause a specific change in variable
    B?
Operational definitions
Definition of the variables of interest
    How are they defined?
    How will they be measured?

    A good operational definition of variables
     defines the procedure so precisely that
     another researcher could replicate it
   Research hypothesis
   Develop the problem statement into a
   specific testable prediction
       States the direction
       Represents a declarative statement

e.g., Brown bullheads exposed to PAH-contaminated
sediments will develop skin tumors at a higher rate than
controls
What is an experiment?

 An inquiry in which an investigator
 chooses the levels (values) of input or
 independent variables and observes
 the values of the output or dependent
 variable(s).
         What is a statistical
        experimental design?
Determine the levels of independent variables (factors)
and the number of experimental units at each combination of
these levels according to the experimental goal.

     What is the output variable?
     Which (input) factors should we study?
     What are the levels of these factors?
     What combinations of these levels should be
     studied?
     How should we assign the studied combinations to
     experimental units?
Experimental unit: the unit we apply the factors on to get the response.
                Example:
          soft drink beverage

•What is the output variable?
  Taste of the drink; score 1 to 10 (from poor to good)

•What factors and at which levels should we study?
                                           A, B
 • Type of sweetener
 • Ratio of syrup to
    water
 • Carbonation level                        Low, High
 • Temperature
         Example:
    soft drink beverage

•What combinations of factors should be studied?

 All 2x2x2x2 combinations.

•How should we assign the studied combinations to
experimental units?

Assign equal number of units to each combination.

 (unit: the “null” beverage or say the plain water)
The Six Steps of Experimental
           Design

  Plan the experiment.
  Design the experiment.
  Perform the experiment.
  Analyze the data from the experiment.
  Confirm the results of the experiment.
  Evaluate the conclusions of the
  experiment.
Plan the Experiment
Identify the dependent or output
variable(s).
Translate output variables to
measurable quantities.
Determine the factors (input or
independent variables) that potentially
affect the output variables that are to
be studied.
Identify potential combined actions
between factors.
           Example:
Which brand of battery should we
             buy?


•What is the output variable?

  Battery life. (in hours)

•What are the input variables (factors)?
   Three available brands; Prices etc.
 Design topics vs. variable
           types

Response             Predictor (input)
(output)       Continuous       Categorical
                Response      Standard
              surfaces (RS) ANOVA designs
                Uniform     (eg. factorial
Continuous
              designs (UD) designs)
                Optimal
              designs (OD)
Categorical               UD,OD
     Prior experimental
         information

The model is known
but the parameters are    Optimal designs
not:
The shape of the
model is somewhat        Response surfaces
clear:

The model is
                          Uniform designs
completely unknown:

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:0
posted:7/12/2012
language:
pages:27