class4-secA-2013rtf.rtf by shenreng9qgrg132


									                        Research Design & Analysis 1:
                             Class 4 (Section A)
• Announcements
    – lab due dates
    – new lab section
    – S.M.I.L.E.
• Review
• Asking questions (Lab assignment)
• Introduction to non-experimental research methods

                                   New Lab Section
                  We have added a new section PSYC2010 LC1
                             Thursday 3:00-4:30
                                  HSH 016

                               Starting this week
                    Please (some of you) move to this section...
The Sensory Motor Instructional Leadership Experience (SMILE) Program is in its seventeenth
  year of operation. In the past, Acadia students from every faculty on campus have
  volunteered to give instruction in physical activity to children with "special needs" in Kings
  County on a weekly basis. The instructor/child ratio is one-to-one and the experience has
  been a very positive one for all involved. Last term over 200 Acadia students provided
  instruction for 180 children every week.
                             Thursday, September 23rd, 1999 at 6:00 p.m.
                                          BAC 244

•   Overall scientific purpose - to build knowledge base.
•   Knowledge base built by research.
•   Best research - scientific method.
•   Scientific method is carried out in steps called the “Research Process.”
                              Scientific explanations
•   Empirical - based on objective observations that are verifiable by others
•   Rational - follows rules of logic and consistent with known facts (data)
•   Testable - through direct observations of from specific predictions
•   Parsimonious
•   General (ideally)
•   Tentative
•   Rigorously evaluated
                   Is it a good (scientific) question?
The “ROTI” Test
• Is it Repeatable (replicable)?
• Is it Observable (empirical)?
• Is it Testable (falsifiable)?
• Is it Important?
   – Plausible?
   – Justifiable?
   – Does it make a contribution?
                         Is this a good question?
• Are there racial differences in the size of newborn male penises?
                  Turning questions into hypotheses
• Hypotheses are falsifiable statements
   – Poperian Science
• Make best guess
   – H0 (Null) vs Hexp (Experimental hypotheses)
• State the nature of the relationship between two (or more) variables
• State expected difference between two (or more) groups
   – directional or non-directional

                       Choosing a research design
Causal vs correlational relationships

   While correlation doesn't imply causation, causation does imply

     Introduction to scientific methods: Non-experimental
• characteristics of non-experimental research
• types of non-experimental research
• what you can learn from these methods
           Characteristics of non-experimental research
• Observe but don’t manipulate
• Lack of control over extraneous variables

Extraneous variables: not directly interested in but which might influence
  the results of your study

                 Types of non-experimental research
• Observational research
   – naturalistic observation
   – participant observation
      • (e.g., Rosenham, 1973)
   – content analysis
• Archival research
• Survey research
   – questionnaires
   – interviews
                      What you can determine/learn
• Co-occurrence of variables
   – correlation
• Ability to make predictions
   – regression
   – predict criterion variable from predictor variable
• Causality still unknown

                       What you cannot determine
If there is a causal relationship between variables
• third variable problem
   – number of appliances and children
   – male pattern baldness and heart disease
• directionality problem
   – health and lice (New Hebrides)
   – watching TV violence and aggression

                         Why use these designs?
• Can be useful in early stages of a research program
• Interested in relation among variables in the “real world”
• Only interested in predicting behaviour (vs. understanding causal
• Experimental methods not possible
   – Organismic variables
   – Ethics
                      Correlations and Regressions
Can use correlation/regression with either manipulated predictor variables
 or natural variation.
Correlational statistics can be applied to any type of design (including
Correlational design occurs when we do not randomly assign participants
 to the level of either variable - i.e., levels of variables are not

                           Correlation coefficients
• Measure the direction and degree of association.
• But can’t infer causality due to
  – directionality problem
  – 3rd variable problem

            Drawing conclusions from correlational designs
Have concerns with directionality, and there are usually many potential
 confounds (uncontrolled extraneous variables - the 3rd variable problem)
Causality can not be inferred
Correlational designs can be used to:
  – discover relations
  – to solve ethical and practical problems
  – to provide greater external/ecological validity

To top