Quasi

1
Quasi Non-Experimental Designs
Quasi-Experimental designs Not quite true
experiments because the different
groups/conditions are not created by
__________________. Groups or conditions are
defined by _____________ variable or a ______
variable. Non-Experimental designs do not
allow you to rule out many threats to
_______________.
2
Quasi Non-Experimental Designs

• Those using ____ group, compare behavior across
  _____
• Time-series designs (quasi-experimental)
• One group pretest-posttest (non-experimental)
• Longitudinal design (non-experimental
• Those using multiple, ____________ groups
• Pretest-posttest nonequivalent control group
  design (quasi-experimental)
• Posttest-only control group design (Static-group
  design) (non-experimental)
• Differential (Causal-comparative) design
  (non-experimental)
• Cross-sectional design (non-experimental)

3
Summary of Quasi Non-experimental designs
4
Time-Series Design

• Definition A quasi experimental design in which
  behavior in one group of participants is measured
  across time _______________ an IV is implemented.
• ___________________ control for multiple threats
  to internal validity
• Allows you to evaluate _________ across time
• Called ___________________ when the IV is not
  created by the experimenter
• Called an _____________________ design if the
  treatment is repeatedly presented and withdrawn
  (provides more evidence that the effect not due
  to coincidence)

5
____________ Design
6
Time-Series Design
Example Effects of anti-smoking campaign on
smoking frequency
No ______ group hard to tell if campaign was
effective
Time Series Effect was just part of periodic
____________
Time Series Effect was just part of downward
________
Time Series Effect occurs only after
____________
7
One-Group Pretest-Posttest Design
1 naturally occurring group Pretest and postest
8
One Group Pretest-Posttest Design

• Observation made in ____ group before and after
  treatment
• No attempt is made to control for many threats to
  ________________

9
Developmental Research Designs study
age-related changes in behavior

• Longitudinal design
• Measuring a variable in individuals over an
  ___________________
• Like a time-series design with no _______________
• Can determine how an individual _________
• No cohort effects
• Very _____________, expensive
• Problems with _____, ________

10
Pretest-Posttest Non-Equivalent Control-Group
Design
11
Pretest-Posttest Nonequivalent Control-Group
Design

• Definition A quasi-experimental design in which
  behavior in two _______________groups is measured
  pre and post-IV
• No random __________ and ___________
• Because you measure behavior before treatment you
  can evaluate group equivalence - it reduces the
  threat of _______________
• e.g., Effects of Flexible vs. fixed work hours on
  productivity in two factories
• e.g., Effects of Home-based vs. School-based
  treatment on problem behavior

12
Posttest Only (Static) Group Design
13
Posttest Only (Static) Group Design

• No random selection and assignment
• 2 __________________, e.g., children in 2 clinics
  
• Groups should be ______
• Susceptible to internal validity threats, e.g.,
  assignment bias (selection threat) - group
  assignment is _______
• Example effects of peer tutoring in two
  classrooms

14
Differential (Casual-comparative) Research

• Definition Research design in which behavior is
  measured in 2 groups that differ on 1 primary
  variable.
• The IV (not true IV) is the ________ or
  ______variable
• Classification variable is not manipulated
  because it is _______ or ___________
• Normal adults vs. psychopaths
• history of abuse vs. no history
• Males vs. females
• heroine abusers vs. polydrug abusers
• Does Group A differ from Group B on measures of
  the DV?

15
Differential (Casual-comparative) Research
16
Differential (Casual-comparative) Research

• Does ____ demonstrate a causal relationship
• There is no guarantee that the two groups will be
  ______________ in all other ways.
• There may be ________________ explanations for
  the observed differences.
• e.g., kids who play sports vs. those who dont
  compare measures of academic success.

17
Conducting differential (causal-comparative
research

• Select groups that vary on __ quantifiable
  variable
• Requires ________ definitions
• Matching participants equate groups on
  ______________factors
• Select _____________ groups more likely to see
  difference in variable of interest

18
 
19
 
20
Developmental Research Designs study
age-related changes in behavior
Age group 1 e.g., 2-3
Age group 2 e.g., 4-5
Age group 3 e.g., 6-7

• Cross-sectional research design
• Measure a variable in different ___________
• Like a ____________ group design
• Cannot determine how any ___________ changes
• Other factors may influence groups other than age
  (______/_____________ effects)