PSY205: Operational Definitions

1
PSY205 Operational Definitions Research Designs
2
Operational definitions

• Variables must be clearly stated in measurable
  terms

3
Operational definitions

• Smelling pleasant scents will decrease stress
• What do we mean by pleasant scents?
• What do we mean by stress?
• How can we define and measure these concepts?
• People may have different ideas about what these
  are

4
Operational definitions

• Aggression of times a child hits a doll
• Memory of words a person can recall from a
  list
• Operational definitions make abstract concepts
  measurable

5
Operational definitions

• Write an operational definition of pleasantness
  and of stress

6
Operational definitions

• Stress
• Physiological measures
• A measure of patience/frustration
• Self-rating
• Pleasantness
• Rating scale

7
Designs

• Correlational studies
• Experiments
• Quasi experiments

8
Designs

• Does watching violence on TV lead to aggression
  in children?

9
Correlational studies

• Ask parents how many hours of violent TV their
  kids watch each day
• Measure how aggressive the kids are
• See if the two are related

10
Correlational studies

• Problem cannot claim that one thing is causing
  the other
• Researcher doesnt manipulate anything
• Can only say that the two things are related

11
Experiments

• Researcher manipulates the conditions
• Assign half of the kids to watch 1 hour/day of
  violent TV, the other half to watch 2 hours/day
• How do we decide which kids watch 1 vs. 2 hours?
• Use random assignment
• Each kid has an equal chance of being in either
  group

12
Ways to use random assignment

• Toss a coin
• Use a random number table
• Random number function on calculator or computer
  is probably OK (provided you dont cheat)
• If you want equal numbers in each group
• pick the groups from a hat
• shuffle surveys before distributing them
• etc

13
The following are NOT random

• Alternating, or anything else that is systematic
• seating position
• time of arrival
• Choosing haphazardly
• Making up your own random sequence

14
Why does randomization equalize groups?

• People assigned arbitrarily to different groups
• no prior reason to expect people to be different
• especially if the groups are large

15
Quasi experiments

• Can manipulate conditions, but cant use random
  assignment
• Test two different teaching methods
• Cant use two methods in the same class
• Use one for section A, one for section B
• You are manipulating the teaching method, but not
  using random assignment