Literature Review Paper

1
(No Transcript)
2
Literature Review Paper

• Due Date May 29
• Use a summary narrative form
• Assignment sheet
• Outline format essential

3
Agenda for This Unit

• Experimental Design
• Causal Reasoning in Experiments
• Notation
• Factorial Designs
• Main Effects
• Interaction Effects

4
Experimental Design

• Defined a study of the effects of variables
  manipulated by a researcher in a situation in
  which all other variables are controlled,
  completed for the purpose of establishing causal
  relationships

5
Experimental Design

• Distinguishing Carefully Designed Studies from
  Experiments
• Manipulation of Variables
• Cause-effect Conclusions

6
Finding True Causes in Experiments

• The Challenge of Control
• A Case Study

7
Notation

• O -
• observation
• X -
• experimental variable
• R -
• randomization

8
Designs

• Reading Designs
• Pre-Experimental vs. True Experimental Designs

9
Randomization

• controlling for extraneous variables
• random assignment
• random selection

10
Comparing Designs
11
Internal Validity

• Defined the degree to which the researcher can
  make an unequivocal statement of experimental
  effect
• Sources
• Mnemonic device
• He said my tush is sagging extra inches.

12
External Validity

• Defined the degree to which research findings
  can be generalized to other similar circumstances
• Sources

13
Other Sources of Invalidity

• Law of the Instrument
• Experimenter Effects (Demand Characteristics)
• Ignoring Initial Differences between Control and
  Experimental Groups

14
Factorial Designs

• Defined experimental designs using more than
  one independent variable

15
One Factor Designs
X1
-

16
A Two Factor Design
X1
-

-
X2

• A 2 x 2 Design

17
A 3 x 2 Design
X1
LOW
MO
HI
-
X2

18
A Three Factor Design
X1
-

-
X2

-
X3

19
The Relationship Between Factorial and Simple
Designs
X1
-

-
X2

20
The Relationship Between Factorial and Simple
Designs
X1
-

-
R X O R O
R X O R O
X2
R X O R O
R X O R O

21
Identification of Offset Control Groups
X1
-

-
R X O R O
R X O R O
X2
R X O R O
R X O R O

22
Identification of Offset Control Groups
X1
-

R X O
-
R X O R O
X2
R X O R O
R X O R O

R O
23
Designs in Which Control Groups are Included
X1
-

R X O
-
R X O R O
X2
R X O R O
R X O R O

24
Main Effects

• Defined dependent variable effects from
  independent variables separately

25
A Main Effect Example
Amount of Attitude Change Advocated
X1
-

30
-
20
Source Character
X2
40
50

26
A Main Effect Example
Amount of Attitude Change Advocated
X1
-

30
-
20
Source Character
X2
40
50

30
40
27
A Main Effect Example
Amount of Attitude Change Advocated
X1
-

30
25
-
20
Source Character
X2
40
50
45

30
40
28
Diagrams of Main Effects

• 20
• 15
• 10
• 5
• Men Women
• Variable 1

D.V.touching
29
Diagrams of Main Effects

• 20
• 15
• 10
• 5
• Men Women
• Variable 1

D.V.touching
30
Diagrams of Main Effects

• 50
• 40
• 30
• 20
• Low High
• Variable 1

D.V.AttitudeChange
Amount of Attitude Change Advocated
31
Diagrams of Main Effects
Variable 2 (Low)

• 50
• 40
• 30
• 20
• Low High
• Variable 1

Source Character
D.V.AttitudeChange
Amount of Attitude Change Advocated
32
Diagrams of Main Effects
Variable 2 (High) Variable 2 (Low)

• 50
• 40
• 30
• 20
• Low High
• Variable 1

Source Character
D.V.AttitudeChange
Amount of Attitude Change Advocated
33
Effects
X1
-

-
21
11
X2
9
19

34
Effects
X1
-

-
21
16
11
X2
9
19
14

35
Effects
X1
-

-
21
16
11
X2
9
19
14

10
20
36
Diagrams of Main Effects

• 20
• 15
• 10
• 5
• Low High
• Variable 1

37
Diagrams of Main Effects
Variable 2 (Low)

• 20
• 15
• 10
• 5
• Low High
• Variable 1

38
Diagrams of Main Effects
Variable 2 (Low) Variable 2 (High)

• 20
• 15
• 10
• 5
• Low High
• Variable 1

39
Interaction Effects

• Defined dependent variable effects from
  independent variables taken together
• Forms Ordinal
• (in the same direction as the
  main effects of variables involved)
• Disordinal
• (not in the same
  direction as the main effects of the
  variables involved)

40
An Interaction Effect Example
X1
-

-
20
20
X2
20
50

41
An Interaction Effect Example
X1
-

-
20
20
20
X2
20
50
35

42
An Interaction Effect Example
X1
-

-
20
20
20
X2
20
50
35

20
35
43
Diagram of the Interaction Effect

• 50
• 40
• 30
• 20
• Low High
• Variable 1

44
Diagram of the Interaction Effect

• 50
• 40
• 30
• 20
• Low High
• Variable 1

Variable 2 (Low)
45
Diagram of the Interaction Effect

• 50
• 40
• 30
• 20
• Low High
• Variable 1

Variable 2 (High) Variable 2 (Low)
46
Another Interaction Effect Example
Sex of Clinician Male
Female
X1
-

Type of Stuttering Clonic Blocking
-
40
20
X2
40
20

47
Another Interaction Effect Example
Sex of Clinician Male
Female
X1
-

Type of Stuttering Clonic Blocking
-
40
20
X2
40
20

30
30
48
Another Interaction Effect Example
Sex of Clinician Male
Female
X1
-

Type of Stuttering Clonic Blocking
-
40
30
20
X2
40
30
20

30
30
49
Diagram of the Interaction Effect

• 40
• 30
• 20
• 10
• Low High
• Variable 1

Male Female Sex of Clinician
50
Diagram of the Interaction Effect
Type of Stuttering Clonic

Variable 2 (Low)

• 40
• 30
• 20
• 10
• Low High
• Variable 1

Male Female Sex of Clinician
51
Diagram of the Interaction Effect
Type of Stuttering Clonic
Blocking
Variable 2 (Low) Variable 2 (High)

• 40
• 30
• 20
• 10
• Low High
• Variable 1

Male Female Sex of Clinician
52
Interpreting Ordinal Interactions

• acceptable to look at the independent variables
  separately
• permissible to interpret main effects for
  independent variables involved in the interaction

53
Interpreting Disordinal Interactions

• must look at both independent variables together
• not permissible to interpret main effects for
  independent variables involved in the interaction

54
OK to Interpret Main Effects

• 50
• 40
• 30
• 20
• Low High
• Variable 1

Variable 2 (High) Variable 2 (Low)
55
Not OK to Interpret Main Effects
Variable 2 (Low) Variable 2 (High)

• 40
• 30
• 20
• 10
• Low High
• Variable 1

56
Effects Example 1

• 20
• 15
• 10
• 5
• Low High
• Variable 2

57
Effects Example 1

• 20
• 15
• 10
• 5
• Low High
• Variable 2

Variable 1 (Low)
58
Effects Example 1

• 20
• 15
• 10
• 5
• Low High
• Variable 2

Variable 1(High) Variable 1 (Low)
59
Effects Example 2

• 20
• 15
• 10
• 5
• Low High
• Variable 2

60
Effects Example 2

• 20
• 15
• 10
• 5
• Low High
• Variable 2

Variable 1 (Low)
61
Effects Example 2

• 20
• 15
• 10
• 5
• Low High
• Variable 2

Variable 1 (High) Variable 1 (Low)