PSYC 209 - 1638

Variables and Validity

27

Relationships Between Variables • Positive Relationship

– As one variable increases, so  does the other variable

Ex. Professor‐student rapport and  professor’s responsiveness  (Wilson, Ryan, & Pugh, 2010)

• Negative Relationship – As one variable increases, the 

other decreases

Ex. Self‐esteem & death anxiety  (Greenberg et al., 1990)

0

5

10

15

20

25

30

1 2 3 4 5 6

0

5

10

15

20

25

30

1 2 3 4 5 6

Relationships Between Variables

• No relationship – Changes in one variable are unrelated to changes in the 

other variable

Ex. Extroversion and emotional stability

0

6

1 2 3 4 5 6

Relationships Between Variables

• Curvilinear relationship – Changes in 1 variable are related to increases AND  decreases in another variable

– Ex. Anxiety and performance on a task (Yerkes &  Dodson, 1908)

30

Cause and Effect among Variables

Crack cocaine use during pregnancy

Birth defects, low birth weight, childhood bx

problems

r = ++

Third Variable

r = ++ r = ++

(Hurt et al., 2013)

Cause and Effect among Variables

• Nonexperimental methods – Reveal relationships between variables – Lead to predictions of behavior – Problems with causality

• Experimental methods – Direct manipulation – Control – Random assignment – Independent & dependent variables

32

Validity

• Construct – Extent to which your operational definition of  the variable reflects the construct

• Internal – Extent to which you can determine a causal  relationship between variables

• External – Extent to which your findings can be generalized

33

Scales of Measurement

The Numbers in Responses

Scales of Measurement

N (nominal) O (ordinal) I  (interval) R (ratio)

28

Nominal

• Responses of subjects are linked to categories

• Can only tell whether something is or isn’t in a  category ( = , ≠ ) – E.g., Gender:

• Claudio = male • Claudio ≠ female  • Claudio ≠ other gender identity 

How Many People Are/Have ‐‐

Ordinal • Categories which are ranked or have an order

• Can indicate relative standing (>, <) – In terms of who was the better chess player,   Claudio > Minda 

– In terms of who was the better chess player, Minda > Randy

– In terms of who was the better chess player, Claudio > Randy

Claudio = 1, Minda = 2, Randy = 3

45 > 30

30 > 0

45 > 0

Order These Pictures  In Terms of The  Most Extreme  Hair Style…

Is the Interval Between 2 & 3 The Same As 

Between 3 & 4?

2

3

4

1

Interval • Ranked categories that have equal, fixed

intervals

• Example: – Day 1 (‐20° C), Day 2 (0° C) and Day 3 (20° C)  – The difference between Day 1 and Day 2 is equal to  the difference between Day 2 and Day 3

– But, the “zero” is not a real zero  it does not  indicate complete absence

What’s in a ZERO?

Ratio

• Categories which are ordered, have an equal  interval AND a true zero point

• Zero point: complete absence of the  attribute being measured

• Examples: – Miles per hour ridden on a bike – Reaction times

Height, Weight, Hours, Minutes…

What Scales Are These? 38

Scales of Measurement for Four Players on a College Basketball Team

Name Average Points Scored

Points Scored (Ranking)

Position on Floor

Jordan 15 1st Shooting guard

Olajuwon 4 3rd Center

Calderón 14 2nd Point guard

Bird 8 4th Small forward

Nominal (=, ≠), Ordinal (>, <), Interval (equal, fixed but no true 0), Ratio (equal, fixed, a true 0)