Understanding Survey Formats and Experimental Designs

Survey Formats

1) open ended

2) forced choice format

3) likert scale

4) semantic differential format

Open Ended Question
Advantage: detailed, rich info
Disadvantage: lot of diverse answers that are hard to measure

Forced Choice Format
Choose the best out of choices

Likert Scale
-disagree- agree

Semantic Differential Format
Numeric scale

Leading Question
Avoid words used in question that could lead someone to answer a specific way

Double Barreled Questions
Asking 2 questions in one

Negatively Worded Question
Confusing question

Question Order
Early statements could impact answers later in the survey

Experimental Design

-Select sample from population
-Assign participants to groups
-Control possible extraneous variables (ER)
-Gather/collect data

Principle of Parsimony (Occam's Razor)
Use simplest explanation of results

The Two-Group Design
At least 2 groups, can be more but not too many

Levels of Two-Group Design
Treatment conditions/groups in an IV

Experimental Group in Two-Group Design
Gets the manipulation

Control Group in Two-Group Design
Does not get manipulation

2 Types of Control Groups
No-treatment control group and placebo control group

No-Treatment Control Group
Participants do not receive manipulation

Placebo Control Group
Participants think they are getting the manipulation (medicine)

Within-Subjects Design
The same participant tests all conditions corresponding to variable

Between-Subjects Design
Different participants are assigned to different conditions corresponding to a variable

Factorial Design
More than 1 level of the IV - 2x2 factorial design
Simples 2/IV X 2/IV
4 groups that will be compared