1. Nominal: classify

2. Ordinal: +order

3. Interval: +determine differences

4. Ratio +determine ratios

Mutually exclusives

Exhaustive

Used to classify responses into 1 of 2 groups

Usually yes or no

-Same as dichotomous but categories are more focused 

-Demographic data

Order response according to a predefined characteristic

Rank order scaling

Graphic: mark on a line

Itemized: very unlikely, unlikely...etc

Semantic differential: opposite words

Likert: sd, d, n, a, sa

1. Explicit Instructions

2. Simple active language

3. Avoid bias

4. Avoid ambiguity

5. No double-barrelled questions

6. No assumptions

7. Personal information

What are the components of an experiment

Independent variable - cause

Dependent variable - effect

3-15 people

Sample size = 15-45

Used when the presence of others will generate ideas

1. Problem statement

2. Create discussion guide

3. Determine group characteristics

1. Prefatory remarks

2. Introduction of participants

3. Set the stage

4. Main discussion

-introduction of the moderator

-explanation of moderator's role

-topic of discussion

-guidelines for discussion

-observation

-rules for reporting/informed consent

What are the components for introduction of the participants?

-name

-basic demographic info

-something relevant to the discussion

-start with questions that probe attitudes/behaviors

-nothing too personal or controversial

open ended questions

stay true to discussion themes

discussion is dynamic, not static

1. covariance = the changes in the IV must correspond with changes in the DV

2. temporal order = IV must have occurred before the DV

3. Elimination of alternative explanations

1. Testing (process of testing/retesting affects a subject's behaviors)

2. History (historical events beyond the control of the researcher have an effect)

3. Maturation (the subject naturally changes during the course of the experiment)

4. Instrumentation (data collection techniques change)

5. Selection (experimental and control groups must be comparable)

6. Mortality (subjects drop out)

To increase internal validity

minimize threats

Quasi- no control group

True- subjects are randomly assigned to control and experimental groups

One group posttest only

One group pretest/posttest

Simulated pretest/posttest

Posttest only w/ control

Pretest/posttest w/ control

Solomon 4 group design

IV -- DV

Simulated pretest/posttest

Threats minimized?

Group 1: Pretest

Group 2: IV -- DV

Testing, mortality

Posttest only w/ control

Threats minimized?

Group 1: DV

Group 2: IV -- DV

Testing, history, maturation, instrumentation

Pretest/posttest with control

Threats minimized?

Group 1: Pretest -- DV

Group 2: Pretest -- IV -- DV

testing, history, maturation, instrumentation, selection

Solomon 4 group design

Threats minimized?

Group 1: Pretest -- IV -- DV

Group 2: Pretest -- DV

Group 3: IV -- DV

Group 4: DV

Testing, interaction, history, maturation, instrumentation, selection, mortality

Group 1: pretest -- DV

Group 2: pretest -- IV₁ -- DV

Group 3: pretest -- IV₂ -- DV

Group 4: pretest -- IV₃ -- DV

Used to examine IVs that interact with each other

G1: factor AX

G2: factor AY

G3: factor BX

G4: factor BY

Each participant gets every level of the IV, so subjects can make their own comparisons

G1: IV₁ -- DV -- IV₂ -- DV -- IV₃ -- DV

G2: IV₂ -- DV -- IV₃ -- DV -- IV₁ -- DV

G3: IV₃ -- DV -- IV₁ -- DV -- IV₂ -- DV

Mean

Median 

Mode

Mean

Outliers

Median

A measure of the strength of relationship between 2 different variables

ranges between -1 and 1

_

x