L4 Surveys and Scale Types

Surveys

Systemic process of recording consumer info via structured questionnaires

List the 4 main categories of distribution methods

1. in-person

2. phone

3. internet

4. mail

List the 4 steps when designing a survey

1. determine survey’s purpose

2. identify target participants

3. determine question design

4. determine dist. method

What are the 5 survey components

1. intro + consent

2. screening questions

3. core questions

4. demographic clarification

5. thank you/closing

Survey errors

1. sampling error

2. measurement error

3. population specification error

Sample (random) error

• Cause: population DOES NOT EQUAL sample 

  • This difference is caused by chance variation 

• Mitigation: increase sample size   

Population specification 9systematic) error

• Cause: personal mistake 

• Mitigation: careful design and implementation 

Examples: 

• Incorrect definition of population (too narrow/wide/different) 

• Incomplete or inaccurate sampling frame

• Sampling procedure is inappropriate or applied incorrectly 

Measurement (systematic) error

• surrogate error

• non-response bias

• interviewer bias

• response bias

• processing error

NOIR

nominal, ordinal, interval, ratio

Nominal scale

numbers are assigned only as uniform labels

Ordinal scale

• Intervals between ranols are meaningless (the difference between each number is the same) 

• Must have clear ordering 

Interval scale

The distances between numbers correspond to the distances between the objects 

• Opposite of ordinal as there are meaningful differences (gaps) 

• No meaningful zero

Ratio level

Numbers have a true 0 point and have equal intervals 

• Equal spacing 

• 0 means “none”/ “nothing exists”