Errors in the research process involving anything but the sample size

Intentional Fieldworker = interviewer falsifies their response

Unintentional Fieldworker = interviwer makes mistakes (ie. fatigue, doesn't know how to administer questions)

Errors committed by the respondents 

Intentional = respondent knowingly provides false answers 

Unintentional = respondent is confused or distracted 

Multiple submissions (submitting many surveys; solve by asking for email)

Bogus Respondents (disguise themselves as another person; solve with prequalified people)

Population misrepresentation (some people don't fit the norm)

Nonresponse

failure on the part of a prospective respondent to take part in a survey

Refusal to Participate

Break Off During Interview (reach a certain point, then no more responses)

Refusal to Answer Certain Questions (Item Omission)

Define the following:

Data Entry

Data Coding 

Data Code Book

Data Entry = making a computer file that holds raw data

Data Coding = assigning computer code values that pertain to the possible responses for each question

Data Code Book = identifies all variable names and code numbers associated with each possible response 

Should ask what the number of valid responses is

Can't see full information for variables 

Questions were too difficult to answer, touched on sensitive issues

Respondents may have had difficulty filling out survey, didn't paid enough attnetion

Reduced sample size and increased sample error

Additional cost for resampling

Systematic biases in observed responses 

Good Sample Size and few MV:

- Drop all missing values

- Check for significant differences

Sample Size Problem

- Define maximum number of MV allowed

- Choose a replacement method (ie. average)

Deals with outliers

Identify and deal with problem cases (too many MV, insufficient variation)