Week 7 - Non Response Flashcards
list the three categories of non respondents
- those who the data collection does not reach
- those who refuse to participate
- those who are unable to perform the task or provide info e.g. due to an impairment
how do you calculate a response rate
the number of people responding divided by the number of people sampled.
Usually represented as a percentage
how is bias formed from non response
- final sample has little relationship to the intended and original sample (those responding are essentially self selected)
- no agreed upon standard for an acceptable response rate (ranging from 25-75%)
list the groups/minorities that are usually underrepresented in surveys
- unemployed
- single people
- recent migrants
- inner city inhabitants
- low SES
- low education levels
- people who aren’t fluent in English
what are the 2 issues that must be addressed to reduce non response rate
- gaining access to selected individuals
2. enlisting their cooperation
list the 3 strategies to correct for non response
- using proxy respondents
- doing statistical adjustments
- re surveying sample of non respondents
what is a proxy respondent
A person in a clinical trial other than the subject/patient who provides responses about the subject/patient on their behalf, but not as an observer.
when is using a proxy respondent not effective
when reporting on info about subjective states e.g.
- feelings
- knowledge
- opinions
when is a proxy respondent effective
when gathering factual information
how do statistical adjustments help to correct for non response
it is possible to weight the answers given by a certain segment of sample based on % of group they represent
how does ‘resurveying a sample of non respondents’ correct for non response
with mail surveys if researcher has attained 60% they may decide another mode of contact/approach a percentage of non respondents using a call
