Issues
With Survey Length Survey
Length: Is it a factor?
In survey development, there is a delicate balance between the amount
of information needed to obtain accurate results and the limits of respondent
motivation to participate. As the length of the survey increases, the
response rate (percent of people who complete the survey) usually decreases.
The more questions asked, the higher the benefit for you (in additional
information obtained), but the higher the cost to the respondents (primarily
in time) to complete the survey.
As such, survey length is often a difficult decision when developing surveys.
The more often that a survey is administered, the fewer the questions
respondents will tolerate.
While survey length decisions are important with all survey types, web-based
surveys are more susceptible to potential threats to validity as a result
of survey length. Attention spans can be astonishingly short in the online
world and people often quit a survey before it is complete. Statistics
have show that early termination rates of online surveys can approach
90% in some instances1.
Before Adaptiqs, surveys could contain only the most vital information
about a topic, and all other information had to be removed so that respondent
fatigue would be kept to a minimum. Surveys that were suspected of being
too long were reduced, even if important questions had to be removed in
the process.
In addition, if someone stopped responding to a survey before it was complete,
there was not a way to obtain information about how they would respond.
With Adaptiqs, surveys are not static questionnaires, but rather intelligent
instruments that adapt to the responses of each individual. As such, overall
survey length can be drastically reduced. Each new piece of information
obtained is combined with previous information and used to determine what
to ask next. By doing so, fewer questions are needed to reach a valid
conclusion.
The foundation of the Adaptiqs technology is an artificial intelligence
engine that uses probability based relationships among survey questions
and the underlying survey constructs. Our technology uses previous data
to administer only the most informative items in the survey. If respondents
decide to quit the survey early, some information can be obtained about
how they might have responded to the unanswered questions had they completed
the entire survey. When this occurs, our technology will provide its best
estimate on their most likely response, along with an estimate of how
confident it is in its prediction.
Consider the situation where you would love to get information from your
customers on 150 items, but you know they will only realistically respond
to 40 or 50 before they stop responding. In the past, you were forced to
reduce the number of questions down to that length regardless of the importance
of the information. With Adaptiqs, it is possible to choose the 40 most
informative questions for each respondent, while providing valuable information
about all items in the survey.
For more information about how Adaptiqs can improve your survey process,
fill out our customer request
form.
