DOING SURVEY RESEARCH ON THE INTERNET:

YES, TIMING DOES MATTER

 

ABSTRACT

 

The Internet provides a medium for conducting survey research with tremendous potential, but at least one downside.  The instantaneous communication capability of the Internet, specifically the use of e-mail, may become overwhelming for users, thereby, creating an unwillingness to complete surveys.

 

We propose that timing is an important factor affecting response rates to Internet surveys.  Rather than sending surveys throughout the week at various times, we believe there is a “best time” to send Internet surveys that will increase response rate.

 

We empirically tested our hypothesis using a randomized field experiment with a sample of 4,994 e-mail addresses.  Response rate for the exploratory sample, randomly assigned into 14 groups (one group for each morning and afternoon of every day of the week), was 2.66%.  The “best time” determined from the exploratory sample was determined to be Wednesday morning.  The entire applied sample was sent at this time, resulting in a response rate of 4.10%.  This 54% increase was significant (p=.0118).  These data provide empirical evidence to support our hypothesis that timing does matter.