elaborated by Hansen et al. (1953) and Lance
& Hattori (2016), all errors except sampling errors are non-sampling errors
which occur when data is missing, when there are analysis errors, input
processing errors and recording errors. In addition, some other non-sampling
errors occur from the measurement instrument, such as errors of unclear
definitions, poorly conceived concepts and defective questionnaires. Unsuitable
definition of the frame is a non-sampling error and finding a frame that
perfectly fits the population is in many cases impossible (Lance & Hattori, 2016).

Furthermore, a non-responsive error happens when sampling units that were
chosen did not respond or interviewed because they were unavailable, unable or
unwilling to do so. Non-response errors can affect survey results in the way that
it decreases the sample size or the amount of data gathered, in response to
specific question results in larger standard errors. Also, a bias can be
brought forward to the point that non-respondents differ from respondents
inside a pre-specified sample. For example, citizens who to not properly follow
tax law will be more prone to refusing or be uncomfortable on filling out a survey measuring tax payment compliance (Lance &
Hattori, 2016). Ultimately, this action will noticeably bias the data towards a
more law abiding net sample than the original sample. Nonresponse bias in
surveys asking for legal information have been evidenced to be even more extreme
if the survey clearly states that the government or another organization of authority
is gathering the data (Lance & Hattori, 2016).

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now