Ferðavenjur Íslendinga - 01.02.1998, Blaðsíða 30

28
Ferðavenjur íslendinga 1996
the non-contact rate is higher for men than women, usually
because of greater absence from home and more difficulties
in reaching the person concerned. Young people also have a
higher non-contact rate than older people. In order to counter
non-response bias on this account the results have been
weighted by sex and age groups.
Another type of non-response is due to incomplete an-
swers to individual questions. Even though a person may be
willing to participate in a survey, he or she may not be ready
to answer some of the questions, either because they are too
personal or people do not remember or know the right
answer. Non-response was not uncommon, for instance,
when people were asked about their income, particularly in
the first stage of the survey, as will be described more closely
below. Instead of disregarding such non-response, it has
been corrected for by means of methods described below.
Other errors. Other errors fall roughly into three categories:
Interviewer errors, processing errors and design errors.
Interviewer errors. Sometimes interviewers record wrong
answers, skip some questions, mix up the order of the
questions or rephrase them in such a manner that their
meaning is altered. In the tourism pattern survey these errors
were dealt with by hiring interviewers with experience from
other Statistics Iceland surveys, by interviewer training and
by carefully explaining the questionnaire to them before the
survey began.
Processing errors. There is always a possibility of making
an error when classifying individual “open” questions that
call for a classification after the interview is over. This
applies in particular to questions on level of education. Such
errors may be the result of insufficient information in the raw
data, unclear instructions regarding the classification codes
and coding mistakes. Every attempt has been made to keep
these processing errors at a minimum by careful editing of the
data and systematic computer search for any inconsistencies.
Design errors. An inadequate structure and design of a
survey may lead to results that are inconsistent with the real
situation. The wording of questions may cause a misunder-
standing and variations in the order of questions may lead to
diverging answers. In the first of the three stages of the
survey the mistake was made of posing the question on
income to all participants without regard to their household
status. This caused considerable non-response and uncer-
tainty in the answers, especially in the case of sampled
teenagers who were asked about the total income of their
household, i.e. their own income and that of their parents.
This mistake was corrected in the two subsequent stages of
the survey by asking only married or cohabiting persons
about the combined income of the respondent in question and
his or her spouse or partner, while other participants were
only questioned about their own income. In order to make the
first stage of the survey comparable with the other two, the
answers to the question on income were discarded in the
processing if a household was composed of more than one
grown-up person and the sampled person was neither mar-
ried nor cohabiting.
8. Concepts
Tourists. A tourist is a person travelling to a place other than
his or her usual environment who stays overnight at the place
visited for at least one night.
Tourism trip. A trip refers to the travel of a person to a place
other than his or her usual environment where he or she stays
for at least one night. The usual environment of a person
consists of the vicinity of his or her home and place of work
or study. Thus a stay by crew members aboard a ship does not
count as a tourism trip nor does a stay in hospital or prison.
Regular trips to private summer houses, however, are in-
cluded in tourism even though some people may consider
their summer house as their second home. The survey ques-
tionnaire included questions on all forms of trips taken by
people aged 11-74 but excluded children under the age of 11
years taking trips without the company of the parent partici-
pating in the survey.
Month. The month of travel is the month of departure even
though the greatest part of the trip took place the following
month.
Type of accommodation. This is classified by the type of
accommodation facility, e.g. a hotel, guesthouse, camping
site or stay with relatives and friends. If a person has used
more than one type of accommodation in the same trip, the
type used for the greatest number of ovemight stays is
selected.
Means of transport. Means of transport refers to that used
for the longest stretch of a trip.
Destination. Destination is the main destination of a trip or
the place where the tourist concemed stays for the greatest
number of nights.
Organization of trip. Respondents were asked whether they
had been assisted by a travel agency in planning a package
tour or in some other aspect of planning their trips, for
instance hiring a car or procuring accommodation.
Tourism expenditure. This refers to the total expenditure in
some way associated with the trip. In cases where more than
one person travelled together, the respondent was questioned
about the total expenditure and the number of persons incur-
ring those expenses. Tourism expenditure, therefore, relates
to one person only.
Income. Married or cohabiting respondents were asked
about the combined incomes of themselves and their spouse
or partner in 1995. Other respondents were asked about their
own income in 1995.
Weights and calculations. Totals and percentages are esti-
mated by applying weights to each answer with reference to
age and sex. Weights are calculated with the following
proportion: