Page 181 - Special Topic Session (STS) - Volume 1

STS423 Ylva A.R. et al.
3.2.2 Implementation of a data import tool
The two options that were discussed concerning the transfer of
data from the farm management software provider to Statistics
Sweden were quite different in terms of the concept of data collection.
In the first option, which was the selected approach, only farmers who
are included in the samples of the surveys are encouraged to use the
tool, even if it is visible in the farm management software to all users.
When the data has been actively exported from the farm management
software and imported by the respondent into the web system for
agricultural statistics, the relevant cells in the questionnaire autofill
with those data. Once the data is in the questionnaire, it has the same
properties as data entered manually by the farmer or interviewer. It
can be verified, validated, changed, deleted, or saved. In this way, the
imported data will be checked for quality in the same way as any other
data entered into the questionnaire. The second option could be
described as a big data concept, since data could be potentially
collected from all farms that have an active licence for the software,
regardless of whether or not they are included in the sample, or even
in the sample frame of the surveys. However, using option 2 as the
only data collection method would lead to a bias in Swedish
agricultural statistics, since the software users are not representative
of the whole target population (see below).
The tool for data import was implemented in the Use of fertilisers
and animal manure 2016 survey, and the following year also in the
Use of pesticides 2017 survey. The information that could be imported
was: field , crop, type of fertiliser/pesticide, date of application and
application rate. The tool made it easier for respondents to fill in the
web questionnaires, especially among those who had farms with
many crops and types of fertilisers/pesticides, since a large part of the
questionnaire could be autofilled. Unexpectedly, use of the tool was
limited – about 5 percent of the total web response in both 2016 and
2017. The actual export function was not a new feature for the
software users, as the farm management software already included a
tool for data export, for example for the yearly applications to the
Swedish Board of Agriculture for area-based agricultural subsidies via
IACS. In Sweden, farms that use professional farm management
software account for a major part of the agricultural area, but
represent a smaller part of the total number of farms, as mainly larger
farms invest in such tools. Hence, only a limited part of the sample is
able to use the tool. Statistics Sweden was aware of this situation
before the function was developed and implemented in the web
system. Despite this, Statistics Sweden had hoped for more extensive

170 | I S I W S C 2 0 1 9