Page 176 - Special Topic Session (STS) - Volume 1
P. 176
STS423 Ylva A.R. et al.
decision was taken to only match blocks that consisted of a single
parcel of temporary grasses. This meant that 94 percent of the total
number of blocks with temporary grasses could be matched and the
age of the temporary grass could be determined. However, when the
same method was applied on blocks with fallow (set -aside) land, only
67 percent of the blocks consisted of one parcel. Therefore, further
tests were carried out to find out whether it was possible to increase
the number of blocks that could be matched. For example, in one test,
block size was used as an additional matching variable. First, the exact
block size was used, and then a deviation of +/-10 percent of the size
was allowed to find matching block pairs among the different years.
Another test included information on buffer zones, as the
geographical positions of these parcels are constant over time. Finally,
after concluding the tests, an optimal method was established. For
fallow land, this method increased the number of blocks that could be
matched to 85 percent.
2.2 Development of a new web system for data collection
In 2015, a project was launched to upgrade and modernise the web
system that had been in use since 2005 in the surveys on crop production
and autumn sown areas (Ländell et al., 2004). One of the main aims of this
modernisation was to facilitate future additions of new modules, such as
new surveys. This opened up the possibility for existing sample surveys in
agri-environmental statistics to join the project. A group of IT experts,
methodologists, statisticians, agronomists and responsible line managers
was established at Statistics Sweden. An agile approach was used, based
on Scrum™ (Schwaber & Sutherland, 2017), with sprints, demos and
frequent meetings within the group to discuss issues such as definitions
and harmonisation, code lists, product backlog items (PBIs), and
prioritisation. The composition of the group could vary to some extent
based on the purpose of the meeting. The aim was to fully replace the old
system used in crop statistics, which was a mixed- mode data collection
system with self-administered web questionnaires and an interface for
telephone interviews, built -in tools for administration, questionnaire
validation, data checking and editing. The new system had to additionally
provide new functionality (see Results, Table 2), mainly based on
suggestions provided by interviewers who had worked on the surveys in
the preceding systems. One of these new functions was a tool for
importing data from professional farm management software.
2.2.1 Tool for data import from farm management software
During the first phase of the overall project, an initial contact had
already been taken with the two dominating companies offering farm
165 | I S I W S C 2 0 1 9