Page 289 - Special Topic Session (STS) - Volume 2

STS493 Sofie d.B. et al.
authorities independently of persons’ consent, but subject to privacy
constraints and legislation. IoT sensor data do not necessarily include
identifying information about the persons to which data correspond. As a
consequence, linkage of IoT sensor data to individual respondents may be
hard or even infeasible without additional information such as time stamps
and/or location coordinates. Nonetheless, hybrid forms of data collection may
arise where respondents are asked for additional information that enables
linkage.
Nowadays, a lot of data within businesses are already available in
electronic format. A first example regarding System-to-System (S2S) data
collection for statistical purposes involves sensor data. Increasingly, electronic
sensors are used to run a business, e.g. by agricultural businesses (like dairy
farms using milking robots). A second example is financial data in a fully
integrated and digitalised business information chain which makes S2S data
communication for financial, tax and statistical reports possible. Drivers for
switching from questionnaires to S2S are: working towards smart business
statistics, (i.e. timely and new statistical output integrated in business
processes), the reduction of response burden, and monitoring and
benchmarking businesses to their counterparts.

4. Examples of both types of sensor data
We consider two examples: two surveys mandatory in the European
Statistical System (ESS): The Household Budget Survey (HBS) and the ICT
survey, and sensor data on potatoes crops as a potential replacement for
mandatory ESS survey on crop yields at an innovative farm in the Netherlands.

4.1 Household Budget Survey and ICT survey
The HBS records all household expenditures and purchases during a week
and large household expenditures and purchases during a longer period up
to a month. The HBS is very burdensome due to the duration of the diary
keeping. Detailed expenditures and purchases are non-central to respondents.
The HBS does not involve complex latent concepts that relate to many survey
questions. The HBS deals with all kinds of purchases and expenditures, both
small and large, and both frequent and infrequent. Some of these purchases
are done on site, such as shops, restaurants and cinemas. Time-location sensor
data may be employed to assist respondents in memorizing or recalling
locations where products or services have been purchased. Some purchases
are done online and part of those may be done through a mobile device. The
use of certain online shopping apps may be tracked to again assist the
respondent. In all of these cases, direct access to the type, amount and cost of
products and services will, generally, not be possible to privacy restrictions on
the apps. Another option is to use the camera to scan shopping receipts. This

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