The work package will cover the following objectives:
Develop and test an ad-hoc module on the detailed usage of mobile devices. The purpose of the survey is to analyse and improve representativeness of MNO data and subsets of MNO data (not having full coverage in terms of providers). Elaborate relevant survey design aspects, i.e. sample design, estimation strategy. Advantages and drawbacks of different options will be compared.
Test the developed questionnaire in a pre-test/friendly user test, so e.g., with employees of participating partners.
Consultation with MNOs and other stakeholders on the questionnaire and the ad-hoc module.
T4.1 "Questionnaire design of the proof-of-concept of ad-hoc surveys as potential non-MNO source"
Assessing the representativeness of MNO data requires identifying potential gaps in the data and developing targeted strategies to address these gaps. To fill in these gaps and the other information needs arising from the methodologies developed in WP3, a questionnaire can be developed to collect the necessary data. The questionnaire can be administered to a representative sample of individuals within the population of interest, and the data collected can be used to adjust and weight the MNO data to more accurately reflect the true distribution of the target variables. This approach can help to ensure that the MNO data is representative and reliable and can be used with confidence for a wide range of applications, of interest of both the parties, e.g., from marketing and advertising from the MNO side, to official statistics purposes from the NSI side. Additionally, the ad-hoc survey could include a block on a specific topic that would be important to be estimated in combination with MNO data, e.g., usage of visits to green areas. Possible topics are:
provider (allowing different nationalities);
number of devices with sim cards;
kind of usage of these devices (stationary, on-the-go, travel behavior, night time usage);
kind of contract (phone card);
info of the person on the contract;
Informed consent on using mobile phone data (for a limited time) to establish a small sample of MNO data.
STAT (leader), Istat, DESTATIS, SCB
T4.2 "Friendly user test of ad-hoc survey, survey design elements, and update of the questionnaire"
The developed model questionnaire will be entered in an online survey tool which will be used for a limited scale friendly user test. Feedback on the questionnaire is the main outcome of this friendly user test, but also the results of the questions will be presented as descriptive statistics. A survey testing process will be performed throughout usual cognitive pretesting and other relevant testing activities in a friendly environment involving domain experts and data collection experts. Develop and propose possible strategies how the survey can be conducted. Including sample considerations, such as the sample size, stratification, etc. The questionnaire might be run as a standalone survey or as a module in another survey, e.g., ICT usage in households. Assess potential benefits of large-scale surveys including the proposed questionnaire module to be able to estimate regional differences in mobile phone usage.
STAT (leader), Istat, DESTATIS, SCB