The Department of Methodology and Statistics is part of the Faculty of Social and Behavioural Sciences of Utrecht University. It is a broad faculty, with programmes in psychology, pedagogical sciences, educational sciences, sociology and cultural anthropology. The Department of Methodology and Statistics offers methodology and statistics’ teaching for all programmes of the faculty. Moreover, the department offers a research Master dedicated to methodology and statistics in the behavioural and social sciences.

The department is responsible for a large research programme. Many researchers are engaged in programmes funded by external parties as the research foundation NWO and others such as ministries, the central testing agency CITO, and Statistics Netherlands. The research programme of the Department of Methods and Statistics has been judged as ‘very good’ or ‘excellent’ on all aspects in the most recent QANU-review, while the department’s Research Master programme has been evaluated as one of the best social science research Masters in the Netherlands (‘keuzegids hoger onderwijs 2011’). The research project advertised here is co-funded by Statistics Netherlands.

As the PhD is joint between Utrecht University and Statistics Netherlands under the WIN (Waarneem-Innovatie Netwerk) programme, also an appointment at Statistics Netherlands will be created. We expect the successful candidate to be willing to work and travel at Statistics Netherlands in The Hague for at least one day per week. Furthermore, secure remote access facilities will be provided to access microdata located at Statistics Netherlands.

This PhD project is focused on the general question how a combination of mobile device sensor data and survey questions can produce accurate official statistics on time use and travel behavior. Undertanding people’s movement is interesting to social scientists, geographers and policy makers. Mobile phones can be used to trace where people go; the data resulting from this are however susceptible to errors. This project has the goal to improve the quality of geospatial and time use data for official statistics. As part of the project, experimental microdata will be evaluated that will be collected in 2018 at Statistics Netherlands. However, the PhD candidate is expected to also actively contribute to future data collection efforts in official statistics using data collected from smartphones. Four subprojects are anticipated:

 

The first subproject will focus on selectivity in missing data in travel sensor data. Unit- and item-nonresponse is a pervasive problem in surveys and Big Data, and especially in data collected with mobile phones. Respondents may not want to participate, run out of battery, not bring their phone, turn tracking off, or there may be technical reasons why data for some trips are completely missing. Moreover, for some trips we may have partial data. For example, we have data obtained from GPS sensors, but are missing survey data on the purpose and details of the trip. The goal of this project is to model and correct for nonresponse in travel behavior data. What makes this project unique is that we have register data available from Statistics Netherlands about all sampled persons, that can be used to model all stages of nonresponse.

The second subproject will focus on measurement error and the relation to choices in the app user interface. The sensor data recorded on the smartphone record the respondent’s position with a precision of about 15 metres under normal situations. However, on some locations precision may be lower or connection to sensors may be lost temporarily (e.g. in subways). Subsequently, statistics about distance travelled, modes of transportation and purpose of travel are affected and contain errors. Respondents have the opportunity to partially adjust such errors in the app user interface. The ability of the respondent to do so, obviously, depends on the type and size of the error and the strategy with which their travel is presented to them. The goal of this project is to explore measurement error and develop methods that reduce its impact. Existing algorithms to deal with measurement error need to be tested and adapted.

The third and fourth subproject may shift attention to measurement of time use or may elaborate the net effect of nonresponse and measurement error in time-location measurements. This choice will depend on the findings in the first two papers, the implementation of apps by Statistics Netherlands and experiments with time use sensor measurements. During the PhD, the scope and research questions will be elaborated.

When shifting attention to time use, travel data is supplemented by data about activities at fixed locations, e.g. at home. Travel data do not provide any input to activities inside a house or building. A combination of bluetooth sensors and beacons, NFC tags and motion sensors may be informative and support derivation of suggestions for time use on site. While the PhD will not develop the IT itself, the analysis of the sensor data and estimation of time use statistics are in scope. Again nonresponse and measurement error may be evaluated and adjusted for the collection of a mix of travel and time use data.

When the focus remains on travel, then nonresponse and measurement error may be evaluated jointly and also be adjusted for jointly. When there is a relation between nonresponse  and measurement error, modeling in separate steps may lead to errors. For example, respondents with more missing data may be more likely to live in urban areas, where measurement error is higher, and because there are more transportation possibilities will also suffer from more classification error. In order to account for these relations in the data, nonresponse and measurement errors need to be corrected for simultaneously.

Regardless of the choice of the third and fourth subprojects, implementation will be a crucial aspect of the research towards the end of the PhD. Respondent consent to share sensor data, respondent willingness to maintain participation throughout the fieldwork period, and user-friendly navigation are key. The research should, therefore, recommend what population subgroups are at risk and what are parts of the data collection that need most care.

We are looking for a candidate, who has:

• a Master degree in Big Data, Computer Sciences, Statistics or Quantitative Social Science;
• a proven interest in Big Data, surveys, and/or official statistics;
• a further interest in Big Data, and geospatial data measured in time is also advantageous;
• a passion for research;
• an ability to read and understand Dutch documentation;
• a proven ability to write and present about their work in English;
• works effective and efficient, is able to work within a team, and to think conceptually.

We offer a temporary fulltime appointment for one year. After a positive evaluation the appointment will be extended to a total of four years maximum. Your gross salary starts with €2,222 per month in the first year and increases to €2,840 per month in the fourth year of employment. In addition, we offer attractive and flexible secondary terms of employment such as an annual holiday pay of 8%, an end-of-year bonus of 8.3% per year and partially paid parental leave. Furthermore, for foreign employees Utrecht University can provide 30 % of the salary including the reimbursement or 30/70th of the salary excluding the reimbursement as a free reimbursement for the extraterritorial reimbursement. This means you will earn a higher net salary every month. Lastly, we also provide an attractive working environment and a vibrant research climate.

A better future for everyone. This ambition motivates our scientists in executing their leading research and inspiring teaching. At Utrecht University, the various disciplines collaborate intensively towards major societal themes. Our focus is on Dynamics of Youth, Institutions for Open Societies, Life Sciences and Sustainability.

The Faculty of Social and Behavioural Sciences is one of the leading faculties in Europe providing research and academic teaching in interdisciplinary social science, cultural anthropology, educational sciences, pedagogical sciences, psychology, and sociology. Research and teaching activities are concentrated in five areas: Behaviour in Social Contexts; Child and Adolescent Studies; Cognitive and Clinical Psychology; Education and Learning; and Methodology and Statistics.

More than 5.600 students are enrolled in a broad range of undergraduate and graduate programmes. The Faculty of Social and Behavioural Sciences has some 850 faculty and staff members, all providing their individual contribution to the training and education of young talent and to the research into and finding solutions for scientific and societal issues.