Doctorate in Business and Socioeconomic Sciences
Dr. Ivo Ponocny, Full Professor.
Full Professor, Dean of the PhD Program, Department of Applied Statistics and Economics
Main areas of expertise: Applied statistics, official statistics, social indicators, evaluation of living conditions, psychometrics, international comparison studies (e.g. PISA)
The assessment of living conditions, well-being and the quality of life and its connection to social sustainability
The system of social indicators as used in the official assessment of the progress of societies by EU or OECD is subject to fundamental changes. Following the recommendations of the so-called Stiglitz report, traditional GDP-focusing reporting will be enriched by additional economic and social indicators, among them ratings of the quality of life given by individuals for themselves. There is much evidence in the scientific literature now that this subjective well-being is poorly predictable by so-called objective conditions, ignoring aspects such as the quality of social relationships, stress at the working place or the availability of time resources. Therefore, subjective indicators are involved in official statistics in the meantime (subjective well-being, life satisfaction), e.g. in the Eurobarometer where a quite simple question has to be responded to. Current knowledge states that the reliability/validity of those indicators is sufficient to apply those indicators, but still full of problems, some of which so fundamental that the validity of the approach is still subject to controversial discussions. The research activities planned shall improve the quality of satisfaction or well-being indicators and help to establish them in societal monitoring.
Currently, the main activities follow up the MODUL Study of Living Conditions which was funded by the Anniversary Fund of Österreichische Nationalbank (Budget: about € 110000) and combined 500 qualitative interviews with 1460 questionnaires and 341 one-week diaries. Recent research relates to several influencing factors on subjective well-being, such as traveling, the geographical surroundings or parental divorce, and to methodological questions such as validity and the benefit of person-centered approaches. Our main achievement by now was to show how little burdening life circumstances manifest themselves in standard ratings about happiness and life satisfaction.
In addition, several reports have been written for Eurostat, in cooperation with Statistics Austria, about social indicators including subjective well-being ratings, and also for the Austrian “Jugendbericht” about well-being of Austrian young people. A further survey will be launched in the framework of the Austrian microcensus; it will help to analyze the role of obligations regarding subjective well-being ratings, and to determine the most important happiness and unhappiness drivers of the Austrian population. The latter will also be a dominant focus in the near future.
International educational studies, especially regarding evaluation studies of national educational systems, such as PISA, PIRLS or PIAAC
In the aftermath of my psychometric developments and my former employment at Statistics Austria, funded research was carried out to be published in the National PISA, PIRLS and PIAAC reports. The first project was a re-assessment of the drastic decrease of Austrian PISA performance from 2000 to 2003, which was found to be partly an artifact (in cooperation with University of Vienna and HIS). Also later work primarily focused on methodological issues such as dimensionality of the scales, the appropriate of the statistical procedures applied and how to describe changes between time points and which statistical measures should be used for that. Contributions were made to the National Reports 2006 and 2009, but also to PIRLS 2006.
As a consequence, I have been member of national as well as international advisory boards since 2005.
A more recent results relates to PIAAC 2009, where it was found that the effect of having children on several performance indicators outweighs the well-known gender differences.
Psychometrics and non-parametric statistics
Being assistant professor at the Institute for Psychology at University of Vienna, I was mainly concerned with psychological test theory, in particular Item Response Theory. Several computer programs were written for applying linear extensions of Rasch-type models and exact goodness-of-fit tests which I had constructed for the dichotomous Rasch model. Another focus lay on the measurement of change with latent models (extended Rasch models, mixture distributions).
For some PISA articles, I demonstrated the benefit of non-parametric hypothesis testing (permutation tests).
A current project relates to optimum tests in cases where standard assumptions are not met. For example, t-tests or Mann-Whitney-tests are outperformed then by specifically designed tests. Though statisticians have dealt with such problems for decades now, there is no unified framework for this, and no user-friendly, freely accessible software, in particular not in R.
Selected PhD ThesesSupervised
- Karin Glaser (second assessor, Univ. Wien). Why terrorism? An Empirical Investigation of Terrorism Conditions and Consequences.
- Richard Mühlmann (second advisor, WU Vienna), Workplace effects of informal eldercare. An empirical approach.
- Klemens Weigl (second advisor, JKU Linz), Interim Analysis. A resampling approach.