Research Theme: Methodological Advances & Innovations
The Methodological advances and innovations theme reflects one central strength of Bangor Business School where our researchers perform at the highest level in terms of methodological rigor with a passion for creating new knowledge for the benefit of fellow academics across diverse fields.
Many of the analytical advancements have been incorporated into company processes. For instance, forecasting research has impacted company processes at Siemens, Carl Kammerling International Ltd, and Roberts of Portdinorwic. Current work with Stockomendation aims to develop a suite of models to analyse stock recommendations for commercialisation.
This theme encompasses social science and economic research, utilizing quantitative methods in the following topics: performance management and optimization which includes advancing empirical methods of forecasting, and extending efficiency model operationalisations. For instance, development of the Theta method and comparing this method to other forecasting methods (Nikolopoulos). Studying the consequences of unobserved heterogeneity when employing different econometric methods in the estimation of two major value-relevance models the Price Regression Model and the Return Regression Model (Vasilakis) Innovative modelling that focuses on new methods and linking previously unrelated topics into new models.
For instance, developing a theoretical model in which sovereign credit news from multiple rating agencies interacts with market heterogeneity, or developing a new tractable method for estimating hedonic cost (Alsakka, O ap Gwilym). One further area relates to psychometrics and scale development, which includes developing new scales or new empirical measures, developing short-forms of existing scales, and the cross-cultural evaluations of established scales (Hanna, Hassan, Shiu). Eleven members of staff are actively researching and publishing in topics under this theme.
Research Theme staff: Dr Rasha Alsakka, Prof John Ashton, Prof Owain ap Gwilym, Dr Sonya Hanna, Prof Louise Hassan, Dr Azhdar Karami, Dr Noemi Mantovan, Prof Kostas Nikolopoulos, Prof Edward Shiu, Dr Chrysovalantis Vasilakis, Prof Jonathan Williams
Predictive Analytics in forLAB: Forecasting with the Theta Method and Temporal Aggregation
Professor Nikolopoulos’ forLAB is a forecasting laboratory in Bangor University focusing on Predictive Analytics. Professor Nikolopoulos’s ground-breaking Theta method is widely used across regional, national and international businesses and governments resulting in significant economic savings and efficiencies. UBER is using the Theta method worldwide for forecasting high frequency time series (down to 10ms) and the raison d'être for its use is the proven accuracy and computational speed. BOSCH is using Theta for forecasting the demand of their very popular series of Power Tools. Amazon WS is benchmarking against Theta. Recently, forLAB focused on Temporal Aggregation methods that have contributed significantly to inventory savings for Siemens too.