Seminar 6.8.2019: Nowcasting methods and Finnish business cycles

Programme and links to the papers closely related to the talks:

*Introduction: Henri Nyberg (University of Turku) ”Introduction to nowcasting methods and Finnish business cycles”

Nowcasting methods

*Paolo Fornaro (ETLA): ”Nowcasting Finnish economic activity: a machine learning approach”

*Petteri Juvonen (Bank of Finland and University of Helsinki): ”Nowcasting the Finnish economy with a large Bayesian vector autoregressive model”

*Jetro Anttonen (ETLA and University of Helsinki): ”ETLAnow project – Nowcasting unemployment with alternative data sources”

Dating and predicting Finnish business cycle turning points

*Harri Pönkä (Ministry of Finance): ”Forecasting the state of the Finnish business cycle”

*Wilma Nissilä (Bank of Finland and University of Turku): ”Dating and predicting Finnish business cycle states with monthly data”

LONGITUDINAL DATA ANALYSIS (5 ECTS)

Change studies and developmental studies are common in scientific enquiry and require longitudinal data. Statistical models based on repeated measurements involve special questions that do not exist in simpler models. What is the unit of analysis, for example? How do we measure change and evaluate its significance? Do the individual profiles in a study population differ in time, and if so, is the difference significant? Can we use ordinary regression models for such analyses?

The Turku Centre of Statistics organises a 3-day introductory intensive course on Longitudinal Data Analysis on 17-18.10 and 31.10.2017. The course is designed for UTUGS PhD students from different disciplines. Basic understanding of statistics is required.

The course consists of a pre-assignment task, 12h lectures presenting basic principles of longitudinal analysis and illustrating different types of longitudinal data and designs, and 6h hands-on computer class exercises. During a week’s break, the participants will complete a home work.

Aims of the course:

The student learns
–          to evaluate the appropriateness of a longitudinal study design
–          to present research questions that can be answered with statistical models
–          to carry out longitudinal analyses and report the essential results of them

Registering: send an email to lazarus.p.unandapo[at]utu.fi with a short description of why you would like to participate the course. Max. number of participants 30 (in computer class), the lectures are free.

Lecturer: Prof. Mervi Eerola
Teachers (computer classes): Lazarus Unandapo, Jouko Katajisto, Markus Matilainen
Place: Quantum, M2 (lectures), IT 112 (computer class)