Tilastopäivät 2017

We thank all participants!

Statistical Days (Tilastopäivät) 2017 was held 18.5.-19.5 at the University of Turku. It was organized jointly by the Turku University Center of Statistics, the Finnish Statistical Society and The Finnish Society of Biostatistics. The keynote speaker was professor Niels Keiding from University of Copenhagen.

MISSING DATA OR MISSING INFORMATION?

Slides are available through links in presentation titles

Thursday 18.5. (University main building: Lecture Hall I)

12:00 – 12:10	Opening words
12:10 – 13:00	Juha Karvanen (JYU) Kari Auranen (UTU)
	– Introduction to missing data – concepts and perspectives
13:00 – 14:00	Niels Keiding (U Copenhagen)
	– Web-based Enrollment and other types of Selection in Surveys and Studies: Consequences for Generalizability
14:00 – 14:30	———— Coffee and tea ————
14:30 – 15:30	Short presentations I
	– Henni Pulkkinen (LUKE): Missing information from the Bayesian perspectice – examples from fisheries stock assessment
	– Mikko Sillanpää (UO): Mapping genes from a single tail sample of the phenotype distribution by generating pseudo observations
	– Jarno Vanhatalo (HU): Spatio-temporal Gaussian process to detect outbreaks of pests: a case study from the Great Barrier Reef
15:30 – 16:30	Awards and winners’ presentations
16:30 – 16:45	———— Coffee and tea ————
16:45 – 17:45	Short presentations II
	– Juha Alho (HU): Statistical Aspects of Suomi24 Discussion Forum
	– Reijo Sund (UEF): Are we missing something with complete register data?
	– Mervi Eerola (UTU): Attrition in longitudinal studies
17:45 – 18:30	Statnet discussion
19:00 – 23:00	Dinner: Kupittaan Paviljonki

Friday 19.5. (University main building: Lecture Hall II)

9:00 – 10:00	Niels Keiding (U Copenhagen)
	– Web-based Enrollment and other types of Selection in Surveys and Studies: Consequences for Generalizability
10.00 – 10.15	———— Coffee and tea ————
10.15 – 11:45	Missing data in health examination surveys  (Chair: Jaakko Reinikainen)
	– Hanna Tolonen (THL):  Selective non-participation in health surveys and available auxiliary information for non-participation adjustment
	– Juho Kopra (JYU): Methods for estimating smoking prevalence under selective non-participation
	– Tommi Härkänen (THL): Systematic handling of missing data in complex study designs
11:45 – 12:45	———— Lunch ————
12:45 – 13:45	How should we evaluate missing information? Comments and critics by the VV club (“Last of the summer wine”)
	– Chair: Elja Arjas
14:15 –	Statnet discussion

Michael Greenacre’s ”Turkutime: The Missing Data Song”

It’s Turkutime,
And Statistics is easy,
Some data are missing,
But randomness is high!

Your data are rich,
Your imputations good-looking,
So hush! Statisticians,
Don’t you cry.