Peter Lugtig - now also on Bluesky

Survey methodologist, smartphone data collection.

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@lakens You may be interested in the change over time in unemployment rates (previous quarter, last year etc.), but ultimately you are estimating unemployment rates at time t. E.g. with https://t.co/MmEdx9ubKj you can do all kinds of tests, but I see this as an estimation problem

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@lakens There is no test here. The idea is to estimate the time-series (or a CI of) unemployment rates. Or, e.g. to estimate at regional level poverty rates. You can of course do tests on all these estimands too (e.g. is estimate 1 the same as estimate 2).

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@lakens See e.g. Lohr, S. L. (2021). Sampling: design and analysis. Chapman and Hall/CRC. (available online if you search).

A policy-making example: if you want to estimate unemployment rate quarterly, you can define that a change of say .1% is 'relevant', implying s.e should be ~.05

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@lakens I think we had this discussion before, but in some fields - such as official stats - its is common to think about the desired size of the s.e., e.g. as expressed in the coëfficiënt of variation. Precision is important there.

[WOinActie @WOinactie]

SAVE THE DATE: 25 november in Den Haag nieuwe protesten tegen de bezuinigingen op het hoger onderwijs #woinactie

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@orioljbosch I also really like the Kish book a lot. Shannon Lohr's book is more accessible, also good and has additional R exercises.

[Peter Lugtig - now also on Bluesky @PeterLugtig]

For my Dutch friends. Sign the petition against the massive cuts to higher education/ Teken de petitie en sluit je aan bij protest op 14 November!
https://t.co/cOvWOZiqBJ

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@cjvanlissa @NWONieuws Congratulations!

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@BastiaanRagas @EllenKoenen Wij doen al jaren https://t.co/D7ShYsJVaq. je betaalt er wel wat voor, maar heel goede ervaringen.

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@DieTukkerfries Sarban aan Oudegracht. Afghaans

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@EikoFried Asking sad ---- happy as two items seems weird, unless you want to do latent variable modeling.
That still leaves the uni vs bipolar issue. Lots of literature on this, but inconclusive imo. I think depends on topic and population. Qualitative test before fielding survey?

[Javier Garcia-Bernardo (@javiergbe.bsky.social) @javiergb_com]

Job opportunity! We are looking for a postdoctoral researcher at @UniUtrecht to investigate how social networks may impact people’s vaccination decisions: https://t.co/rj6atUjXT1 Come work with us! Deadline: October 4th

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@CarinaCornesse @gesis_org Woohoo! Congrats Carina and @gesis_org!

[Angelo Moretti @AngeloMoretti3]

📆 The deadline is approaching!
📢 Submit now a session proposal to the 11th Conference of the European Survey Research Association (ESRA) @ESRAsurvey
ℹ More info: https://t.co/2LWy2CqhaV

#ESRA #Conference #Utrecht #Survey #Statistics

[Peter Lugtig - now also on Bluesky @PeterLugtig]

Anyone interested in surveys and data collection in a Big-data age: we are organizing ESRA; a big conference in Utrecht in July 2025. Below is a call for session proposals. Deadline: 15 sept. #esra25

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@patricksturg 😐

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@AngeloMoretti3 @SIStatistica @camillasalvat @MS_Utrecht @camillasalvat Congratulations!

[Bob Sacamano (Almost Blue) @bob_saccamanno]

Don Dillman, a great scholar who influenced the way many of us conduct our surveys, passed away this morning. Here is a heritage video @AAPOR did in 2014 on this remarkable, kind, and wise survey researcher who taught us all. https://t.co/7vCI2aGno6

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@EikoFried And about the composite effect: normalising just transforms the score, but doesnt take away the experiment varies multiple things at once. (the no. of points + the answer type). Would be nice to replicate this with 1-7 VAS.

[Peter Lugtig - now also on Bluesky @PeterLugtig]

@EikoFried One place to start is https://t.co/MQlMvzAHAD. This is a 'meta-analytic' tool of 100s of question-and-answer experiments, including things like the number of answer options, answer scale, end-point labels etc.