Alan Zaslavsky and his course drew me into survey statistics. We targeted on a easy query on the coronary heart of statistics: how can we make statements a couple of inhabitants, given a pattern? We have to signify everybody, as a result of all of us matter and are all distinctive. However not everybody could be in our pattern. As Andrew says, that is what makes it so onerous.
Why have a look at survey knowledge ? As my teammate David Shor says right here, tremendous politically engaged individuals are overrepresented within the media. Survey knowledge supplies a counterbalance to that, aiming to signify everybody.
So let’s bounce in with a new weblog sequence !
Who is that this weblog sequence for ? Primarily of us who already know some survey statistics and wish to study extra collectively. People who’ve heard of, might outline, and even use these ideas, however have questions on them:
- sampling designs like stratification (Sharon Lohr’s Sampling e-book)
- nonrandom (“comfort”) samples (Meng 2018)
- survey weights (Sharon Lohr’s Sampling e-book, Gelman 2007)
- design results (Sharon Lohr’s Sampling e-book)
- poststratification (Holt & Smith 1979) and MRP
- design-based vs model-based inference (Little 2004)
I’ll try and introduce ideas alongside the best way. However there will likely be gaps, which we should always chat about within the feedback. And I’ll ask you questions, so please take part. It’s those that make make survey statistics (and something) nice.
p.s. The title comes from It’s The Individuals: A Pacific Crest Path Movie by Elina Osborne. She quotes the Māori proverb “What’s a very powerful factor ? It’s folks, it’s folks, it’s folks”. I simply did a protracted “solo” hike of the Virginia part of the Appalachian Path, the place I discovered a lot from hikers and path city residents. As a survey statistician, I get to maintain listening and studying.
