What better data opens: Structure, Identification, and Spatial Causality for Urban Research

Prijzen vanaf
23,36

Gesponsorde links

VERGELIJK ALLE AANBIEDERS (2)

Beschrijving

Bol Better data doesn't just add numbers. It opens the question every urban policy is really asking: what did it actually do? Volume 1, Starting With What You Have, showed what can be learned from the data a city already holds - and drew a firm line at claims of cause. This companion volume crosses that line, carefully. It works through the designs that turn association into credible causal evidence: panels built from free satellite archives, difference-in-differences and staggered rollouts, matching, instrumental variables, regression discontinuity, synthetic control, spatial regression, and the spillovers that urban interventions almost always produce. Every method is tested on data where the true answer is known before the analysis is run. Naive estimates miss by as much as 213 per cent; the designs in this book recover the truth, most within five per cent - and every chapter states plainly which untestable assumption that result depends on. In this book, you will learn how to: - Diagnose which causal designs your data can actually support, before you run a single regression - Build a monthly urban panel from free satellite data, going back to 2012, for any city on earth - Estimate treatment effects with difference-in-differences, staggered adoption, matching, and instrumental variables - Use regression discontinuity and synthetic control when no comparison group seems to exist - Measure spillovers instead of assuming - by design - that they don't happen - Follow every method in Python, with R equivalents and a no-code route through free software (GeoDa, QGIS, jamovi, gretl) Each chapter follows the same structure: the concept explained twice, once in plain language and once with the mathematics; the identification assumption stated before the data; a full worked example with real code; and a section on what the design still cannot tell you. Written for readers beginning graduate study and for early-career researchers - especially those who must work where data infrastructure is thin. No prior econometrics is assumed. All code and data are open (MIT / CC0) at github.com/leejoolim2/what-better-data-opens. Every dataset is synthetic, built on real public schemas, with a known answer planted in it - so you can check your own analysis against the book's, not just trust it. What Better Data Opens is Volume 2 in the Urban Research Methods series, following Starting With What You Have.

Vergelijk aanbieders (2)

Gesponsorde links · Sommige shops betalen ons een vergoeding

Sorteren op:

€ 23,36 Gratis verzending

€ 24,00 € 2,99 verzendkosten Totaal € 26,99

Beschrijving (1)

Better data doesn't just add numbers. It opens the question every urban policy is really asking: what did it actually do? Volume 1, Starting With What You Have, showed what can be learned from the data a city already holds - and drew a firm line at claims of cause. This companion volume crosses that line, carefully. It works through the designs that turn association into credible causal evidence: panels built from free satellite archives, difference-in-differences and staggered rollouts, matching, instrumental variables, regression discontinuity, synthetic control, spatial regression, and the spillovers that urban interventions almost always produce. Every method is tested on data where the true answer is known before the analysis is run. Naive estimates miss by as much as 213 per cent; the designs in this book recover the truth, most within five per cent - and every chapter states plainly which untestable assumption that result depends on. In this book, you will learn how to: - Diagnose which causal designs your data can actually support, before you run a single regression - Build a monthly urban panel from free satellite data, going back to 2012, for any city on earth - Estimate treatment effects with difference-in-differences, staggered adoption, matching, and instrumental variables - Use regression discontinuity and synthetic control when no comparison group seems to exist - Measure spillovers instead of assuming - by design - that they don't happen - Follow every method in Python, with R equivalents and a no-code route through free software (GeoDa, QGIS, jamovi, gretl) Each chapter follows the same structure: the concept explained twice, once in plain language and once with the mathematics; the identification assumption stated before the data; a full worked example with real code; and a section on what the design still cannot tell you. Written for readers beginning graduate study and for early-career researchers - especially those who must work where data infrastructure is thin. No prior econometrics is assumed. All code and data are open (MIT / CC0) at github.com/leejoolim2/what-better-data-opens. Every dataset is synthetic, built on real public schemas, with a known answer planted in it - so you can check your own analysis against the book's, not just trust it. What Better Data Opens is Volume 2 in the Urban Research Methods series, following Starting With What You Have.


Productspecificaties

Merk Independently Published
EAN
  • 9798176475449
Maat


Prijshistorie

* Prijshistorie bevat geen data van Amazon.

Prijzen voor het laatst bijgewerkt op:

Uitgelichte Keuze
23,36
Naar shop