Skip to contents

Hypothetical potential outcomes for 30 policing hotspots, depending on their own treatment status (Z) and on whether a nearby hotspot is treated (A). Because exposure to spillover is itself random, each hotspot has a probability of landing in each of the four exposure conditions.

Usage

table_8_4

Format

A data frame with 30 rows and 17 columns:

hotspot

Hotspot number, 1 to 30

x, y

Coordinates in the plane

hotspots_within_500m

Number of other hotspots within 500m

hotspots_within_750m

Number of other hotspots within 750m

A

Realized spillover exposure, "0" or "1"

Z

Realized direct treatment, "0" or "1"

exposure

Two-character exposure condition, A followed by Z

prob_A_0_Z_0, prob_A_0_Z_1, prob_A_1_Z_0, prob_A_1_Z_1

Probability of each exposure condition under the randomization

Y_A_0_Z_0, Y_A_0_Z_1, Y_A_1_Z_0, Y_A_1_Z_1

Potential outcome in each exposure condition

Y

Revealed outcome

Source

ISPS Data Archive study D081 doi:10.60600/YU/VUH8KD , file 42273 (GerberGreenBook_Chapter8_Table_8_4_8_5). The coordinates are not stored in the archive as data; they are regenerated from the set.seed(1337) draw in the archive's own Chapter 8 figure code (file 42377). Built by data-raw/book_tables.R.