Tiêu đề 9.3. Matching.pdf ✅

pacmann.io © 2022 – Pacmann AI 13 Matching
pacmann.io © 2022 – Pacmann AI 14 Matching Goal ● Goal of matching: Make treatment and control groups comparable. ● Matching methods assign weights to observations to achieve better comparability. ● To assess the effect of a treatment, compute a weighted average of the results observed in both the treatment and control groups. ● By comparing these averages, we can estimate the effect of the treatment.
pacmann.io © 2022 – Pacmann AI 15 Weighted Mean ● Weighted mean calculation is done by multiply each observation by its weight. ● Add up the weighted values and divide by the sum of weights. ● Illustration: ○ Equal weight: values 1, 2, 3, and 4 with equal weights will have mean the same as we calculate usual mean. Their mean is 2.5 val weight 1 1 2 1 3 1 4 1 10 4
pacmann.io © 2022 – Pacmann AI 16 Weighted Mean ● Weighted mean calculation is done by multiply each observation by its weight. ● Add up the weighted values and divide by the sum of weights. ● Illustration: ○ Equal weight: values 1, 2, 3, and 4 with equal weights will have mean the same as we calculate usual mean ○ Difference weight: values 1, 2, 3, and 4 will have different weights. Their mean would be 2.6 val weight 1 0.5 2 2 3 1.5 4 1 13 5