Skip to main content

Table 8 Rosenbaum bounds for test of unconfoundedness

From: Wage returns to mid-career investments in job training through employer supported course enrollment: evidence for Canada

Gamma (Γ)

sig+

sig-

t-hat+

t-hat-

CI+

CI-

Diff-in-Diff model for Log Wage: Panel 4

1

0

0

0.087013

0.087013

0.068185

0.105809

1.1

1.80E-11

0

0.064636

0.109332

0.0457

0.128248

1.2

3.10E-06

0

0.044168

0.129788

0.025141

0.148683

1.3

0.004783

0

0.025342

0.148488

0.006238

0.16737

1.34

3.13E-02

0

0.018169

0.155517

−0.00097

0.174479

1.36

0.110433

0

0.01226

0.158212

−0.007433

0.174479

Diff model for Log Wage: Panel 5

1

0

0

0.09267

0.09267

0.074243

0.111161

1.1

6.30E-14

0

0.070572

0.114806

0.052146

0.133467

1.2

6.10E-08

0

0.050499

0.135114

0.031944

0.153673

1.3

0.000405

0

0.031988

0.153632

0.013394

0.172309

1.38

0.028381

0

0.018194

0.167513

−0.00053

0.186268

1.46

0.107181

0

0.011871

0.185605

−0.006911

−0.204624

  1. Note:
  2. 1. People in Panel 4 are tracked from 2002 to 2006 for five years, which is the same as that in Panel 5 where individuals are followed from 2005 to 2009 for five years.
  3. 2. gamma - log odds of differential assignment due to unobserved factors.
  4. 3. sig−/+ − lower and upper bound significance level.
  5. 4. t-hat−/+ − lower and upper bound Hodges-Lehmann point estimate.
  6. 5. CI−/+ − lower and upper bound confidence interval (a = .95).
Back to article page