Average Marginal Effects (AME) for selection
| Phase I | Phase II | |||
|---|---|---|---|---|
| Factor | AME (pp) | p-value | AME (pp) | p-value |
| Log (headcount) | 2.54 | 0.00 | 1.03 | 0.28 |
| Initial TRL | 0.69 | 0.13 | 3.57 | 0.00 |
| Anticipated TIL | −1.85 | 0.00 | −2.05 | 0.12 |
| Year fixed effects | Yes | Yes | ||
| Observations | 8097 | 1657 | ||
| Tjur D | 0.02 | 0.09 | ||
| Phase I | Phase | |||
|---|---|---|---|---|
| Factor | ||||
| Log (headcount) | 2.54 | 0.00 | 1.03 | 0.28 |
| Initial | 0.69 | 0.13 | 3.57 | 0.00 |
| Anticipated | −1.85 | 0.00 | −2.05 | 0.12 |
| Year fixed effects | Yes | Yes | ||
| Observations | 8097 | 1657 | ||
| Tjur D | 0.02 | 0.09 | ||
The probability of selection estimated with logistic regressions, using 2009–2016 NASA SBIR/STTR proposal data. Initial TRL Technology Readiness Level as reported by the PI in the proposal. Anticipated TIL Anticipated TRL Improvement Level as stated in the proposal. The treatment group is microfirms (1-10 employees) and untreated observations are weighted with the entropy balancing algorithm using the firm history and the number of patents as covariates. Standard errors are clustered by firm. Average marginal effects (AMEs) are reported in percentage points. The figure of merit is Tjur’s coefficient of discrimination D
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.