Using inverse probability weights and propensity scores ( Fitzgerald et al ., 1998 ; Cuddeback et al ., 2004 ; Foster and Fang, 2004 ; Miller and Hollist, 2007 ; Baulch and Quisumbing, 2011 ), we found that attrition bias did not substantially alter our conclusions, and it affected only a few estimates close to the margin of significance ( p = 0.05).
← all excerpts
The Effects of Doctoral Teaching Development on Early-Career STEM Scholars' College Teaching Self-efficacy.
1
0.0500
0.0500