Skip to main content
. 2021 Jun 23;9(1):14. doi: 10.1186/s40536-021-00107-3

Table 3.

Univariate findings after matching for covariates (PISA 2015; N = 1256)

Construct WS (N = 149) non WS (N = 146)a
M SD M SD d
Covariates for propensity score matching
 Gender (1 = female/0 = male) 0.55 (0.04) 0.50 0.54 (0.04) 0.50 0.02
 Immigration background (1 = yes/0 = no) 0.17 (0.03) 0.38 0.18 (0.04) 0.39 − 0.03
 Index of parental education in years of schooling 15.73 (0.20) 2.39 15.81 (0.19) 1.80 − 0.03
 Index of parental occupational status 67.51 (1.42) 17.13 67.17 (1.69) 18.24 0.02
 Index of cultural possessions at home 1.06 (0.09) 1.06 0.98 (0.12) 1.11 0.08
Analysis variables for mediation modeling
 Inquiry-based science instruction 0.68 (0.09) 0.95 − 0.22 (0.12) 1.10 0.82
 Enjoyment of science 0.26 (0.09) 1.10 − 0.04 (0.15) 1.22 0.24
 Interest in broad science topics 0.53 (0.09) 0.72 0.23 (0.09) 0.99 0.30
 Science achievement 503.30 (6.57) 75.98 534.89 (10.12) 92.96 -0.32

Results are based on imputed manifest scores weighted by PISA total student weight. Standard errors are in parentheses. Cohen's d is computed relative with the SD of the weighted total Austrian PISA 2015 sample before matching. Scores for science achievement are based on Plausible Values; scores for all other analysis variables are based on Warm's Mean Weighted Likelihood Estimates (WLE)

WS Waldorf students

a146 students per data set and 1107 different students across all 10 data sets