Skip to main content
. 2021 Jun 26;14(1):9. doi: 10.1186/s12052-021-00148-w

Table 2.

Multiple linear regression model results for post-class tree thinking concept inventory scores

TTCI Post-Score (Model 1) TTCI Post-Score (Model 2)
CC MI CC MI
Β
(Std. error)
p-value Β
(Std. error)
p-value Β
(Std. error)
p-value Β
(Std. error)
p-value
Intercept

1.96

(1.44 – 2.49)

 < 0.001

1.895

(1.66 – 2.13)

 < 0.001

2.50

(1.86 – 3.14)

 < 0.001

2.428

(2.14–2.72)

 < 0.001
TTCI Pre-Score

0.59

(0.49 – 0.69)

 < 0.001

0.606

(0.558 – 0.654)

 < 0.001

0.46

(0.32 – 0.60)

 < 0.001

0.468

(0.40 - 0.54)

 < 0.001
Year (Reference: Year 1)

− 0.40

(− 0.82 to 0.01)

0.059

− 0.22

(– 0.39 to – 0.05)

0.197

− 0.41

(– 0.83 to 0.00)

0.050

− 0.254

(− 0.42 – 0.09)

0.134
I-SEAH (Centered)

0.02

(− 0.03 to 0.08)

0.393

0.032

(0.01 – 0.06)

0.196

0.01

(– 0.03 to 0.05)

0.560

0.012

(− 0.01 – 0.03)

0.502
Treatment (Reference: Animal)

0.35

(− 0.07 to 0.77)

0.102

0.281

(0.09 – 0.48)

0.153

− 0.77

(– 1.65 to 0.12)

0.090

− 0.692

(– 1.094 to − 0 .29)

0.086
Treatment X I-SEAH

− 0.03

(− 0.11 to 0.05)

0.477

− 0.042

(− 0.08 − 0.01)

0.231 - - - -
Treatment X TTCI Pre-Score - - - -

0.28

(0.08 – 0.47)

0.006

0.249

(0.157 – 0.342)

0.008
Observations 282 673 275 673
R2 / adjusted R2 0.343/0.331 0.376 / 0.364

Full model results predicting student TTCI post-scores using complete-case data (CC) and multiple imputation (MI). Year and treatment are binary variables with reference values in parentheses. MI results are pooled following Rubin (2004). Coefficients significant at p ≤ 0.05 are bolded