Skip to main content

Table 6 Full conditional models of competence development in Accounting

From: Determinants of competence development in accounting in upper secondary education

Fixed Effects

Level 1 (within-student)

Level 2 (between-student)

Level 3 (between-class)

   

Est.

p

Predicting initial status

Predicting growth

Predicting initial status

Predicting growth

R2

AIC

BIC

Est.

p

Est.

p

Est.

p

Est.

p

%

  

M1: maths

na

na

0.09

0.032

−0.05

0.134

0.57

0.000

0.34

0.001

39

1091

1294

M2: grades

0.00

0.917

−0.28

0.000

−0.05

0.128

−0.40

0.001

−0.16

0.028

48

1085

1296

M3: asc

−0.02

0.634

0.04

0.458

0.02

0.647

0.11

0.461

−0.02

0.727

39

1095

1306

M4: ssat

0.02

0.478

−0.01

0.756

0.04

0.215

0.22

0.083

0.05

0.543

43

1093

1305

M5: intri

0.01

0.809

0.11

0.047

−0.03

0.509

0.09

0.681

−0.05

0.750

39

1095

1306

M6: wfeff

0.01

0.855

−0.01

0.692

−0.03

0.413

0.09

0.431

−0.02

0.737

39

1094

1305

M7: comp

0.07

0.027

−0.06

0.431

0.01

0.744

0.32

0.138

0.00

0.988

40

1094

1305

M8: auto

−0.02

0.625

0.05

0.418

0.01

0.714

0.27

0.089

0.02

0.899

40

1094

1305

M9: relat

0.05

0.191

−0.01

0.783

0.02

0.426

−0.01

0.965

0.05

0.624

39

1095

1306

M10: struc

−0.04

0.270

0.05

0.432

−0.03

0.369

0.35

0.000

0.12

0.036

47

1087

1298

M11: coak

0.04

0.087

−0.08

0.051

0.03

0.261

−0.03

0.854

0.01

0.937

39

1095

1307

M12: coau

0.02

0.402

−0.07

0.079

−0.01

0.727

0.41

0.000

0.25

0.000

39

1067

1279

M13: taska

−0.03

0.389

0.08

0.226

−0.01

0.775

0.31

0.060

0.10

0.059

49

1085

1296

  1. Levels 1, 2, and 3 comprise time-varying predictors, time-invariant student variables and context effects, respectively. All models M1 to M2 include all level 1 and level 2 models, but only the class mean in mathematics and the class mean of the variable indicated in the model’s name. Italic values indicate significant at a level of 5 % error probability. Underlined italic values indicate significant at a level of 10 % error probability. Pseudo R2 was calculated according to Singer and Willett (2003) by taking the square of the correlation between the observed WBB and the predicted WBB. Prediction was based on fixed effects only. See Table 2 for abbreviations