Skip to main content

An IERI – International Educational Research Institute Journal

Table 3 Comparison of models for prediction of reading achievement and self-concept

From: The relationship among personal achievement motives, school relational goal structures and learning outcomes: a multilevel analysis with PISA 2018 data

Model

Npar

AIC

BIC

LogLik

China B–S–J–Z (NLevel1 = 11,700, NLevel2 = 360)

    

Predicting reading achievement

MC0: without predictors

3

133,451

133,473

 − 66,722

MC11.:without interaction

12

132,763

132,851

 − 66,369

MC12: with interaction COMPFO*WOKRMAST

14

132,734

132,837

 − 66,353

Predicting reading self-concept

MC0: without predictors

3

29,225

29,247

 − 14,609

MC21: without interaction

12

27,194

27,282

 − 13,585

MC22: with interaction COMPFO*WOKRMAST

14

27,127

27,231

 − 13,550

MC23: with interaction COMPFO*COMPETE

14

27,155

27,259

 − 13,564

USA (NLevel1 = 4170, NLevel2 = 162)

    

Predicting reading achievement

MU0: without predictors

3

50,041

50,060

 − 25,017

MU1: without interaction

12

49,520

49,596

 − 24,747

Predicting reading self-concept

MU0: without predictors

3

11,958

11,977

 − 5976

MU1: without interaction

12

11,399

11,475

 − 5688

  1. Perception variables were aggregated at school level and centered with grand means. COMPFO = perception of competition at school level. Competitiveness, work mastery, fear of failure were centered with group means. WORKMAST = work mastery motive, COMPETE = competitiveness motive. Predictors in models without interactions were economic, social and cultural status, gender, school track, competitiveness, work mastery, fear of failure, aggregated perception of competition and cooperation. For U.S. students the immigration status was also included as predictors in the models. Models without interactions were nested in models with interactions