Skip to main content
. Author manuscript; available in PMC: 2012 Jul 26.
Published in final edited form as: J Teach Phys Educ. 2010 Oct 1;29(4):399–416. doi: 10.1123/jtpe.29.4.399

Table 3.

Multilevel Models Predicting Student Engagement in Physical Education

Student engagement in PE
Variables Student
only
(Model 1)
Class
content
only
(Model 2)
Student &
class content
(Model 3)
Intercept terms
Reference intercept (% class time) 100.85 100.88 100.89
Inactive instruction −.5910** −.5806**
Fitness −.1376 −.1253
Skill practice .1168 .1199
Game play −.2014* −.1878*
Class management .0112 .0108
Slope terms
Perceived competence in PEa 9.3693**** 9.2808****
Body imageb 1.0015* 1.0686*
Perceived competence in PE × inactive instruction .4429***
Perceived competence in PE × fitness .0443
Perceived competence in PE × skills practice −.1766*
Perceived competence in PE × game play .0542
Perceived competence in PE × class management .0027
Body image × inactive instruction −.0210
Body image × fitness −.0106
Body image × skills practice −.0133
Body image × game play −.0203
Body image × class management −.0009

Notes: Unstandardized regression coefficients; continuous class content variables were centered at the grand mean and student-level variables were group mean centered (within teacher-group);

*

p<.05;

**

p<.0l;

***

p<.001;

****

p<.0001

a

Perceived competence in PE (M = 3.84; SD = 0.98; range: 1–5);

b

Body image (M = 3.96; SD = 0.90; range: 1–5).