Skip to main content

Table 2 Hierarchical linear model

From: “…Because we do so together”: A Mixed-Methods Analysis of Assistant Teacher’s Work Environment, Conditions, and Teamwork Experiences

 

Coefficient

Std. error

(Within effects) dimension scores

 Observation

0.064

0.067

 Materials

− 0.003

0.074

 Support services for children and families

0.310**

0.112

 Staffing

− 0.043

0.087

 Professional development

− 0.051

0.058

 Applying learning

0.446**

0.154

 Decision-making

0.065

0.058

 Input

0.035

0.106

 Economic well-being

0.146 + 

0.080

 Wellness supports

− 0.095

0.120

 Quality of work life

0.140

0.109

(Between effects) mean dimension scores

  

 Observation

0.068

0.043

 Materials

− 0.065

0.051

 Support services for children and families

0.168***

0.064

 Staffing

− 0.070

0.051

 Professional development

− 0.028

0.045

 Applying learning

0.611***

0.104

 Decision-making

0.102**

0.041

 Input

0.059

0.076

 Economic well-being

− 0.031

0.062

 Wellness supports

0.181*

0.085

 Quality of work life

0.072

0.083

Sociodemographic characteristics

 Wage per hour

0.006

0.013

 Educational attainment (ref. category: some college or less)

  Associate degree

0.066

0.118

  Bachelor’s degree or higher

0.011

0.131

 Race (ref. category: Hispanic)

  

  POC

− 0.051

0.143

  White

− 0.120

0.104

Center-level variables

  

 Bilingualism at the center (ref. category: no one is bilingual)

  At least one educator is bilingual

− 0.166

0.102

  All are bilingual

− 0.208

0.210

 Educational attainment of the staff at the center (ref. category: no one has a BA)

  At least one teaching staff has a BA

0.122

0.119

  All teaching staff has a BA

0.245

0.220

 Intercept

0.339

0.474

Goodness-of-fit statistics

  

 R-square level 1

0.5152

 

 R-square level 2

0.5190

 

Observations

266

 
  1. Dependent variable Teamwork and SEQUAL dimensions are measured in a 1–6 scale from strongly disagree (1) to strongly agree (6)
  2. Sensitivity analysis in Additional file 1: Table S1 suggests a limited impact of missing data since the descriptive statistics in our model do not substantially change from the full sample
  3. ***p < 0.001, **p < 0.01, *p < 0.05, +p < 0.1