Skip to main content

Table 8 Regression results using attitude variables based on factor analysis

From: Having a bad attitude? The relationship between attitudes and sickness absence

 

(1)

(2)

(3)

(4)

 

Self-certified

Self-certified

GP-certified

GP-certified

 

coef.

coef.

coef.

coef.

 

(se)

(se)

(se)

(se)

Female

−0.311

−0.219

2.330*

2.439*

 

(0.19)

(.018)

(1.28)

(1.46)

Immigrant

0.001

0.020

0.736

0.302

 

(0.20)

(0.19)

(1.25)

(1.30)

Factor 1: Lenient attitudes towards sick leave

 

0.314***

 

0.339

  

(0.10)

 

(0.73)

Factor 2: Low intrinsic motivation

 

0.192*

 

−0.232

  

(0.11)

 

(0.81)

Factor 3: Lenient towards cheating

 

−0.076

 

0.834

  

(0.11)

 

(0.78)

Factor 4: Being stressed

 

0.103

 

0.326

  

(0.14)

 

(0.89)

Factor 5: High homeload

 

−0.240

 

1.239

  

(0.15)

 

(1.07)

_cons

5.402***

5.918***

18.001*

18.920*

Controls a

Yes

Yes

Yes

Yes

Attitudes

No

Yes

No

Yes

R-squared

0.157

0.188

0.133

0.150

Low education

0.703***

0.668***

1.211

0.642

 

(0.22)

(0.21)

(1.20)

(1.19)

Controls b

Yes

Yes

Yes

Yes

Attitudes

No

Yes

No

Yes

R-squared

0.154

0.184

0.135

0.150

Old

−0.075

0.046

2.617*

3.052*

 

(0.22)

(0.23)

(1.37)

(1.58)

Controls c

Yes

Yes

Yes

Yes

Attitudes

No

Yes

No

Yes

R-squared

0.141

0.170

0.107

0.119

N d

226

226

226

226

  1. *p < 0.10, **p < 0.05, ***p < 0.01
  2. aage, age 2, education, na_educ, single_hh, na_singlehh, child_cust, na_custody, immig_sec, na_immig, student, secondjob, director, shifts, district_a, ass_living, mental_health
  3. bfemale, age, age 2, na_educ, single_hh, na_singlehh, child_cust, na_custody, immig_sec, immigrant, na_immig, student, secondjob, director, shifts, district_a, ass_living, mental_health
  4. cfemale, education, na_educ, single_hh, na_singlehh, child_cust, na_custody, immig_sec, immigrant, na_immig, student, secondjob, director, shifts, district_a, ass_living, mental_health
  5. dWe use frequency weights in the regressions due to the fact that we observe the employees for varying amounts of time, putting more weight on the employees that we have more information about. The frequency weights duplicate the observations according to the employees’ number of working days. This means that the total number of observations, if we take the frequency weights into account, is 104,229. We use cluster robust standard errors to count for the serial correlation that occurs due to the frequency weights
Back to article page