Table III.

Utility of attribute levels for the sample and the two preference clusters

Sample (n = 198)Working hours and worktask-oriented type
(n = 17) p =0.95a
Personal notion ofwork-oriented type
(n = 181)
p =0.95a
AttributesLevelsbUtilitycImportancedUtilitycImportancedUtilitycImportanced
Personal notion1. Traditional work approach47.0425.830.2615.363.5332.7
of work2. Employee recognition expressed by high salary10.67 0.00 26.01 
 3. Reputation and public image0.00 14.47 0.00 
 4. Balancing professional
and private aspirations
61.19 33.61 83.55 
Work tasks1. Tasks specific to the company7.895.67.0714.98.093.2
 2. Good task/ compensation ratio11.05 16.49 6.97 
 3. Varied and widely recognized tasks0.00 0.00 0.26 
 4. Tasks manageable with reasonable effort13.25 32.68 0.00 
Working hours1. Core working hours28.4726.452.9731.610.5424.1
 2. Paid overtime35.73 69.51 11.27 
 3. Main business hours0.00 0.00 0.00 
 4. Flexible working hours62.68 56.84 61.64 
Work-related1. Mutual trust56.2923.746.5421.258.2522.8
relations2. Barter system (information, know-how and working times)28.62 36.23 21.53 
 3. Skills-related reputation among colleagues0.00 0.00 0.00 
 4. Shared appreciation of health, society
and environment
29.12 34.88 23.20 
Career opportunities1. Expert among colleagues27.8218.41.1517.142.8417.3
 2. Skills demanded in the job market43.72 37.63 44.30 
 3. Skills-based public reputation and recognition0.00 0.00 0.00 
 4. Job in the local region36.45 29.69 38.02 

Notes:

a

p-values of 95 per cent indicate that the cluster is strongly supported by the data;

b

The attribute levels are based on the respective convention. The family (domestic) convention forms the first level, the market convention makes up the second level, the fame convention is depicted in the third level and the green convention is represented by the fourth attribute level;

c

Arithmetic mean of rescaled part-worths per cluster. The part-worth estimates are rescaled so that the sum of the part-worth values across the five attributes for each respondent equals 500. These estimates do not affect the magnitude of any part-worth, but provide a common scale across all part-worth values for comparison across attribute levels and respondents;

d

Arithmetic mean of importance of each attribute per cluster. The importance of each attribute reflects how much a job attribute influences the choice of a job profile. Important weights are calculated by computing the difference between the largest and the smallest part-worth for each attribute, summing the differences and normalizing to 100. Attribute importance scores sum to 100 across all five attributes for each respondent. Model fit: RChi2 = 2,869.83, df = 15, p 0.001; RMcF2= 0.653.

or Create an Account

Close Modal
Close Modal